Abstract
Chapter 4 continues Chap. 3, utilizing the multi-response permutation procedures developed in Chap. 2 for analyzing completely randomized data at the interval level of measurement. In Chap. 4, multi-response permutation procedures are used to analyze regression residuals generated by ordinary least squares (OLS) and least absolute deviation (LAD) regression models. Experimental designs presented and analyzed in Chap. 4 include one-way randomized, one-way randomized with a covariate, one-way randomized-block, two-way randomized-block, two-way factorial, Latin square, split-plot, and two-factor nested designs.
Access provided by Autonomous University of Puebla. Download chapter PDF
Similar content being viewed by others
Multi-Response Permutation Procedures (MRPP) were introduced in Chap. 2 and applied to interval-level, completely randomized data in Chap. 3. While multi-response permutation procedures are generally thought of as providing tests of differences among g treatment groups as demonstrated in Chap. 3, they also have applications in ordinary least squares (OLS) linear regression analyses with v = 2 and least absolute deviations (LAD) linear regression analyses with v = 1. In this fourth chapter of Permutation Statistical Methods, MRPP analyses of LAD regression residuals are illustrated with a variety of experimental designs, including one-way completely randomized with and without a covariate, one-way and two-way randomized-block, two-way factorial, Latin square, and two-factor nested analysis-of-variance designs. Also considered are multivariate multiple regression designs.
4.1 LAD Linear Regression
OLS linear regression has long been recognized as a useful tool in many fields of research. The optimal properties of OLS regression are well known when the errors are normally distributed. However, in practice the assumption of multivariate normality is rarely justified. LAD linear regression is an attractive alternative to OLS regression as it is extremely robust to deviations from normality as well as to the presence of extreme values [297, p. 172].
It is widely recognized that estimators of OLS regression parameters can be severely affected by unusual values in either the criterion variable or in one or more of the predictor variables. This is due in large part to the weight given to each data point when minimizing the sum of squared errors. In contrast, LAD regression is much less sensitive to the effects of unusual-value errors due to the fact that the errors are not squared. Moreover, LAD regression has been shown to be superior to OLS regression when errors are generated from heavy-tailed or outlier-producing distributions, such as the Cauchy and double-exponential distributions; see, for example, articles by Blattburg and Sargent [46], Dielman [94, 95], Dielman and Pfaffenberger [96], Dielman and Rose [97], Mathew and Nordström [264], Mielke , Berry , Landsea , and Gray [303], Pfaffenberger and Dinkel [337], Rice and White [346], Rosenberg and Carlson [352], Rousseeuw [355], Taylor [394], and Wilson [432].
As described by Sheynin , the initial known use of regression by Daniel Bernoulli (c. 1734) for astronomical prediction problems involved LAD regression based on ordinary Euclidean distances between the observed and predicted response values [372]. Further developments in LAD regression were due to Roger Joseph (Rogerius Josephus) Boscovich (c. 1755), Pierre-Simon Laplace (c. 1789), and Carl Friedrich Gauss (c. 1809). The American mathematician and astronomer Nathaniel Bowditch (c. 1809) was highly critical of OLS regression because, as he argued, squared regression residuals unduly emphasized questionable observations in comparison with the absolute regression residuals associated with LAD regression [372].
Consider the general multivariate regression model given by
where \(\mathbf{y}_{i}^{\,{\prime}} = (y_{1i},\,\ldots,\,y_{ri})\) denotes the row vector of r observed response measurements for the ith of N objects, \(\mathbf{x}_{i}^{\,{\prime}} = (x_{1i},\,\ldots,\,x_{si})\) is the row vector of s predictor values for the ith object, \(\boldsymbol{\beta }^{{\prime}} = (\beta _{1},\,\ldots,\,\beta _{t})\) is the row vector of t parameters, \(\boldsymbol{h}^{{\prime}} = (h_{1},\,\ldots,\,h_{r})\) is the row vector of r model functions of \(\boldsymbol{\beta }\) and x i for the ith object, and \(\mathbf{e}_{i}^{\,{\prime}} = (e_{1i},\,\ldots,\,e_{ri})\) denotes the r errors between the response variables and model functions for the ith object, i = 1, …, N objects. The special case of a multivariate linear regression model is given by
where \(\boldsymbol{f}(\mathbf{x}_{i})\) denotes a column vector of p distinct functions of s predictors (x i ) for the ith object, i = 1, …, N, and B is an r×p matrix of parameters in which (B j1, …, B jp ) is the row vector of p parameters associated with the jth response measurement, j = 1, …, r.
Let y i denote a column vector of r observed response measurement scores and let \(\tilde{\mathbf{y}}_{i}\) denote a column vector of r predicted response values for the ith object, i = 1, …, N. Thus, the general and linear predicted multivariate regression models are given by
respectively, where \(\boldsymbol{\tilde{\beta }}\) and \(\tilde{\mathbf{B}}\) are estimated parameters that are intended to provide good fits between the y i and \(\tilde{\mathbf{y}}_{i}\) values relative to a selected goodness-of-fit criterion. The null hypothesis (H 0) underlying each criterion dictates that each of the N! possible, equally-likely pairings of the predicted sequential ordering (\(\tilde{\mathbf{y}}_{1},\,\ldots,\,\tilde{\mathbf{y}}_{N}\)) with the fixed observed sequential ordering (y 1, …, y N ) occurs with equal probability, i.e., 1∕N! .
Let \(\Delta (\mathbf{\tilde{y}_{i}},\mathbf{y}_{i})\) for i = 1, …, N denote the distance function between the predicted and observed response measurement values and consider the generalized Minkowski distance function given by
where w ≥ 1 and v > 0. Since v = 1 yields the Minkowski metric [12], the choice of v = 1 is preferred since v > 1 yields distance functions that do not satisfy the triangle inequality property of a metric. Consequently, the distance function of choice utilizes v = 1 and w = 2, i.e., an ordinary Euclidean distance function.
Let the average distance function between \((\tilde{\mathbf{y}}_{1},\,\ldots,\,\tilde{\mathbf{y}}_{N})\) and \((\mathbf{y}_{1},\,\ldots,\,\mathbf{y}_{N})\) be given by
As noted previously, a distance function with v > 1 is not a metric function. If the distance function associated with LAD regression is squared (i.e., v = 2), then the estimated parameters that minimize δ yield an OLS regression model.
The criterion for fitting multivariate regression models based on δ is the chance-corrected measure of agreement between the observed and predicted response measurement values given by
where μ δ is the expected value of δ over the N! possible pairings under the null hypothesis. An efficient computational expression for obtaining μ δ that involves a sum of N 2 rather than N! terms is given by
4.1.1 Linear Regression and Agreement
A simple interpretation of \(\mathfrak{R}\) can be described for \(r = s = 1\) since the same interpretation holds for any r and s. In the case involving perfect agreement, \(\tilde{y}_{i} = y_{i}\) for i = 1, …, N, δ = 0. 00, and \(\mathfrak{R} = 1.00\). This implies that the functional relationship between \(\tilde{y}\) and y can be described by a straight line that passes through the origin with a slope of 45∘, as depicted in Fig. 4.1 with N = 5 bivariate \((y,\,\tilde{y})\) values: (2, 2), (4, 4), (6, 6), (8, 8), and (10, 10). For the N = 5 data points depicted in Fig. 4.1, the intercept is \(\tilde{\beta }_{0} = 0.00\), the unstandardized slope is \(\tilde{\beta }_{1} = +1.00\), the squared Pearson product-moment correlation coefficient is \(r_{y\tilde{y}}^{2} = +1.00\), and the agreement percentage is also 1. 00, i.e., all five of the y and \(\tilde{y}\) paired values agree.
In this context, the squared Pearson product-moment correlation coefficient, \(r_{y\tilde{y}}^{2}\), has also been used as a measure of agreement. However, \(r_{y\tilde{y}}^{2} = +1.00\) implies a linear relationship between y and \(\tilde{y}\), where both the intercept and slope are arbitrary. While perfect agreement is described by \(\mathfrak{R} = +1.00\), \(r_{y\tilde{y}}^{2} = +1.00\) describes a linear relationship that may or may not reflect perfect agreement as depicted in Fig. 4.2 with N = 5 \((y,\,\tilde{y})\) values: (2, 4), (4, 5), (6, 6), (8, 7), and (10, 8). For the N = 5 bivariate data points depicted in Fig. 4.2, the intercept is \(\tilde{\beta }_{0} = +3.00\), the unstandardized slope is \(\tilde{\beta }_{1} = +0.50\), the squared Pearson product-moment correlation coefficient is \(r_{y\tilde{y}}^{2} = +1.00\), and the agreement percentage is 0. 20, i.e., only one (6, 6) of the N = 5 y and \(\tilde{y}\) paired values agree.
Comparisons of \(\mathfrak{R}\) with other measures of agreement and the advantages of \(\mathfrak{R}\) relative to the other agreement measures were detailed in a 1996 article by Watterson [416].
While the agreement measure \(\mathfrak{R}\) provides a description of the functional relationship between \((\tilde{\mathbf{y}}_{1},\,\ldots,\,\tilde{\mathbf{y}}_{N})\) and \((\mathbf{y}_{1},\,\ldots,\,\mathbf{y}_{N})\), it does not indicate how extreme an observed value of \(\mathfrak{R}\), say \(\mathfrak{R}_{\text{o}}\), is relative to the N! possible values of \(\mathfrak{R}\) under the null hypothesis. Since μ δ is invariant under the null hypothesis and the observed value of δ is given by
the exact probability value for \(\mathfrak{R}_{\text{o}}\) is given by
where M = N! . Because an exact probability value requires generating N! arrangements of the observed data, calculation of an exact value is prohibitive even for small values of N, e.g., \(M = N! = 15! = 1,307,674,368,000\).
When M is very large, an approximate probability value for δ may be obtained from a resampling permutation procedure. Let L denote a random sample of all possible arrangements of the observed data, where L is typically a large number, e.g., L = 1, 000, 000. Then, an approximate resampling probability value is given by
Also, when M is very large and P is exceedingly small, a resampling-approximation permutation procedure based on fitting the first three exact moments of the discrete permutation distribution to a Pearson type III distribution provides approximate probability values, as detailed in Chap. 1, Sect. 1.2.2; see also references [284] and [300].
4.2 Example LAD Regression Analyses
In this section, example analyses illustrate the permutation approach to typical multiple regression problems. The first example analyzes a small set of multivariate response measurement scores using LAD regression and generates a resampling permutation probability value; the second example analyzes the same small set of multivariate response measurement scores using OLS regression and also generates a resampling permutation probability value; the third example analyzes the same set of multivariate response measurement scores using OLS regression, but provides a conventional approximate probability value based on Snedecor’s F distribution.
4.2.1 Example Analysis 1
Consider the multiple regression data listed in Fig. 4.3 where s = 2 observed response measurement scores have been obtained for each of N = 12 objects, y 1, …, y N denotes the observed response measurement scores for the N objects, and \(\mathbf{x}_{i}^{\,{\prime}} = (x_{1i},\,\ldots,\,x_{2i})\) is the row vector of s = 2 predictor variables for the ith of N objects.
Because there are \(M = 12! = 479,001,600\) possible, equally-likely arrangements of the N = 12 multivariate response measurement scores in Fig. 4.3, an exact permutation approach is impractical and a resampling procedure is mandated.
A LAD regression analysis of the multivariate response measurement scores listed in Fig. 4.3 yields estimated regression coefficients of
Footnote 1 Figure 4.4 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 12.
Following Eq. (4.1) on p. 125 with v = 1, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.4 is δ o = 1. 50.
If all M possible arrangements of the N = 12 observed LAD regression residuals listed in Fig. 4.4 occur with equal chance, the approximate resampling probability value of δ o = 1. 50 calculated on L = 1, 000, 000 random arrangements of the observed LAD regression residuals is
Following Eq. (4.3) on p. 117, the exact expected value of the M = 479, 001, 600 δ values is μ δ = 1. 8294 and, following Eq. (4.2) on p. 117, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is
indicating 18 % agreement between the observed and predicted y values above that expected by chance.
4.2.2 Example Analysis 2
For a second example analysis of the multivariate response measurement scores listed in Fig. 4.3 on p. 120, consider an OLS regression analysis based on a resampling permutation procedure. An OLS regression analysis of the multivariate response measurement scores listed in Fig. 4.3 yields estimated regression coefficients of
Figure 4.5 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 12.
Following Eq. (4.1) on p. 117 with v = 2, the observed value of the MRPP test statistic computed on the OLS regression residuals listed in Fig. 4.5 is δ o = 3. 1502. If all M possible arrangements of the N = 12 observed OLS regression residuals listed in Fig. 4.5 occur with equal chance, the approximate resampling probability value of δ o = 3. 1502 computed on L = 1, 000, 000 random arrangements of the observed OLS regression residuals is
For comparison, the approximate resampling probability value based on LAD regression in Example 1 is P = 0. 0191.
Following Eq. (4.3) on p. 117, the exact expected value of the M = 479, 001, 600 δ values is μ δ = 5. 2942 and, following Eq. (4.2) on p. 117, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating approximately 41 % agreement between the observed and predicted y values above that expected by chance.
4.2.3 Example Analysis 3
Finally, consider a conventional OLS regression analysis of the multivariate response measurement scores listed in Fig. 4.3 on p. 120. An OLS regression analysis yields estimated regression coefficients of
the regression residuals are listed in Fig. 4.5, and the observed squared multiple correlation coefficient is \(R_{y.x_{1},x_{2}}^{2} = 0.2539\). \(R_{y.x_{1},x_{2}}^{2}\) may be transformed into an F-ratio by
Assuming independence, normality, and homogeneity of variance, F is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = s = 2\) and \(\nu _{2} = N - s - 1 = 12 - 2 - 1 = 9\) degrees of freedom. Under the null hypothesis, the observed value of F o = 1. 5313 yields an approximate probability value of P = 0. 2677.
Note that the asymptotic probability value based on OLS regression in Example 3 is P = 0. 2677, while a resampling analysis of the same data in Example 2 yielded a probability value, again based on OLS regression, of P = 0. 0961, a marked difference. Moreover, a LAD regression analysis of the same data in Example 1 yielded an approximate resampling probability value of P = 0. 0191, once again demonstrating the different results possible with v = 1 and v = 2, both with and without a permutation analysis.
4.3 LAD Regression and Analysis of Variance Designs
It is well known that experimental designs that would ordinarily be analyzed by some form of analysis of variance can also be analyzed by OLS multiple regression using either dummy- or effect-coding schemes. The same is true of LAD regression. In this section a variety of analysis-of-variance designs are analyzed using MRPP, LAD regression, and either dummy or effect coding of treatment groups; included are one-way randomized, one-way randomized with a covariate, one-way randomized-block, two-way randomized-block, two-way factorial, Latin square, split-plot, and two-factor nested analysis-of-variance designs.
4.3.1 One-Way Randomized Design
Consider a one-way completely randomized experimental design with fixed effects in which N = 26 objects have been randomly assigned to one of g = 3 treatment groups with n 1 = 8 and \(n_{2} = n_{3} = 9\). The design and data are adapted from Stevens [387, p. 70] and are given in Fig. 4.6.
For a one-way randomized experimental design, the appropriate regression model is given by
where y i denotes the ith of N responses possibly affected by a treatment; x ij is the jth of m covariates associated with the ith response, where x i1 = 1 if the model includes an intercept; β j denotes the jth of m regression parameters; and e i designates the error associated with the ith of N responses. If the estimates of β 1, …, β m that minimize
are denoted by \(\tilde{\beta }_{1},\,\ldots,\,\tilde{\beta }_{m}\), then the N residuals of the LAD regression model are given by \(e_{i} = y_{i} -\tilde{ y}_{i}\) for i = 1, …, N, where the predicted value of y i is given by
In contrast, OLS regression estimators of β 1, …, β m minimize
the N residuals of the OLS regression model are given by \(e_{i} = y_{i} -\hat{ y}_{i}\) for i = 1, …, N, and the predicted value of y i is given by
If the N regression residuals are partitioned into g disjoint treatment groups of sizes n 1, …, n g , where n i ≥ 2 for i = 1, …, g and
then the permutation test depends on test statistic
where
is a positive weight for the ith of g treatment groups that minimizes the variability of δ,
and \(\xi _{i}\) is the average pairwise Euclidean difference among the n i residuals in the ith of g treatment groups defined by
where v = 1 for LAD regression and
The null hypothesis specifies that each of the
allocations of the N residuals to the g treatment groups is equally likely with n i , i = 1, …, g, residuals preserved for each arrangement of the observed data. The exact probability value of an observed value of δ, δ o, is given by
As previously, when M is large, an approximate probability value of δ may be obtained from a resampling procedure, where
and L denotes the number of resampled test statistic values. Typically, L is set to a large number to ensure accuracy, e.g., L = 1, 000, 000. When M is very large and P is exceedingly small, a resampling-approximation permutation procedure may produce no δ values equal to or less than δ o, even with L = 1, 000, 000, yielding an approximate resampling probability value of P = 0. 00. In such cases, moment-approximation permutation procedures based on fitting the first three exact moments of the discrete permutation distribution to a Pearson type III distribution provide approximate probability values, as detailed in Chap. 1, Sect. 1.2.2 [284, 300].
An index of the effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is given by the chance-corrected measure
where μ δ is the arithmetic average of the δ values calculated on all M equally-likely arrangements of the observed response measurements, i.e.,
A design matrix of dummy codes for an MRPP regression analysis of the N = 26 response measurement scores in Fig. 4.6 is given in Fig. 4.7 where the first columns of 1 values provide for an intercept. The second columns contain the N = 26 univariate response measurement scores listed according to the original random assignment of the N = 26 objects to the g = 3 treatment groups with the first n 1 = 8 scores, the next n 2 = 9 scores, and the last n 3 = 9 scores associated with the first, second, and third treatment groups, respectively.
Because the purpose of the analysis is to test for possible differences among the g = 3 treatment groups, a reduced regression model is constructed without a variate for treatments. Therefore, for a single-factor experiment the design matrix for the reduced model is composed solely of a code for the intercept. The MRPP regression analysis examines the N = 26 regression residuals for possible differences among the g = 3 treatment levels; consequently, no dummy codes for treatments are included in Fig. 4.7 as this information is implicit in the ordering of the g = 3 treatment groups in the three columns labeled “Score” with n 1 = 8 and \(n_{2} = n_{3} = 9\) values.
An exact permutation solution is impractical for the univariate response measurements listed in Fig. 4.7 since there are
possible, equally-likely arrangements of the N = 26 univariate response measurement scores; consequently, a resampling procedure is the default in this case.
LAD Regression Analysis
An MRPP resampling analysis of the LAD regression residuals calculated on the univariate response measurement scores listed in Fig. 4.7 yields an estimated LAD regression coefficient of \(\tilde{\beta }_{0} = +12.00\). Figure 4.8 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 26.
Following Eq. (4.5) on p. 125 and employing ordinary Euclidean distance between residuals with v = 1, the N = 26 LAD regression residuals listed in Fig. 4.8 yield g = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.8 with v = 1 and treatment-group weights
is
If all M possible arrangements of the N = 26 observed LAD regression residuals listed in Fig. 4.8 occur with equal chance, the approximate resampling probability value of δ o = 5. 2308 computed on L = 1, 000, 000 random arrangements of the observed LAD regression residuals with n 1 = 8 and \(n_{2} = n_{3} = 9\) residuals preserved for each arrangement is
Following Eq. (4.7) on p. 126, the exact expected value of the Mδ values is μ δ = 6. 1262 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is
indicating approximately 15 % agreement between the observed and predicted y values above that expected by chance.
OLS Regression Analysis
For comparison, consider an MRPP resampling analysis of OLS regression residuals calculated on the N = 26 univariate response measurement scores listed in Fig. 4.7 on p. 127. The MRPP regression analysis yields an estimated OLS regression coefficient of \(\hat{\beta }_{0} = +14.2692\). Figure 4.9 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 26.
Following Eq. (4.5) on p. 125 and employing squared Euclidean distance between residuals with v = 2, the N = 26 OLS regression residuals listed in Fig. 4.9 yield g = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the OLS regression residuals listed in Fig. 4.9 with v = 2 and treatment-group weights
is
If all M possible arrangements of the N = 26 observed OLS regression residuals listed in Fig. 4.9 occur with equal chance, the approximate resampling probability value of δ o = 53. 6425 computed on L = 1, 000, 000 random arrangements of the observed OLS regression residuals with n 1 = 8 and \(n_{2} = n_{3} = 9\) residuals preserved for each arrangement is
For comparison, the approximate resampling probability value based LAD regression, v = 1, L = 1, 000, 000, and \(C_{i} = n_{i}/N\) for i = 1, 2, 3 is P = 0. 0121.
Following Eq. (4.7) on p. 126, the exact expected value of the M = 75, 957, 810, 500 δ values is μ δ = 60. 5692 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating approximately 11 % agreement between the observed and predicted y values above that expected by chance.
Conventional ANOVA Analysis
A conventional fixed-effects one-way analysis of variance calculated on the N = 26 univariate response measurement scores listed in Fig. 4.6 on p. 124 yields an observed F-ratio of F o = 2. 6141. Assuming independence, normality, and homogeneity of variance, F is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = g - 1 = 3 - 1 = 2\) and \(\nu _{2} = N - g = 26 - 3 = 23\) degrees of freedom. Under the null hypothesis, the observed value of F o = 2. 6141 yields an approximate probability value of P = 0. 0948, which is similar to that produced by the MRPP resampling analysis of the OLS regression residuals.
4.3.2 One-Way Randomized Design with a Covariate
A covariate experimental design permits the testing of differences among the treatment groups after the effect of the covariate has been removed from the analysis. Consider a one-way completely randomized design with a covariate in which N = 47 objects are randomly assigned to one of g = 5 treatment groups. The experimental data are listed in Table 4.1 and are adapted from a 1984 study by Conti and Musty [78].
A design matrix of dummy codes for analyzing treatments is given in Fig. 4.10, where the first column of 1 values provides for an intercept, the second column contains the covariate (Pre-test) values, and the third column contains the (Post-test) scores listed according to the original random assignment of the N = 47 objects to the g = 5 treatment groups with the first n 1 = 10 scores, the next n 2 = 10 scores, the next n 3 = 9 scores, the next n 4 = 8 scores, and the last n 5 = 10 scores associated with the g = 5 treatment groups, respectively.
The MRPP regression analysis examines the N = 47 regression residuals for possible differences among the g = 5 treatment levels; consequently, no dummy codes for treatments are included in Fig. 4.10 as this information is implicit in the ordering of the g = 5 treatment groups in the two paired columns labeled “Pre” and “Post.”
Because there are
possible, equally-likely arrangements of the N = 47 univariate response measurement scores listed in Table 4.1, an exact permutation approach is not possible and a resampling analysis is mandated.
LAD Regression Analysis
An MRPP resampling analysis of the LAD regression residuals calculated on the N = 47 response measurement scores listed in Fig. 4.10 yields estimated LAD regression coefficients of
Table 4.2 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 47.
Following Eq. (4.5) on p. 125 and employing ordinary Euclidean distance between residuals v = 1, the LAD regression residuals listed in Table 4.2 yield g = 5 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Table 4.2 with v = 1 and treatment-group weights
is
If all M possible arrangements of the observed LAD regression residuals listed in Table 4.2 occur with equal chance, the approximate resampling probability value of δ o = 0. 7962 computed on L = 1, 000, 000 random arrangements of the observed LAD regression residuals with \(n_{1} = n_{2} = n_{5} = 10\), n 3 = 9, and n 4 = 8 residuals preserved for each arrangement is
Following Eq. (4.7) on p. 126, the exact expected value of the Mδ values is μ δ = 0. 9178 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is
indicating approximately 13 % agreement between the observed and predicted y values above that expected by chance.
OLS Regression Analysis
For comparison, consider an MRPP resampling analysis of the OLS regression residuals calculated on the N = 47 univariate response measurement scores listed in Fig. 4.10 on p. 132. The MRPP regression analysis yields estimated OLS regression coefficients of
Table 4.3 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 47.
Following Eq. (4.5) on p. 125 and employing squared Euclidean distance between residuals with v = 2, the OLS regression residuals listed in Table 4.3 yield g = 5 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the OLS regression residuals listed in Table 4.3 with v = 2 and treatment-group weights
is
If all M possible arrangements of the N = 47 observed OLS regression residuals listed in Table 4.3 occur with equal chance, the approximate resampling probability value of δ o = 1. 0482 computed on L = 1, 000, 000 random arrangements of the observed OLS regression residuals with \(n_{1} = n_{2} = n_{5} = 10\), n 3 = 9, and n 4 = 8 residuals preserved for each arrangement is
For comparison, the approximate resampling probability value based on LAD regression, v = 1, L = 1, 000, 000, and \(C_{i} = n_{i}/N\) for i = 1, …, 5 is P = 0. 0041.
Following Eq. (4.7) on p. 126, the exact expected value of the Mδ values is μ δ = 1. 2761 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating approximately 18 % agreement between the observed and predicted y values above that expected by chance.
Conventional ANOVA Analysis
A conventional fixed-effects one-way analysis of covariance calculated on the N = 47 univariate response measurement scores listed in Table 4.1 on p. 131 yields an observed F-ratio of F o = 4. 6978. Assuming independence, normality, and homogeneity of variance, F is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = g - 1 = 5 - 1 = 4\) and \(\nu _{2} = N - g - 1 = 47 - 5 - 1 = 41\) degrees of freedom. Under the null hypothesis, the observed value of F o = 4. 6978 yields an approximate probability value of P = 0. 0033.
4.3.3 One-Way Randomized-Block Design
One-way randomized-block designs are common in experimental research and have long been valuable statistical tools in such fields as agriculture and genetics. E.J.G. Pitman , for example, developed a permutation approach for one-way randomized-block designs in 1938 [342]. With modern developments in embryo transplants and cloning where subjects can be genetically matched on a large number of important characteristics, randomized-block designs have become very practical and efficient.Footnote 2
Consider a one-way randomized-block design where b = 6 objects (blocks) are evaluated over a = 3 treatments with r = 1 response measurement. The design and data are adapted from a study by Anderson , Sweeney , and Williams [9, p. 471] and are given in Fig. 4.11.
A design matrix of dummy codes for an MRPP regression analysis is given in Fig. 4.12, where the first column of 1 values provides for an intercept, the next five columns contain dummy codes for the b = 6 blocks, and the last column contains the univariate response measurement scores listed according to the original random assignment of the N = 18 objects to the a = 3 treatment levels of Factor A with the first \(n_{A_{1}} = 6\) objects, the next \(n_{A_{2}} = 6\) objects, and the last \(n_{A_{3}} = 6\) objects associated with treatment levels \(A_{1}\), A 2, and A 3, respectively. The MRPP regression analysis examines the N = 18 regression residuals for possible differences in the a = 3 treatment levels; consequently, there are no dummy codes for treatments in Fig. 4.12 as this information is implicit in the ordering of the a = 3 treatment levels of Factor A in the last column.
Because there are
possible, equally-likely arrangements of the N = 18 univariate response measurement scores listed in Fig. 4.11, an exact permutation approach is not practical.
LAD Regression Analysis
An MRPP resampling analysis of the LAD regression residuals calculated on the univariate response measurement scores listed in Fig. 4.12 yields estimated LAD regression coefficients of
for Factor A. Figure 4.13 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 18.
Following Eq. (4.5) on p. 125 and employing ordinary Euclidean distance between residuals with v = 1, the N = 18 LAD regression residuals listed in Fig. 4.13 yield a = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.13 with v = 1 and treatment-group weights
is
If all M possible arrangements of the N = 18 observed LAD regression residuals listed in Fig. 4.13 occur with equal chance, the approximate resampling probability value of δ A = 1. 2889 computed on L = 1, 000, 000 random arrangements of the observed LAD regression residuals with \(n_{A_{1}} = n_{A_{2}} = n_{A_{3}} = 6\) residuals preserved for each arrangement is
Following Eq. (4.7) on p. 126, the exact expected value of the M = 17, 153, 136 δ values is μ δ = 1. 6078 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is
indicating approximately 20 % agreement between the observed and predicted y values above that expected by chance.
An Exact Test
Although an exact permutation analysis of the N = 18 LAD regression residuals listed in Fig. 4.13 is impractical, it is not impossible. In fact, exact permutation methods are oftentimes more efficient than resampling permutation methods because the L = 1, 000, 000 calls to a pseudorandom number generator, necessary for a resampling test, are not required by an exact test.
Following Eq. (4.5) on p. 125, an exact permutation analysis of the N = 18 LAD regression residuals listed in Fig. 4.13 yields a = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic based on v = 1 and treatment-group weights
is
If all arrangements of the N = 18 observed LAD regression residuals listed in Fig. 4.13 occur with equal chance, the exact probability value of δ A = 1. 2889 computed on the M = 17, 153, 136 possible arrangements of the observed LAD regression residuals with \(n_{A_{1}} = n_{A_{2}} = n_{A_{3}} = 6\) residuals preserved for each arrangement is
For comparison, the resampling probability value computed on L = 1, 000, 000 random arrangements of the observed LAD regression residuals listed in Fig. 4.13 is P = 0. 0560.
OLS Regression Analysis
For comparison, consider an MRPP resampling analysis of OLS regression residuals calculated on the N = 18 univariate response measurement scores listed in Fig. 4.12 on p. 137. The MRPP regression analysis yields estimated OLS regression coefficients of
for Factor A. Figure 4.14 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 18.
Following Eq. (4.5) on p. 125 and employing ordinary Euclidean distance between residuals with v = 2, the N = 18 OLS regression residuals listed in Fig. 4.14 yield a = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the OLS regression residuals listed in Fig. 4.14 with v = 2 and treatment-group weights
is
If all M possible arrangements of the N = 18 observed OLS regression residuals listed in Fig. 4.14 occur with equal chance, the approximate resampling probability value of δ A = 2. 5333 computed on L = 1, 000, 000 random arrangements of the observed OLS regression residuals with \(n_{A_{1}} = n_{A_{2}} = n_{A_{3}} = 6\) residuals preserved for each arrangement is
For comparison, the approximate resampling probability value based on LAD regression, v = 1, L = 1, 000, 000, and \(C_{i} = n_{A_{i}}/N\) for i = 1, 2, 3 is P = 0. 0560.
Following Eq. (4.7) on p. 126, the exact expected value of the M = 17, 153, 136 δ values is μ δ = 5. 5556 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating approximately 54 % agreement between the observed and predicted y values above that expected by chance.
An Exact Test
Although an exact permutation analysis of the N = 18 OLS regression residuals listed in Fig. 4.14 is impractical, it is not impossible. Following Eq. (4.5) on p. 125, an exact permutation analysis of the N = 18 OLS regression residuals listed in Fig. 4.14 yields a = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic based on v = 2 and treatment-group weights
is
If all arrangements of the N = 18 observed OLS regression residuals listed in Fig. 4.14 occur with equal chance, the exact probability value of δ A = 2. 5333 computed on the M = 17, 153, 136 possible arrangements of the observed OLS regression residuals with \(n_{A_{1}} = n_{A_{2}} = n_{A_{3}} = 6\) residuals preserved for each arrangement is
For comparison, the approximate resampling probability value computed on L = 1, 000, 000 random arrangements of the observed OLS regression residuals listed in Fig. 4.14 is also P = 0. 0050.
Conventional ANOVA Analysis
A conventional randomized-block analysis of variance calculated on the N = 18 univariate response measurement scores listed in Fig. 4.11 on p. 137 yields an observed F-ratio of F A = 5. 5263. Assuming independence and normality, F A is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = a - 1 = 3 - 1 = 2\) and \(\nu _{2} = (b - 1)(a - 1) = (6 - 1)(3 - 1) = 10\) degrees of freedom. Under the null hypothesis, the observed value of F A = 5. 5263 yields an approximate probability value of P = 0. 0242.
4.3.4 Two-Way Randomized-Block Design
Consider a balanced two-way randomized-block design in which n = 3 subjects (S ) are tested over a = 3 levels of Factor A and the experiment is repeated b = 3 times for Factor B. The design and data are adapted from Myers and Well [315, p. 260] and are given in Table 4.4. A complete permutation analysis of a two-way randomized-block design requires three separate analyses comprised of (1) the main effect of Factor A, (2) the main effect of Factor B, and (3) the A×B interaction effect.
Analysis of Factor A
A design matrix of dummy codes for analyzing Factor A is given on the left side of Table 4.5, where the first column of 1 values provides for an intercept and the second and third columns contain dummy codes for Factor B. The last column on the left side of Table 4.5 lists the N = 9 response measurement summations over the b = 3 levels of Factor B (e.g., \(3.10 + 1.90 + 1.60 = 6.60\)) and ordered by the a = 3 treatment levels of Factor A with the first \(n_{A_{1}} = 3\) summations, the next \(n_{A_{2}} = 3\) summations, and the last \(n_{A_{3}} = 3\) summations associated with treatment levels A 1, A 2, and A 3, respectively. The MRPP regression analysis examines the N = 9 regression residuals for possible differences in the a = 3 treatment levels of Factor A; consequently, no dummy codes are provided for Factor A as this information is implicit in the ordering of the a = 3 treatment levels of Factor A in the last column on the left side of Table 4.5.
An exact permutation solution is reasonable for the response measurement summations listed on the left side of Table 4.5 since there are only
possible, equally-likely arrangements of the N = 9 response measurement summations for Factor A with \(n_{A_{1}} = n_{A_{2}} = n_{A_{3}} = 3\) response measurement summations preserved for each arrangement of the observed data.
LAD Regression Analysis
An MRPP analysis of the LAD regression residuals calculated on the N = 9 response measurement summations on the left side of Table 4.5 yields estimated LAD regression coefficients of
for Factor A. Figure 4.15 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 9.
Following Eq. (4.5) on p. 125 and employing squared Euclidean distance between residuals with v = 1, the N = 9 LAD regression residuals listed in Fig. 4.15 yield a = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.15 with v = 1 and treatment-group weights
is
If all arrangements of the N = 9 observed LAD regression residuals listed in Fig. 4.15 occur with equal chance, the exact probability value of δ A = 1. 8889 computed on the M = 1, 680 possible arrangements of the observed LAD regression residuals with \(n_{A_{1}} = n_{A_{2}} = n_{A_{3}} = 3\) residuals preserved for each arrangement is
Following Eq. (4.7) on p. 126, the exact expected value of the M = 1, 680 δ values is μ δ = 2. 9889 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is
indicating approximately 37 % agreement between the observed and predicted y values above that expected by chance.
OLS Regression Analysis
For comparison, consider an MRPP analysis of OLS regression residuals calculated on the N = 9 response measurement summations for Factor A listed on the left side of Table 4.5. Again, since there are only M = 1, 680 possible arrangements of the response measurement summations, an exact permutation test is selected. The MRPP regression analysis yields estimated OLS regression coefficients of
for Factor A. Figure 4.16 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 9.
Following Eq. (4.5) on p. 125 and employing squared Euclidean distance between residuals with v = 2, the N = 9 OLS regression residuals listed in Fig. 4.16 yield a = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the OLS regression residuals listed in Fig. 4.16 with v = 2 and treatment-group weights
is
If all arrangements of the N = 9 observed OLS regression residuals listed in Fig. 4.16 occur with equal chance, the exact probability value of δ A = 6. 6237 computed on the M = 1, 680 possible arrangements of the observed OLS regression residuals with \(n_{A_{1}} = n_{A_{2}} = n_{A_{3}} = 3\) residuals preserved for each arrangement is
For comparison, the exact probability value based on LAD regression, v = 1, M = 1, 680, and \(C_{i} = n_{A_{i}}/N\) for i = 1, 2, 3 is P = 0. 0036.
Following Eq. (4.7) on p. 126, the exact expected value of the M = 1, 680 δ values is μ δ = 14. 7250 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating approximately 55 % agreement between the observed and predicted y values above that expected by chance.
Conventional ANOVA Analysis
A conventional randomized-block analysis of variance calculated on the N = 27 univariate response measurement scores for Factor A in Table 4.4 on p. 143 yields an observed F-ratio of F A = 3. 9282. Assuming independence and normality, F A is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = a - 1 = 3 - 1 = 2\) and \(\nu _{2} = (n - 1)(a - 1) = (3 - 1)(3 - 1) = 4\) degrees of freedom. Under the null hypothesis, the observed value of F A = 3. 9282 yields an approximate probability value of P = 0. 1138.
Analysis of Factor B
The right side of Table 4.5 on p. 144 contains a design matrix of dummy codes for analyzing Factor B, where the first column of 1 values provides for an intercept and the next two columns contain dummy codes for Factor A. The last column on the right side of Table 4.5 lists the N = 9 response measurement summations over the a = 3 levels of Factor A (e.g., \(3.10 + 2.90 + 2.40 = 8.40\)) and ordered by the b = 3 treatment levels with the first \(n_{B_{1}} = 3\) summations, the next \(n_{B_{2}} = 3\) summations, and the last \(n_{B_{3}} = 3\) summations associated with treatment levels, B 1, B 2, and B 3, respectively. The MRPP regression analysis examines the N = 9 regression residuals for possible differences among the b = 3 treatment levels of Factor B; consequently, no dummy codes are provided for Factor B as this information is implicit in the ordering of the b = 3 treatment levels of Factor B in the last column on the right side of Table 4.5.
An exact permutation solution is ideal for the response measurement summations on the right side of Table 4.5 since there are only
possible, equally-likely arrangements of the N = 9 response measurement summations for Factor B with \(n_{B_{1}} = n_{B_{2}} = n_{B_{3}}\) response measurement summations preserved for each arrangement of the observed data.
LAD Regression Analysis
An MRPP analysis of the LAD regression residuals calculated on the N = 9 response measurement summations on the right side of Table 4.5 on p. 144 yields estimated LAD regression coefficients of
for Factor B. Figure 4.17 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 9.
Following Eq. (4.5) on p. 125 and employing ordinary Euclidean distance between residuals with v = 1, the N = 9 LAD regression residuals listed in Fig. 4.17 yield b = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.17 with v = 1 and treatment-group weights
is
If all arrangements of the N = 9 observed LAD regression residuals listed in Fig. 4.17 occur with equal chance, the exact probability value of δ B = 0. 7556 computed on the M = 1, 680 possible arrangements of the observed LAD regression residuals with \(n_{B_{1}} = n_{B_{2}} = n_{B_{3}} = 3\) residuals preserved for each arrangement is
Following Eq. (4.7) on p. 126, the exact expected value of the M = 1, 680 δ values is μ δ = 2. 5889 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is
indicating approximately 71 % agreement between the observed and predicted y values above that expected by chance.
OLS Regression Analysis
For comparison, consider an MRPP analysis of OLS regression residuals calculated on the N = 9 response measurement summations for Factor B listed on the right side of Table 4.5 on p. 144. Again, since there are only M = 1, 680 possible arrangements of the response measurement summations, an exact permutation test is preferred. The MRPP regression analysis yields estimated OLS regression coefficients of
for Factor B. Figure 4.18 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 9.
Following Eq. (4.5) on p. 125 and employing squared Euclidean distance between residuals with v = 2, the N = 9 OLS regression residuals listed in Fig. 4.18 yield b = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the OLS regression residuals listed in Fig. 4.18 with v = 2 and treatment-group weights
is
If all arrangements of the N = 9 observed OLS regression residuals listed in Fig. 4.18 occur with equal chance, the exact probability value of δ B = 1. 0770 computed on the M = 1, 680 possible arrangements of the observed OLS regression residuals with \(n_{B_{1}} = n_{B_{2}} = n_{B_{3}} = 3\) residuals preserved for each arrangement is
For comparison, the exact probability value based on LAD regression, v = 1, M = 1, 680, and \(C_{i} = n_{B_{i}}/N\) for i = 1, 2, 3 is also P = 0. 0036.
Following Eq. (4.7) on p. 126, the exact expected value of the M = 1, 680 δ values is μ δ = 9. 9150 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating approximately 89 % agreement between the observed and predicted y values above that expected by chance.
Conventional ANOVA Analysis
A conventional randomized-block analysis of variance calculated on the N = 27 univariate response measurement scores for Factor B listed in Table 4.4 on p. 143 yields an observed F-ratio of F B = 22. 5488. Assuming independence and normality, F B is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = b - 1 = 3 - 1 = 2\) and \(\nu _{2} = (n - 1)(b - 1) = (3 - 1)(3 - 1) = 4\) degrees of freedom. Under the null hypothesis, the observed value of F B = 22. 5488 yields an approximate probability value of P = 0. 0066, which is similar to the LAD and OLS regression probability value of P = 0. 0036.
Analysis of the A×B Interaction
A design matrix of dummy codes for analyzing the interaction of Factors A and B is given in Table 4.6, where the first column of 1 values provides for an intercept and the second and third columns contain dummy codes for subjects (S ). The fourth and fifth columns contain dummy codes for Factor A, the sixth and seventh columns contain dummy codes for Factor B, and the next eight columns contain dummy codes for the S×A and S×B interactions. The last column in Table 4.6 lists the response measurement scores ordered by the \(ab = (3)(3) = 9\) levels of the A×B interaction.
The MRPP regression analysis examines the N = 27 regression residuals for possible differences among the nine treatment levels of the A×B interaction; consequently, no dummy codes are provided for the A×B interaction as this information is implicit in the ordering of the treatment levels of the A×B interaction in the last column of Table 4.6.
Because there are
possible, equally-likely arrangements of the N = 27 univariate response measurement scores for the A×B interaction listed in Table 4.6, an exact permutation solution is not possible.
LAD Regression Analysis
An MRPP resampling analysis of the LAD regression residuals calculated on the N = 27 univariate response measurement scores listed in Table 4.6 yields estimated LAD regression coefficients of
for the interaction of Factors A and B. Figure 4.19 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 27.
Following Eq. (4.5) on p. 125 and employing ordinary Euclidean distance between residuals with v = 1, the N = 27 LAD regression residuals listed in Fig. 4.19 yield \(ab = (3)(3) = 9\) average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.19 with v = 1 and treatment-group weights
is
If all M possible arrangements of the N = 27 observed LAD regression residuals listed in Fig. 4.19 occur with equal chance, the approximate resampling probability value of δ A×B = 0. 1926 computed on L = 1, 000, 000 random arrangements of the observed LAD regression residuals with \(n_{(A\times B)_{1}} = \cdots = n_{(A\times B)_{9}} = 3\) residuals preserved for each arrangement is
Following Eq. (4.7) on p. 126, the exact expected value of the Mδ values is μ δ = 0. 2063 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is
indicating approximately 7 % agreement between the observed and predicted y values above that expected by chance.
OLS Regression Analysis
For comparison, consider an MRPP analysis of OLS regression residuals calculated on the N = 27 univariate response measurement scores for the A×B interaction listed in Table 4.6. The MRPP regression analysis yields estimated OLS regression coefficients of
for the interaction of Factors A and B. Figure 4.20 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 27.
Following Eq. (4.5) on p. 125 and employing squared Euclidean distance between residuals with v = 2, the N = 27 OLS regression residuals listed in Fig. 4.20 yield \(ab = (3)(3) = 9\) average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the OLS regression residuals listed in Fig. 4.20 with v = 2 and treatment-group weights
is
If all M possible arrangements of the N = 27 observed OLS regression residuals listed in Fig. 4.20 occur with equal chance, the approximate resampling probability value of δ A×B = 0. 0724 computed on L = 1, 000, 000 random arrangements of the observed OLS regression residuals with \(n_{(A\times B)_{1}} = \cdots = n_{(A\times B)_{9}} = 3\) residuals preserved for each arrangement is
For comparison, the approximate resampling probability value based on LAD regression, v = 1, L = 1, 000, 000, and \(C_{i} = n_{(A\times B)_{i}}/N\) for i = 1, …, 9 is P = 0. 2355.
Following Eq. (4.7) on p. 126, the exact expected value of the Mδ values is μ δ = 8. 9231 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating approximately 19 % agreement between the observed and predicted y values above that expected by chance.
Conventional ANOVA Analysis
A conventional randomized-block analysis of variance calculated on the N = 27 response measurement scores for the A×B interaction listed in Table 4.4 on p. 143 yields an observed F-ratio of F A×B = 1. 5591. Assuming independence and normality, F A×B is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = (a - 1)(b - 1) = (3 - 1)(3 - 1) = 4\) and \(\nu _{2} = (n - 1)(a - 1)(b - 1) = (3 - 1)(3 - 1)(3 - 1) = 8\) degrees of freedom. Under the null hypothesis, the observed value of F A×B = 1. 5591 yields an approximate probability value of P = 0. 2744.
4.3.5 Two-Way Factorial Design
Consider a 2×3 fixed-effects factorial design with n = 4 subjects in each treatment combination for a total of N = 24 subjects. The univariate response measurement scores for Factors A and B are listed in Fig. 4.21, and the design matrices and data for Factors A and B are given in Table 4.7; the design and data are adapted from Keppel [214, p. 197]. While design matrices of either dummy or effect codes are appropriate for one-way completely randomized and randomized-block designs, the main effects of factorial designs are best analyzed with effect codes when estimation of the effects of each factor is adjusted for all other factors in the model to obtain the unique contribution of each factor [31, 37, 294].Footnote 3 A permutation analysis of factorial designs requires three separate analyses comprising (1) the main effect of Factor A, (2) the main effect of Factor B, and (3) the A×B interaction effect.
Analysis of Factor A
A design matrix of effect codes for analyzing Factor A is given on the left side of Table 4.7, where the first column of 1 values provides for an intercept. The second and third columns contain effect codes for Factor B, the fourth and fifth columns contain effect codes for the A×B interaction, and the last column on the left side of Table 4.7 contains the N = 24 univariate response measurement scores listed according to the original random assignment of the subjects to the a = 2 levels of Factor A with the first \(n_{A_{1}} = 12\) scores and the last \(n_{A_{2}} = 12\) scores associated with treatment levels A 1 and A 2, respectively. The MRPP regression analysis examines the N = 24 regression residuals for possible differences between the a = 2 treatment levels of Factor A; consequently, no effect codes are provided for Factor A as this information is implicit in the ordering of the a = 2 treatment levels of Factor A in the last column on the left side of Table 4.7.
An exact permutation solution is feasible for the univariate response measurement scores listed on the left side of Table 4.7 since there are only
possible, equally-likely arrangements of the N = 24 response measurement scores for Factor A.
LAD Regression Analysis
An MRPP analysis of the LAD regression residuals calculated on the N = 24 univariate response measurement scores on the left side of Table 4.7 yields estimated LAD regression coefficients of
for Factor A. Figure 4.22 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 24.
Following Eq. (4.5) on p. 125 and employing ordinary Euclidean distance between residuals with v = 1, the N = 24 LAD regression residuals listed in Fig. 4.22 yield a = 2 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.22 with v = 1 and treatment-group weights
is
If all arrangements of the N = 24 observed LAD regression residuals listed in Fig. 4.22 occur with equal chance, the exact probability value of δ A = 5. 0758 computed on the M = 2, 704, 156 possible arrangements of the observed LAD regression residuals with \(n_{A_{1}} = n_{A_{2}} = 12\) residuals preserved for each arrangement is
Following Eq. (4.7) on p. 126, the exact expected value of the M = 2, 704, 156 δ values is μ δ = 5. 0725 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is
indicating slightly less than chance agreement between the observed and predicted y values.
OLS Regression Analysis
For comparison, consider an MRPP analysis of OLS regression residuals calculated on the N = 24 univariate response measurement scores for Factor A on the left side of Table 4.7. The MRPP regression analysis yields estimated OLS regression coefficients of
for Factor A. Figure 4.23 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 24.
Following Eq. (4.5) on p. 125 and employing squared Euclidean distance between residuals with v = 2, the N = 24 OLS regression residuals listed in Fig. 4.23 yield a = 2 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the OLS regression residuals listed in Fig. 4.23 with v = 2 and treatment-group weights
is
If all arrangements of the N = 24 observed OLS regression residuals listed in Fig. 4.23 occur with equal chance, the exact probability value of δ A = 30. 00 computed on the M = 2, 704, 156 possible arrangements of the observed OLS regression residuals with \(n_{A_{1}} = n_{A_{2}} = 12\) residuals preserved for each arrangement is
For comparison, the exact probability value based on LAD regression, v = 1, M = 2, 704, 156, and \(C_{i} = n_{A_{i}}/N\) for i = 1, 2 is P = 0. 3843.
Following Eq. (4.7) on p. 126, the exact expected value of the M = 2, 704, 156 δ values is μ δ = 30. 7826 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating approximately 3 % agreement between the observed and predicted y values above that expected by chance.
Conventional ANOVA Analysis
A conventional fixed-effects factorial analysis of variance calculated on the N = 24 Factor A response measurement scores listed in Fig. 4.21 on p. 156 yields an observed F-ratio of F A = 1. 3091. Assuming independence, normality, and homogeneity of variance, F A is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = a - 1 = 2 - 1 = 1\) and \(\nu _{2} = N - ab = 24 - (2)(3) = 18\) degrees of freedom. Under the null hypothesis, the observed value of F A = 1. 3091 yields an approximate probability value of P = 0. 2675, which is similar to the OLS regression probability value of P = 0. 2357.
Analysis of Factor B
The right side of Table 4.7 on p. 157 contains a design matrix of effect codes for analyzing Factor B, where the first column of 1 values provides for an intercept. The second column contains effect codes for Factor A, the third and fourth columns contain effect codes for the A×B interaction, and the last column on the right side of Table 4.7 contains the N = 24 univariate response measurement scores listed according to the original random assignment of the subjects to the b = 3 levels of Factor B with the first \(n_{B_{1}} = 8\) scores, the next \(n_{B_{2}} = 8\) scores, and the last \(n_{B_{3}} = 8\) scores associated with treatment levels, B 1, B 2, and B 3, respectively. The MRPP regression analysis examines the N = 24 regression residuals for possible differences among the b = 3 treatment levels of Factor B; consequently, no effect codes are provided for Factor B as this information is implicit in the ordering of the b = 3 treatment levels of Factor B in the last column on the right side of Table 4.7.
Because there are
possible, equally-likely arrangements of the N = 24 response measurement scores for Factor B listed on the right side of Table 4.7, an exact permutation approach is not practical.
LAD Regression Analysis
An MRPP resampling analysis of the N = 24 LAD regression residuals calculated on the univariate response measurement scores on the right side of Table 4.7 on p. 157 yields estimated LAD regression coefficients of
for Factor B. Figure 4.24 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 24.
Following Eq. (4.5) on p. 125 and employing ordinary Euclidean distance between residuals with v = 1, the N = 24 LAD regression residuals listed in Fig. 4.24 yield b = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.24 with v = 1 and treatment-group weights
is
If all M possible arrangements of the N = 24 observed LAD regression residuals listed in Fig. 4.24 occur with equal chance, the approximate resampling probability value of δ B = 5. 00 computed on L = 1, 000, 000 random arrangements of the observed LAD regression residuals with \(n_{B_{1}} = n_{B_{2}} = n_{B_{3}} = 8\) residuals preserved for each arrangement is
Following Eq. (4.7) on p. 126, the exact expected value of the Mδ values is μ δ = 5. 3333 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is
indicating approximately 6 % agreement between the observed and predicted y values above that expected by chance.
OLS Regression Analysis
For comparison, consider an MRPP analysis of OLS regression residuals calculated on the N = 24 univariate response measurement scores for Factor B listed on the right side of Table 4.7 on p. 157. The MRPP regression analysis yields estimated OLS regression coefficients of
for Factor B. Figure 4.25 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 24.
Following Eq. (4.5) on p. 125 and employing squared Euclidean distance between residuals with v = 2, the N = 24 OLS regression residuals listed in Fig. 4.25 yield b = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the OLS regression residuals listed in Fig. 4.25 with v = 2 and treatment-group weights
is
If all M possible arrangements of the N = 24 observed OLS regression residuals listed in Fig. 4.25 occur with equal chance, the approximate resampling probability value of δ B = 31. 4286 computed on L = 1, 000, 000 random arrangements of the observed OLS regression residuals with \(n_{B_{1}} = n_{B_{2}} = n_{B_{3}} = 8\) residuals preserved for each arrangement is
For comparison, the approximate resampling probability value based on LAD regression, v = 1, L = 1, 000, 000, and \(C_{i} = n_{B_{i}}/N\) for i = 1, 2, 3 is P = 0. 1250.
Following Eq. (4.7) on p. 126, the exact expected value of the M = 9, 465, 511, 770 δ values is μ δ = 38. 4348 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating approximately 18 % agreement between the observed and predicted y values above that expected by chance.
Conventional ANOVA Analysis
A conventional fixed-effects factorial analysis of variance calculated on the N = 24 Factor B response measurement scores listed in Fig. 4.21 on p. 156 yields an observed F-ratio of F B = 3. 0545. Assuming independence, normality, and homogeneity of variance, F B is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = b - 1 = 3 - 1 = 2\) and \(\nu _{2} = N - ab = 24 - (2)(3) = 18\) degrees of freedom. Under the null hypothesis, the observed value of F B = 3. 0545 yields an approximate probability value of P = 0. 0721.
Analysis of the A×B Interaction
A design matrix of effect codes for analyzing the A×B interaction of the data listed in Fig. 4.21 on p. 156 is given in Fig. 4.26, where the first column of 1 values provides for an intercept, the second column contains effect codes for Factor A, the third and fourth columns contain effect codes for Factor B, and the last column lists the N = 24 univariate response measurement scores listed according to the original random assignment of the subjects to the \(ab = (2)(3) = 6\) levels of the A×B interaction. The MRPP regression analysis examines the N = 24 regression residuals for possible differences among the six treatment levels of the A×B interaction; consequently, no effect codes are provided for the A×B interaction as this information is implicit in the ordering of the treatment levels of the A×B interaction in the last column of Fig. 4.26.
Because there are
possible, equally-likely arrangements of the N = 24 univariate response measurement scores for the A×B interaction listed in Fig. 4.26, an exact permutation approach is clearly not possible.
LAD Regression Analysis
An MRPP resampling analysis of the LAD regression residuals calculated on the univariate response measurement scores in Fig. 4.26 yields estimated LAD regression coefficients of
for the interaction of Factors A and B. Figure 4.27 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 24.
Following Eq. (4.5) on p. 125 and employing ordinary Euclidean distance between residuals with v = 1, the N = 24 LAD regression residuals listed in Fig. 4.27 yield \(ab = (2)(3) = 6\) average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.27 with v = 1 and treatment-group weights
is
If all M possible arrangements of the N = 24 observed LAD regression residuals listed in Fig. 4.27 occur with equal chance, the approximate resampling probability value of δ A×B = 5. 3333 computed on L = 1, 000, 000 random arrangements of the observed LAD regression residuals with \(n_{(A\times B)_{1}} = \cdots = n_{(A\times B)_{6}} = 4\) residuals preserved for each arrangement is
Following Eq. (4.7) on p. 126, the exact expected value of the Mδ values is μ δ = 5. 50 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is
indicating approximately 3 % agreement between the observed and predicted y values above that expected by chance.
OLS Regression Analysis
For comparison, consider an MRPP analysis of OLS regression residuals calculated on the N = 24 univariate response measurement scores of the A×B interaction listed in Fig. 4.26. The MRPP regression analysis yields estimated OLS regression coefficients of
for the interaction of Factors A and B. Figure 4.28 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 24.
Following Eq. (4.5) on p. 125 and employing squared Euclidean distance between residuals with v = 2, the N = 24 OLS regression residuals listed in Fig. 4.28 yield \(ab = (2)(3) = 6\) average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the OLS regression residuals listed in Fig. 4.28 with v = 2 and treatment-group weights
is
If all M possible arrangements of the observed OLS regression residuals listed in Fig. 4.28 occur with equal chance, the approximate resampling probability value of δ A×B = 36. 6666 computed on L = 1, 000, 000 random arrangements of the observed OLS regression residuals with \(n_{(A\times B)_{1}} = \cdots = n_{(A\times B)_{6}} = 4\) residuals preserved for each arrangement is
For comparison, the approximate resampling probability value based on LAD regression, v = 1, L = 1, 000, 000, and \(C_{i} = n_{(A\times B)_{i}}/N\) for i = 1, …, 6 is P = 0. 3477.
Following Eq. (4.7) on p. 126, the exact expected value of the Mδ values is μ δ = 41. 2174 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating approximately 11 % agreement between the observed and predicted y values above that expected by chance.
Conventional ANOVA Analysis
A conventional fixed-effects factorial analysis of variance calculated on the N = 24 univariate response measurement scores listed in Fig. 4.21 on p. 156 yields an observed F-ratio of F A×B = 3. 9273. Assuming independence, normality, and homogeneity of variance, F A×B is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = (a - 1)(b - 1) = (2 - 1)(3 - 1) = 2\) and \(\nu _{2} = ab(n - 1) = (2)(3)(4 - 1) = 18\) degrees of freedom. Under the null hypothesis, the observed value of F A×B = 3. 9273 yields an approximate probability value of P = 0. 0384, which differs greatly from the LAD and OLS regression probability values of P = 0. 3477 and P = 0. 2242, respectively.
4.3.6 Latin Square Design
A Latin square experimental design assigns treatments to subjects so the treatments occur in a balanced fashion within a square block or field; thus, n treatments appear once in each of n rows and n columns. The Latin square is the design of choice when controlling for two blocking factors. Consider an ordinary balanced Latin square experiment involving repeated measurements in which n = 4 subjects (S ) are each tested b = 4 times on Factor A. The design and data are adapted from Ferguson [115, p. 349] and are given in Table 4.8, where B refers to the ordinal position in which the levels of Factor A are administered. Thus, the first subject, S 1, receives the b = 4 treatments in the order A 2, A 4, A 1, A 3, and so on. Due to the balanced nature of Latin square designs, the assumption is that there is no interaction between blocking Factors A and B, or between either blocking factor and the treatments.
Analysis of Factor A
A design matrix of dummy codes for analyzing Factor A is given in Fig. 4.29, where the first column of 1 values provides for an intercept, the second through fourth columns contain dummy codes for Subjects, the fifth through seventh columns contain dummy codes for Factor B, and the last column lists the univariate response measurement scores ordered by the a = 4 levels of Factor A, with the first \(n_{A_{1}} = 4\) scores, the next \(n_{A_{2}} = 4\) scores, the next \(n_{A_{3}} = 4\) scores, and the last \(n_{A_{4}} = 4\) scores associated with treatment levels A 1, A 2, A 3, and A 4, respectively. The MRPP regression analysis examines the N = 16 regression residuals for possible differences among the a = 4 treatment levels of Factor A; consequently, no dummy codes are provided for Factor A as this information is implicit in the ordering of the a = 4 treatment levels of Factor A in the last column of Fig. 4.29.
Because there are
possible, equally-likely arrangements of the N = 16 univariate response measurement scores listed in Fig. 4.29, an exact permutation approach is not practical.
LAD Regression Analysis
An MRPP resampling analysis of the N = 16 LAD regression residuals calculated on the univariate response measurement scores listed in Fig. 4.29 yields estimated LAD regression coefficients of
for Factor A. Figure 4.30 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 16.
Following (4.5) on p. 125 and employing ordinary Euclidean distance between residuals with v = 1, the N = 16 LAD regression residuals listed in Fig. 4.30 yield a = 4 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.30 with v = 1 and treatment-group weights
is
If all M possible arrangements of the N = 16 observed LAD regression residuals listed in Fig. 4.30 occur with equal chance, the approximate resampling probability value of δ A = 6. 2083 computed on L = 1, 000, 000 random arrangements of the observed LAD regression residuals with \(n_{A_{1}} = \cdots = n_{A_{4}} = 4\) residuals preserved for each arrangement is
Following Eq. (4.7) on p. 126, the exact expected value of the M = 63, 063, 000 δ values is μ δ = 8. 2750 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is
indicating approximately 25 % agreement between the observed and predicted y values above that expected by chance.
OLS Regression Analysis
For comparison, consider an MRPP resampling analysis of the OLS regression residuals calculated on the N = 16 univariate response measurement scores listed in Fig. 4.29 on p. 171. The MRPP regression analysis yields estimated OLS regression coefficients of
for Factor A. Figure 4.31 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 16.
Following Eq. (4.5) on p. 125 and employing squared Euclidean distance between residuals with v = 2, the N = 16 OLS regression residuals listed in Fig. 4.31 yield a = 4 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the OLS regression residuals listed in Fig. 4.31 with v = 2 and treatment-group weights
is
If all M possible arrangements of the N = 16 observed OLS regression residuals listed in Fig. 4.31 occur with equal chance, the approximate resampling probability value of δ A = 3. 9792 computed on L = 1, 000, 000 random arrangements of the observed OLS regression residuals with \(n_{A_{1}} = \cdots = n_{A_{4}} = 4\) residuals preserved for each arrangement is
For comparison, the approximate resampling probability value based on LAD regression, v = 1, L = 1, 000, 000, and \(C_{i} = n_{A_{i}}/N\) for i = 1, …, 4 is P = 0. 0273.
Following Eq. (4.7) on p. 126, the exact expected value of the M = 63, 063, 000 δ values is μ δ = 68. 0083 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating approximately 95 % agreement between the observed and predicted y values above that expected by chance.
Conventional ANOVA Analysis
A conventional Latin square analysis of variance calculated on the N = 16 univariate response measurement scores listed in Table 4.8 on p. 170170 yields an observed F-ratio of F A = 40. 7277. Assuming independence and normality, F A is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = a - 1 = 4 - 1 = 3\) and \(\nu _{2} = (a - 2)(a - 1) = (4 - 2)(4 - 1) = 6\) degrees of freedom. Under the null hypothesis, the observed value of F A = 40. 7277 yields an approximate probability value of \(P = 0.2204\times 10^{-3}\).
Analysis of Factor B
A design matrix of dummy codes for analyzing Factor B is given in Fig. 4.32, where the first column of 1 values provides for an intercept, the second through fourth columns contain dummy codes for Subjects, the fifth through seventh columns contain dummy codes for Factor A, and the last column lists the univariate response measurement scores ordered by the b = 4 treatment levels of Factor B, with the first \(n_{B_{1}} = 4\) scores, the next \(n_{B_{2}} = 4\) scores, the next \(n_{B_{3}} = 4\) scores, and the last \(n_{B_{4}} = 4\) associated with treatment levels B 1, B 2, B 3, and B 4, respectively. The MRPP regression analysis examines LAD regression residuals for possible differences among the b = 4 treatment levels of Factor B; consequently, no dummy codes are provided for Factor B as this information is implicit in the ordering of the b = 4 treatment levels of Factor B in the last column of Fig. 4.32.
Because there are
possible, equally-likely arrangements of the N = 16 univariate response measurement scores listed in Fig. 4.32, an exact permutation approach is not practical.
LAD Regression Analysis
An MRPP resampling analysis of the N = 16 LAD regression residuals calculated on the univariate response measurement scores in Fig. 4.32 yields estimated LAD regression coefficients of
for Factor B. Figure 4.33 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 16.
Following Eq. (4.5) on p. 125 and employing ordinary Euclidean distance between residuals with v = 1, the N = 16 LAD regression residuals listed in Fig. 4.33 yield b = 4 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.33 with v = 1 and treatment-group weights
is
If all M possible arrangements of the N = 16 observed LAD regression residuals listed in Fig. 4.33 occur with equal chance, the approximate resampling probability value of δ B = 1. 7917 computed on L = 1, 000, 000 random arrangements of the observed LAD regression residuals with \(n_{B_{1}} = \cdots = n_{B_{4}} = 4\) residuals preserved for each arrangement is
Following Eq. (4.7) on p. 126, the exact expected value of the M = 63, 063, 000 δ values is μ δ = 1. 8583 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N is
indicating approximately 4 % agreement between the observed and predicted y values above that expected by chance.
OLS Regression Analysis
For comparison, consider an MRPP resampling analysis of the OLS regression residuals calculated on the N = 16 univariate response measurement scores listed in Fig. 4.29 on p. 171. The MRPP regression analysis yields estimated OLS regression coefficients of
for Factor B. Figure 4.34 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 16.
Following Eq. (4.5) on p. 125 and employing squared Euclidean distance between residuals with v = 2, the N = 16 OLS regression residuals listed in Fig. 4.34 yield b = 4 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the OLS regression residuals listed in Fig. 4.34 with v = 2 and treatment-group weights
is
If all M possible arrangements of the N = 16 observed OLS regression residuals listed in Fig. 4.34 occur with equal chance, the approximate resampling probability value of δ B = 3. 9792 computed on L = 1, 000, 000 random arrangements of the observed OLS regression residuals with \(n_{B_{1}} = n_{B_{2}} = n_{B_{3}} = n_{B_{4}} = 4\) residuals preserved for each arrangement is
For comparison, the approximate resampling probability value based on LAD regression, v = 1, L = 1, 000, 000, and \(C_{i} = n_{B_{i}}/N\) for i = 1, …, 4 is P = 0. 4953.
Following Eq. (4.7) on p. 126, the exact expected value of the M = 63, 063, 000 δ values is μ δ = 4. 0750 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating only approximately 2 % agreement between the observed and predicted y values above that expected by chance.
Conventional ANOVA Analysis
A conventional Latin square analysis of variance calculated on the N = 16 univariate response measurement scores listed in Table 4.8 on p. 170 yields an observed F-ratio of F B = 0. 5602. Assuming independence and normality, F B is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = b - 1 = 4 - 1 = 3\) and \(\nu _{2} = (b - 2)(b - 1) = (4 - 2)(4 - 1) = 6\) degrees of freedom. Under the null hypothesis, the observed value of F B = 0. 5602 yields an approximate probability value of P = 0. 6606. The LAD regression, OLS regression, and F-ratio probability values of P = 0. 4953, P = 0. 3789, and P = 0. 6606, respectively, all indicate that the order in which the treatments were distributed did not matter.
4.3.7 Split-Plot Design
Imagine a testing experiment with two treatment factors, A and B, with a and b treatment levels, respectively, so that there are ab treatment combinations. If each testing session requires h hours of a subject’s time and every subject is to be treated under all treatment conditions, each subject will require ab testing sessions and abh hours of testing time. When this is unreasonable, then with S subjects available, assign \(n = S/A\) subjects to each level of Factor A and test each subject under all levels of Factor B. The design is a repeated-measures split-plot design in which subjects are randomly assigned to the a treatment levels of Factor A (i.e., plots), and each subject is then tested under all b levels of Factor B (i.e., subplots). The design is also called a mixed factorial design with one between-subjects factor (A) and one within-subjects factor (B), or an A×(B×S) design [214].
Consider a split-plot experiment in which Factor A has a = 3 treatment levels, Factor B has b = 3 treatment levels, n = 12 subjects are randomly assigned to each of the a = 3 levels of Factor A, and each subject is tested at all b = 3 levels of Factor B. The design and data are adapted from Keppel and Zedeck and are given in Fig. 4.35 [215, p. 303].
Analysis of Factor A
A design matrix of effect codes for an MRPP regression analysis of Factor A is given in Fig. 4.36, where the first column of 1 values provides for an intercept and the second column lists the total of response measurement summations over the b levels of Factor B (e.g., \(53 + 51 + 35 = 139\)). The summations are ordered by the a = 3 treatment levels of Factor A with the first \(n_{A_{1}} = 4\) summations, the second \(n_{A_{2}} = 4\) summations, and the last \(n_{A_{3}} = 4\) summations associated with treatment levels A 1, A 2, and A 3, respectively. The MRPP regression analysis examines the N = 12 regression residuals for possible differences among the a = 3 treatment levels of Factor A; consequently, no effect codes are provided for Factor A as this information is implicit in the ordering of the a = 3 treatment levels of Factor A in the second column of Fig. 4.36.
An exact permutation solution is feasible for the response measurement summations listed in Fig. 4.36 since there are only
possible, equally-likely arrangements of the N = 12 response measurement summations for Factor A.
LAD Regression Analysis
An MRPP analysis of the N = 12 LAD regression residuals calculated on the response measurement summations in Fig. 4.36 yields an estimated LAD regression coefficient of \(\tilde{\beta }_{0} = +101.00\) for Factor A. Figure 4.37 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 12.
Following Eq. (4.5) on p. 125 and employing ordinary Euclidean distance between residuals with v = 1, the N = 12 LAD regression residuals listed in Fig. 4.37 yield a = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.37 with v = 1 and treatment-group weights
is
If all arrangements of the N = 16 observed LAD regression residuals listed in Fig. 4.37 occur with equal chance, the exact probability value of δ A = 19. 4444 calculated on the M = 34, 650 possible arrangements of the observed LAD regression residuals with \(n_{A_{1}} = n_{A_{2}} = n_{A_{3}} = 4\) residuals preserved for each arrangement is
Following Eq. (4.7) on p. 126, the exact expected value of the M = 34, 650 δ values is μ δ = 27. 00 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is
indicating approximately 28 % agreement between the observed and predicted y values above that expected by chance.
OLS Regression
For comparison, consider an MRPP analysis of OLS regression residuals calculated on the response measurement summations for Factor A in Fig. 4.36. The MRPP regression analysis yields an estimated OLS regression coefficient of \(\hat{\beta }_{0} = +100.50\) for Factor A. Figure 4.38 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 12.
Following Eq. (4.5) on p. 125 and employing squared Euclidean distance between residuals with v = 2, the N = 12 OLS regression residuals listed in Fig. 4.38 yield a = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the OLS regression residuals listed in Fig. 4.38 with v = 2 and treatment-group weights
is
If all arrangements of the N = 12 observed OLS regression residuals listed in Fig. 4.38 occur with equal chance, the exact probability value of δ A = 527. 3333 computed on the M = 34, 650 possible arrangements of the observed OLS regression residuals with \(n_{A_{1}} = n_{A_{2}} = n_{A_{3}} = 4\) residuals preserved for each arrangement is
For comparison, the exact probability value based on LAD regression, v = 1, M = 34, 650, and \(C_{i} = n_{A_{i}}/N\) for i = 1, 2, 3 is P = 0. 0194.
Following Eq. (4.7) on p. 126, the exact expected value of the M = 34, 650 δ values is μ δ = 1, 082. 7273 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating approximately 51 % agreement between the observed and predicted y values above that expected by chance.
Conventional ANOVA Analysis
A conventional split-plot analysis of variance calculated on the N = 12 univariate response measurement scores listed in Fig. 4.35 on p. 180 yields an observed F-ratio of F A = 6. 7927. Assuming independence, normality, and homogeneity of variance, F A is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = a - 1 = 3 - 1 = 2\) and \(\nu _{2} = a(n - 1) = 3(4 - 1) = 9\) degrees of freedom. Under the null hypothesis, the observed value of F A = 6. 7927 yields an approximate probability value of P = 0. 0159.
Analysis of Factor B
A design matrix of effect codes for an MRPP regression analysis of Factor B is given in Table 4.9, where the first column of 1 values provides for an intercept, the next 11 columns contain effect codes for Subjects nested within Factor A, and the next four columns contain effect codes for the A×B interaction. The last column lists the N = 36 univariate response measurement scores ordered by the b = 3 treatment levels of Factor B, with the first \(n_{B_{1}} = 12\) scores, the next \(n_{B_{2}} = 12\) scores, and the last \(n_{B_{3}} = 12\) scores associated with treatment levels B 1, B 2, and B 3, respectively. The MRPP regression analysis examines the N = 36 regression residuals for possible differences among the b = 3 treatment levels of Factor B; consequently, no effect codes are provided for Factor B as this information is implicit in the ordering of the b = 3 treatment levels of Factor B in the last column of Table 4.9.
Because there are
possible, equally-likely arrangements of the N = 36 univariate response measurement scores listed in Table 4.9, an exact permutation approach is not possible.
LAD Regression Analysis
An MRPP resampling analysis of the LAD regression residuals calculated on the N = 36 univariate response measurement scores in Table 4.9 yields estimated LAD regression coefficients of
for Factor B. Figure 4.39 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 36.
Following Eq. (4.5) on p. 125 and employing ordinary Euclidean distance between residuals with v = 1, the N = 36 LAD regression residuals listed in Fig. 4.39 yield b = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.39 with v = 1 and treatment-group weights
is
If all M possible arrangements of the N = 36 observed LAD regression residuals listed in Fig. 4.39 occur with equal chance, the approximate resampling probability value of δ B = 7. 8283 computed on L = 1, 000, 000 random arrangements of the observed LAD regression residuals with \(n_{B_{1}} = n_{B_{2}} = n_{B_{3}} = 12\) residuals preserved for each arrangement is
which may be interpreted as a probability of less than one in a million.
When M is very large and the probability of an observed δ is extremely small, as in this case, resampling permutation procedures sometimes result in zero probability, even with L = 1, 000, 000 random arrangements of the observed regression residuals. A reanalysis of Factor B using L = 10, 000, 000 random arrangements of the observed data yielded an identical resampling probability value of P = 0. 00. Moment-approximation permutation procedures, described briefly in Chap. 1, Sect. 1.2.2, can often provide results in these extreme situations. The moment-approximation of a test statistic requires computation of the exact moments of the test statistic, assuming equally-likely arrangements of the observed regression residuals [284, 300]. Usually, the first three exact moments are used: the exact mean, μ δ , the exact variance, \(\sigma _{\delta }^{2}\), and the exact skewness, γ δ , of δ. The three moments are then used to fit a specified distribution, such as a Pearson type III distribution, that approximates the underlying discrete permutation distribution and provides an approximate probability value. For Factor B, a moment-approximation procedure yields δ B = 7. 8283, μ δ = 12. 5460, \(\sigma _{\delta }^{2} = 0.1675\), \(\gamma _{\delta } = -1.3580\), a standardized test statistic of
and a Pearson type III approximate probability value of \(P = 0.1495\times 10^{-6}\).
Following Eq. (4.7) on p. 126, the exact expected value of the Mδ values is μ δ = 12. 5460 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is
indicating approximately 38 % agreement between the observed and predicted y values above that expected by chance.
OLS Regression Analysis
For comparison, consider an MRPP analysis of OLS regression residuals calculated on the N = 36 response measurement summations for Factor B in Table 4.9 on p. 184. The MRPP regression analysis yields estimated OLS regression coefficients of
for Factor B. Figure 4.40 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 36.
Following Eq. (4.5) on p. 125 and employing squared Euclidean distance between residuals with v = 2, the N = 36 OLS regression residuals listed in Fig. 4.40 yield b = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the OLS regression residuals listed in Fig. 4.40 with v = 2 and treatment-group weights
is
If all M possible arrangements of the N = 36 observed OLS regression residuals listed in Fig. 4.40 occur with equal chance, the approximate resampling probability value of δ B = 31. 0909 computed on L = 1, 000, 000 random arrangements of the observed OLS regression residuals with \(n_{B_{1}} = n_{B_{2}} = n_{B_{3}} = 12\) residuals preserved for each arrangement is
i.e., a probability of less than one in a million. For comparison, the approximate resampling probability value based on LAD regression, v = 1, L = 1, 000, 000, and \(C_{i} = n_{B_{i}}/N\) for i = 1, 2, 3 is also P = 0. 00.
As with the analysis of the LAD regression residuals listed in Fig. 4.39 on p. 186, when M is large and the probability of an observed δ is very small, an alternative moment procedure based on the exact mean, μ δ , exact variance, \(\sigma _{\delta }^{2}\), and exact skewness, γ δ , of δ can be employed to obtain approximate probability values; see Chap. 1, Sect. 1.2.2. For Factor B, a moment-approximation procedure yields δ B = 31. 0909, μ δ = 199. 2857, \(\sigma _{\delta }^{2} = 134.8578\), \(\gamma _{\delta } = -1.7697\), a standardized test statistic of
and a Pearson type III approximate probability value of \(P = 0.5420\times 10^{-7}\).
Following Eq. (4.7) on p. 126, the exact expected value of the Mδ values is μ δ = 199. 2857 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating approximately 84 % agreement between the observed and predicted y values above that expected by chance.
Conventional ANOVA Analysis
A conventional split-plot analysis of variance calculated on the N = 36 univariate response measurement scores listed in Fig. 4.35 on p. 180 yields an observed F-ratio of F B = 52. 1842. Assuming independence and normality, F B is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = b - 1 = 3 - 1 = 2\) and \(\nu _{2} = a(n - 1)(b - 1) = 3(4 - 1)(3 - 1) = 18\) degrees of freedom. Under the null hypothesis, the observed value of F B = 52. 1842 yields an approximate probability value of \(P = 0.3224\times 10^{-7}\).
Analysis of the A×B Interaction
A design matrix of effect codes for an MRPP regression analysis of the A×B interaction is given in Table 4.10, where the first column of 1 values provides for an intercept, the next 11 columns contain effect codes for Subjects nested within Factor A, and the next two columns contain effect codes for Factor B. The last column lists the N = 36 univariate response measurement scores ordered by the \(ab = (3)(3) = 9\) levels of the A×B interaction. The MRPP regression analysis examines the N = 36 regression residuals for possible differences among the nine treatment levels of the A×B interaction; consequently, no effect codes are provided for the A×B interaction as this information is implicit in the ordering of the treatment levels of the A×B interaction in the last column of Table 4.10.
Because there are
possible, equally-likely arrangements of the N = 36 univariate response measurement scores listed in Table 4.10, an exact permutation approach is not possible.
LAD Regression Analysis
An MRPP resampling analysis of the N = 36 LAD regression residuals calculated on the univariate response measurement scores in Table 4.10 yields estimated LAD regression coefficients of
for the interaction of Factors A and B. Figure 4.41 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 36.
Following Eq. (4.5) on p. 125 and employing ordinary Euclidean distance between residuals with v = 1, the N = 36 LAD regression residuals listed in Fig. 4.41 yield \(ab = (3)(3) = 9\) average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.41 with v = 1 and treatment-group weights
is
If all M possible arrangements of the N = 36 observed LAD regression residuals listed in Fig. 4.41 occur with equal chance, the approximate resampling probability value of δ A×B = 5. 2593 computed on L = 1, 000, 000 random arrangements of the observed LAD regression residuals with \(n_{(A\times B)_{1}} = \cdots = n_{(A\times B)_{9}} = 4\) residuals preserved for each arrangement is
Following Eq. (4.7) on p. 126, the exact expected value of the Mδ values is μ δ = 5. 6825 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is
indicating approximately 7 % agreement between the observed and predicted y values above that expected by chance.
OLS Regression Analysis
For comparison, consider an MRPP analysis of OLS regression residuals calculated on the N = 36 response measurement scores for the A×B interaction in Table 4.10. The MRPP regression analysis yields estimated OLS regression coefficients of
for the interaction of Factors A and B. Figure 4.42 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 36.
Following Eq. (4.5) on p. 125 and employing squared Euclidean distance between residuals with v = 2, the N = 36 OLS regression residuals listed in Fig. 4.42 yield \(ab = (3)(3) = 9\) average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the OLS regression residuals listed in Fig. 4.42 with v = 2 and treatment-group weights
is
If all M possible arrangements of the N = 36 observed OLS regression residuals listed in Fig. 4.42 occur with equal chance, the approximate resampling probability value of δ A×B = 38. 00 calculated on L = 1, 000, 000 random arrangements of the observed OLS regression residuals with \(n_{(A\times B)_{1}} = \cdots = n_{(A\times B)_{9}} = 4\) residuals preserved for each arrangement is
For comparison, the approximate resampling probability value based on LAD regression, v = 1, L = 1, 000, 000, and \(C_{i} = n_{(A\times B)_{i}}/N\) for i = 1, …, 9 is P = 0. 1402.
Following Eq. (4.7) on p. 126, the exact expected value of the Mδ values is μ δ = 47. 6286 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating approximately 20 % agreement between the observed and predicted y values above that expected by chance.
Conventional ANOVA Analysis
A conventional split-plot analysis of variance calculated on the N = 36 univariate response measurement scores listed in Fig. 4.35 on p. 180 yields an observed F-ratio of F A×B = 2. 8114. Assuming independence, normality, and homogeneity of variance, F A×B is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = (a - 1)(b - 1) = (3 - 1)(3 - 1) = 4\) and \(\nu _{2} = a(n - 1)(b - 1) = 3(4 - 1)(3 - 1) = 18\) degrees of freedom. Under the null hypothesis, the observed value of F A×B = 2. 8114 yields an approximate probability value of P = 0. 0565, which is similar to the probability value of P = 0. 0723 obtained with the OLS regression analysis.
4.3.8 Nested Design
It is sometimes necessary to compare treatment groups when one independent variable is nested under a second independent variable. Two-factor nested analysis-of-variance designs occur whenever one factor is not completely crossed with the second factor. Consider a nested design to compare a = 3 levels of Factor A on scores obtained from b = 3 levels of Factor B, with B 1, B 2, and B 3 of Factor B in level A 1; B 4, B 5, and B 6 of Factor B in level A 2, and B 7, B 8, and B 9 of Factor B in level A 3. Thus, Factor B is said to be nested under Factor A. The univariate data for this example are listed in Table 4.11 for a sample of n = 4 objects randomly chosen from each of the \(ab = (3)(3) = 9\) levels of Factors A and B.
Analysis of Factor A
A design matrix of effect codes for an MRPP regression analysis of Factor A is given in Table 4.12, where the first column of 1 values provides for an intercept, the next two columns contain the effect codes for Factor B, and the third column contains the univariate response measurement scores listed according to the original random assignment of the n = 36 objects to the a = 3 levels of Factor A with the first \(n_{A_{1}} = 12\) scores, the next \(n_{A_{2}} = 12\) scores, and the last \(n_{A_{3}} = 12\) scores associated with the a = 3 levels of Factor A, respectively. The MRPP regression analysis examines the N = 36 regression residuals for possible differences among the a = 3 treatment levels of Factor A; consequently, no effect codes are provided for Factor A as this information is implicit in the ordering of the a = 3 levels of Factor A in the rightmost columns of Table 4.12.
Because there are
possible, equally-likely arrangements of the N = 36 univariate response measurement scores listed in Table 4.11, an exact permutation approach is not possible.
LAD Regression Analysis
An MRPP resampling analysis of the N = 36 LAD regression residuals calculated on the univariate response measurement scores listed in Table 4.12 yields estimated LAD regression coefficients of
for Factor A. Figure 4.43 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 36.
Following Eq. (4.5) on p. 125 and employing ordinary Euclidean distance between residuals with v = 1, the N = 36 LAD regression residuals listed in Fig. 4.43 yield a = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.43 with v = 1 and treatment-group weights
is
If all M possible arrangements of the N = 36 observed LAD regression residuals listed in Fig. 4.43 occur with equal chance, the approximate resampling probability value of δ A = 3. 3131 computed on L = 1, 000, 000 random arrangements of the observed LAD regression residuals with \(n_{A_{1}} = n_{A_{2}} = n_{A_{3}} = 12\) residuals preserved for each arrangement is
Following Eq. (4.7) on p. 126, the exact expected value of the Mδ values is μ δ = 3. 2508 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size between the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is
indicating slightly less than chance agreement between the observed and predicted y values.
OLS Regression Analysis
For comparison, consider an MRPP resampling analysis of OLS regression residuals calculated on the N = 36 univariate response measurement scores listed in Table 4.12. The MRPP regression analysis yields estimated OLS regression coefficients of
for Factor A. Figure 4.44 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 36.Footnote 4
Following Eq. (4.5) on p. 125 and employing squared Euclidean distance between residuals with v = 2, the N = 36 OLS regression residuals listed in Fig. 4.44 yield a = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the OLS regression residuals listed in Fig. 4.44 with v = 2 and treatment-group weights
is
If all M possible arrangements of the N = 36 observed OLS regression residuals listed in Fig. 4.44 occur with equal chance, the approximate resampling probability value of δ A = 16. 9697 computed on L = 1, 000, 000 random arrangements of the observed OLS regression residuals with \(n_{A_{1}} = n_{A_{2}} = n_{A_{3}} = 12\) residuals preserved for each arrangement is
A reanalysis of the data based on L = 10, 000, 000 random arrangements of the N = 36 observed regression residuals listed in Fig. 4.44 with \(n_{A_{1}} = n_{A_{2}} = n_{A_{3}} = 12\) residuals preserved for each arrangement also yields an approximate resampling probability value of P = 1. 00.
A probability value of P = 1. 00 is not very informative. In such cases, an alternative moment procedure based on the exact mean, μ δ , exact variance, \(\sigma _{\delta }^{2}\), and exact skewness, γ δ , of δ can be employed to obtain approximate probability values; see Chap. 1, Sect. 1.2.2. For Factor A, a moment-approximation procedure yields δ A = 16. 9697, μ δ = 16. 00, \(\sigma _{\delta }^{2} = 0.8472\), \(\gamma _{\delta } = -1.7012\), an observed standardized test statistic of
and a Pearson type III approximate probability value of P = 0. 9487.
For comparison, the approximate resampling probability value based on LAD regression, v = 1, L = 1, 000, 000, and \(C_{i} = n_{A_{i}}/N\) for i = 1, …, a is P = 0. 7048.
Following Eq. (4.7) on p. 126, the exact expected value of the Mδ values is μ δ = 16. 00 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating slightly less than chance agreement between the observed and predicted y values.
Conventional ANOVA Analysis
A conventional fixed-effects nested analysis of variance calculated on the N = 36 response measurement scores for Factor A listed in Table 4.11 on p. 196 yields an observed F-ratio of F A = 3. 6818. Assuming independence, normality, and homogeneity of variance, F A is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = a - 1 = 3 - 1 = 2\) and \(\nu _{2} = ab(n - 1) = (3)(3)(4 - 1) = 27\) degrees of freedom. Under the null hypothesis, the observed value of F A = 3. 6818 yields an approximate probability value of P = 0. 0386.
Analysis of Factor B | A
A design matrix of effect codes for an MRPP regression analysis of Factor B, nested under Factor A, is given in Fig. 4.45, where the first column of 1 values provides for an intercept, the next two columns contain effect codes for Factor A, the next four columns contain effect codes for the A×B interaction, and the last column contains the univariate response measurement scores listed according to the b = 3 levels of Factor B with the first \(n_{B\vert A_{1}} = 12\) scores, the next \(n_{B\vert A_{2}} = 12\) scores, and the last \(n_{B\vert A_{3}} = 12\) scores associated with the b = 3 levels of Factor B, respectively. The MRPP regression analysis examines the N = 36 regression residuals for possible differences among the b = 3 treatment levels of Factor B; consequently, no effect codes are provided for Factor B as this information is implicit in the ordering of the b = 3 levels of Factor B in the last column of Fig. 4.45.
LAD Regression Analysis
Again, because there are
possible, equally-likely arrangements of the N = 36 response measurement scores listed in Fig. 4.45, an exact permutation approach is not possible. An MRPP resampling analysis of the N = 36 LAD regression residuals calculated on the univariate response measurement scores in Fig. 4.45 yields estimated LAD regression coefficients of
for Factor B | A. Figure 4.46 lists the observed y i values, LAD predicted \(\tilde{y}_{i}\) values, and residual e i values for i = 1, …, 36.
Following Eq. (4.5) on p. 125 and employing ordinary Euclidean distance between residuals with v = 1, the N = 36 LAD regression residuals listed in Fig. 4.46 yield a = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.46 with v = 1 and treatment-group weights
is
If all M possible arrangements of the N = 36 observed LAD regression residuals listed in Fig. 4.46 occur with equal chance, the approximate resampling probability value of δ B | A = 2. 00 computed on L = 1, 000, 000 random arrangements of the observed LAD regression residuals with \(n_{B\vert A_{1}} = n_{B\vert A_{2}} = n_{B\vert A_{3}} = 12\) residuals preserved for each arrangement is
Following Eq. (4.7) on p. 126, the exact expected value of the Mδ values is μ δ = 2. 0127 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, i = 1, …, N, is
indicating approximately chance agreement between the observed and predicted y values.
OLS Regression Analysis
For comparison, consider an MRPP resampling analysis of OLS regression residuals calculated on the N = 36 response measurement scores listed in Table 4.12 on p. 197. The MRPP regression analysis yields estimated OLS regression coefficients of
for Factor B | A. Figure 4.47 lists the observed y i values, OLS predicted \(\hat{y}_{i}\) values, and residual e i values for i = 1, …, 36.
Following Eq. (4.5) on p. 125 and employing squared Euclidean distance between residuals with v = 2, the N = 36 OLS regression residuals listed in Fig. 4.47 yield a = 3 average distance-function values of
Following Eq. (4.4) on p. 125, the observed value of the MRPP test statistic calculated on the OLS regression residuals listed in Fig. 4.47 with v = 2 and treatment-group weights
is
If all M possible arrangements of the N = 36 observed OLS regression residuals listed in Fig. 4.47 occur with equal chance, the approximate resampling probability value of δ B | A = 5. 8182 computed on L = 1, 000, 000 random arrangements of the observed OLS regression residuals with \(n_{B\vert A_{1}} = n_{B\vert A_{2}} = n_{B\vert A_{3}} = 12\) residuals preserved for each arrangement is
For comparison, the approximate resampling probability value based on LAD regression, v = 1, L = 1, 000, 000, and \(C_{i} = n_{B\vert A_{i}}/N\) for i = 1, 2, 3 is P = 0. 3616.
Following Eq. (4.7) on p. 126, the exact expected value of the Mδ values is μ δ = 5. 4857 and, following Eq. (4.6) on p. 126, the observed chance-corrected measure of effect size for the y i and \(\hat{y}_{i}\) values, i = 1, …, N, is
indicating slightly less than chance agreement between the observed and predicted y values.
Conventional ANOVA Analysis
A conventional fixed-effects nested analysis of variance calculated on the N = 36 univariate response measurement scores for Factor B | A listed in Table 4.11 on p. 196 yields an observed F-ratio of F B | A = 10. 6362. Assuming independence, normality, and homogeneity of variance, F B | A is approximately distributed as Snedecor’s F under the null hypothesis with \(\nu _{1} = a(b - 1) = 3(3 - 1) = 6\) and \(\nu _{2} = ab(n - 1) = (3)(3)(4 - 1) = 27\) degrees of freedom. Under the null hypothesis, the observed value of F B | A = 10. 6362 yields an approximate probability value of \(P = 4.5461\times 10^{-6}\).
4.4 Multivariate Multiple Regression Designs
An extension of LAD multiple regression to include multiple dependent variables, as well as multiple independent variables, i.e., multivariate multiple LAD regression, is developed in this section. The extension was prompted by a multivariate Least Sum (of) Euclidean Distances (LSED) algorithm developed by Kaufman , Taylor , Mielke , and Berry in 2002 [198].
Consider the multivariate multiple regression model given by
for i = 1, …, N and k = 1, …, r, where y ik represents the ith of N measurements for the kth of r response variables, possibly affected by a treatment; x ij is the jth of m covariates associated with the ith response, where x i1 = 1 if the model includes an intercept; β jk denotes the jth of m regression parameters for the kth of r response variables; and e ik designates the error associated with the ith of N measurements for the k of r response variables.
If estimates of β jk that minimize
are denoted by \(\tilde{\beta }_{jk}\) for j = 1, …, m and k = 1, …, r, then the Nr-dimensional residuals of the LSED multivariate multiple regression model are given by
for i = 1, …, N and k = 1, …, r.
Let the Nr-dimensional residuals, (e i1, …, e ir ) for i = 1, …, N obtained from an LSED multivariate multiple regression model, be partitioned into g treatment groups of sizes n 1, …, n g , where n i ≥ 2 for i = 1, …, g and
The MRPP analysis of the multivariate multiple regression residuals depends on statistic
where \(C_{i} = n_{i}/N\) is a positive weight for the ith of g treatment groups and \(\xi _{i}\) is the average pairwise Euclidean distance among the n i r-dimensional residuals in the ith of g treatment groups defined by
where
The null hypothesis specifies that each of the
possible allocations of the Nr-dimensional residuals to the g treatment groups is equally likely. An exact MRPP probability value associated with the observed value of δ, δ o, is given by
As previously, when M is large an approximate probability value may be obtained from a resampling permutation procedure. Let L denote a large random sample drawn from all M possible arrangements of the observed data, then an approximate resampling probability value is given by
As with univariate multiple regression models, the criterion for fitting multivariate multiple regression models based on δ is the chance-corrected measure of effect size between the observed and predicted response measurement values given by
where μ δ is the expected value of δ over the N! possible pairings under the null hypothesis, given by
4.4.1 Example Analysis
To illustrate a multivariate LSED multiple regression analysis, consider an unbalanced two-way randomized-block experimental design in which N = 16 subjects (S ) are tested over a = 3 levels of Factor A, the experiment is repeated b = 2 times for Factor B, and there are r = 2 response measurement scores for each subject. The design and data are adapted from Mielke and Berry [297, p. 184] and are given in Fig. 4.48. The design is intentionally kept small to illustrate the multivariate multiple regression procedure.
Analysis of Factor A
A design matrix of dummy codes for an MRPP regression analysis of Factor A is given in Fig. 4.49, where the first column of 1 values provides for an intercept, the next column contains the dummy codes for Factor B, and the third and fourth columns contain the bivariate response measurement scores listed according to the original random assignment of the N = 16 subjects to the a = 3 levels of Factor A with the first \(n_{A_{1}} = 5\) scores, the next \(n_{A_{2}} = 7\) scores, and the last \(n_{A_{3}} = 4\) scores associated with the a = 3 levels of Factor A, respectively. The MRPP regression analysis examines the N = 16 regression residuals for possible differences among the a = 3 treatment levels of Factor A; consequently, no dummy codes are provided for Factor A as this information is implicit in the ordering of the a = 3 levels of Factor A in the last two columns of Fig. 4.49.
Because there are only
possible, equally-likely arrangements of the N = 16 bivariate response measurement scores listed in Fig. 4.49, an exact permutation approach is feasible. An MRPP analysis of the N = 16 LAD regression residuals calculated on the bivariate response measurements for Factor A in Fig. 4.49 yields estimated LAD regression coefficients of
for Factor A. Figure 4.50 lists the observed y ik values, LAD predicted \(\tilde{y}_{ik}\) values, and residual e ik values for i = 1, …, 16 and k = 1, 2.
Following Eq. (4.9) on p. 208 and employing ordinary Euclidean distance between residuals with v = 1, the N = 16 LAD regression residuals listed in Fig. 4.50 yield a = 3 average distance-function values of
Following Eq. (4.8) on p. 208, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.50 with v = 1 and treatment-group weights
is
If all arrangements of the N = 16 observed LAD regression residuals listed in Fig. 4.50 occur with equal chance, the exact probability value of δ A = 12. 8587 computed on the M = 1, 441, 440 possible arrangements of the observed LAD regression residuals with \(n_{A_{1}} = 5\), \(n_{A_{2}} = 7\), and \(n_{A_{3}} = 4\) residuals preserved for each arrangement is
Following Eq. (4.11) on p. 209, the exact expected value of the M = 1, 441, 440 δ values is μ δ = 18. 1020 and, following Eq. (4.10) on p. 209, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, \(i = 1,\ldots,N\), is
indicating approximately 29 % agreement between the observed and predicted values above that expected by chance.
Analysis of Factor B
A design matrix of dummy codes for an MRPP regression analysis of Factor B is given in Fig. 4.51, where the first column of 1 values provides for an intercept, the next two columns contain the dummy codes for Factor A, and the fourth and fifth columns contain the bivariate response measurement scores listed according to the original random assignment of the N = 16 subjects to the b = 2 levels of Factor B with the first \(n_{B_{1}} = 7\) scores and the last \(n_{B_{2}} = 9\) scores associated with the b = 2 levels of Factor B, respectively. The MRPP regression analysis examines the N = 16 regression residuals for possible differences between the b = 2 treatment levels of Factor B; consequently, no dummy codes are provided for Factor B as this information is implicit in the ordering of the b = 2 levels of Factor B in the last two columns of Fig. 4.51.
Because there are only
possible, equally-likely arrangements of the N = 16 response measurement scores listed in Fig. 4.51, an exact permutation approach is feasible. An MRPP analysis of the N = 16 LAD regression residuals calculated on the bivariate response measurements for Factor B in Fig. 4.51 yields estimated LAD regression coefficients of
for Factor B. Figure 4.52 lists the observed y ik values, LAD predicted \(\tilde{y}_{ik}\) values, and residual e ik values for i = 1, …, 16 and k = 1, 2.
Following Eq. (4.9) on p. 208 and employing ordinary Euclidean distance between residuals with v = 1, the N = 16 LAD regression residuals listed in Fig. 4.52 yield b = 2 average distance-function values of
Following Eq. (4.4) on p. 208, the observed value of the MRPP test statistic calculated on the LAD regression residuals listed in Fig. 4.52 with v = 1 and treatment-group weights
is
If all arrangements of the N = 16 observed LAD regression residuals listed in Fig. 4.52 occur with equal chance, the exact probability value of δ B = 12. 0535 computed on the M = 11, 440 possible arrangements of the observed LAD regression residuals with \(n_{B_{1}} = 7\) and \(n_{B_{2}} = 9\) residuals preserved for each arrangement is
Following Eq. (4.11) on p. 209, the exact expected value of the M = 11, 440 δ values is μ δ = 12. 2923 and, following Eq. (4.10) on p. 209, the observed chance-corrected measure of effect size for the y i and \(\tilde{y}_{i}\) values, \(i = 1,\ldots,N\), is
indicating approximately 2 % agreement between the observed and predicted values above that expected by chance.
4.5 Coda
Chapter 4 applied the Multi-Response Permutation Procedures (MRPP) developed in Chap. 2 to interval-level response measurements, utilizing dummy and effect coding of treatment groups to generate regression residuals from LAD regression models, subsequently analyzed with MRPP. Considered in this chapter were one-way randomized, one-way randomized with a covariate, one-way randomized-block, two-way randomized-block, two-way factorial, Latin square, split-plot, and two-factor nested designs. Chapter 4 concluded with example multivariate multiple regression designs.
Comparisons of permutation-based LAD regression with ordinary Euclidean distance between response measurements, permutation-based OLS regression with squared Euclidean distance between response measurements, and conventional OLS regression with squared Euclidean distance between response measurements in Chap. 4, revealed that considerable differences can exist among the three approaches that are not systematic. Oftentimes, one of the three approaches yielded the lowest of the three probability values, while other times the same approach yielded the highest probability value. Sometimes the three approaches yielded the same, or nearly the same, probability value, as was the case with the analysis of Factor B in the two-way randomized-block design example, and other times the three probability values were markedly different, as was the case with the analysis of the A×B interaction in the two-way factorial design example. In general, permutation-based LAD regression, coupled with MRPP and ordinary Euclidean distance between response measurements, is recommended due to the lack of restrictive assumptions and robustness that is possible with extreme values.
Chapter 5
Chapter 5 establishes the relationships between the MRPP test statistics, δ and \(\mathfrak{R}\), and selected conventional tests and measures designed for the analysis of completely randomized data at the ordinal level of measurement. Considered in Chap. 5 are the Wilcoxon two-sample rank-sum test, the Kruskal–Wallis multiple-sample rank-sum test, the Mood rank-sum test for dispersion, the Brown–Mood median test, the Mielke power-of-rank functions, the Whitfield two-sample rank-sum test, and the Cureton rank-biserial test.
Notes
- 1.
For the remainder of this chapter, a tilde over a β (\(\tilde{\beta }\)) indicates an unstandardized LAD regression coefficient, while a caret over a β (\(\hat{\beta }\)) indicates an unstandardized OLS regression coefficient.
- 2.
All the biologically inherited information is not carried in the genes of a cell’s nucleus. A small number of genes are carried by intra-cellular bodies, the mitochondria. Thus, the result of cloning is not, strictly speaking, a perfect genetic clone of the donor organism.
- 3.
This method of estimation is known as Method I as presented in a seminal article by Overall and Spiegel in 1969 [330].
- 4.
Note that in the case of Factor A, LAD regression and OLS regression yield the same regression coefficients. Therefore, the observed regression residuals are the same for both analyses.
References
Agresti, A.: Measures of nominal-ordinal association. J. Am. Stat. Assoc. 76, 524–529 (1981)
Agresti, A.: Categorical Data Analysis, 2nd edn. Wiley, New York (2002)
Agresti, A., Finley, B.: Statistical Methods for the Social Sciences. Prentice-Hall, Upper Saddle River (1997)
Agresti, A., Liu, I.: Modeling a categorical variable allowing arbitrarily many category choices. Biometrics 55, 936–943 (1999)
Agresti, A., Liu, I.: Strategies for modeling a categorical variable allowing multiple category choies. Sociol. Method Res. 29, 403–434 (2001)
Altman, D.G., Bland, J.M.: Measurement in medicine: the analysis of method comparison studies. Statistician 32, 307–317 (1983)
Anderson, T.W.: An Introduction to Multivariate Statistical Analysis, 2nd edn. Wiley, New York (1984)
Anderson, T.W.: Two of Harold Hotelling’s contributions to multivariate analysis. Tech. Rep. 40, Stanford University, Stanford (1990)
Anderson, D.R., Sweeney, D.J., Williams, T.A.: Introduction to Statistics: Concepts and Applications. West, New York (1994)
Ansari, A.R., Bradley, R.A.: Rank-sum tests for dispersion. Ann. Math. Stat. 31, 1174–1189 (1960)
Anscombe, F.J.: Rejection of outliers. Technometrics 2, 123–147 (1960)
Arabie, P.: Was Euclid an unnecessarily sophisticated psychologist? Psychometrika 56, 567–587 (1991)
Arbuckle, J., Aiken, L.S.: A program for Pitman’s permutation test for differences in location. Behav. Res. Methods Instrum. 7, 381 (1975)
Author: Resampling Stats User’s Guide. Resampling Stats, Arlington (1999)
Author: StatXact for Windows. Cytel Software, Cambridge (2000)
Bailer, A.J.: Testing variance equality with randomization tests. J. Stat. Comput. Simul. 31, 1–8 (1989)
Bakan, D.: The test of significance in psychological research. Psychol. Bull. 66, 423–437 (1966)
Bakeman, R., Robinson, B.F., Quera, V.: Testing sequential association: estimating exact p values using sampled permutations. Psychol. Methods 1, 4–15 (1996)
Bartko, J.J.: On various intraclass correlation reliability coefficients. Psychol. Bull. 83, 762–765 (1976)
Bartko, J.J.: Measurement and reliability: statistical thinking considerations. Schizophr. Bull. 17, 483–489 (1991)
Bartlett, M.S.: A note on tests of significance in multivariate analysis. Proc. Camb. Philos. Soc. 34, 33–40 (1939)
Bernardin, H.J., Beatty, R.W.: Performance Appraisal: Assessing Human Behavior at Work. Kent, Boston (1984)
Berry, K.J., Mielke, P.W.: Moment approximations as an alternative to the F test in analysis of variance. Br. J. Math. Stat. Psychol. 36, 202–206 (1983)
Berry, K.J., Mielke, P.W.: An APL function for Radlow and Alf’s exact chi-square test. Behav. Res. Methods Instrum. Comput. 17, 131–132 (1985)
Berry, K.J., Mielke, P.W.: Goodman and Kruskal’s tau-b statistic: a nonasymptotic test of significance. Sociol. Methods Res. 13, 543–550 (1985)
Berry, K.J., Mielke, P.W.: Subroutines for computing exact chi-square and Fisher’s exact probability tests. Educ. Psychol. Meas. 45, 153–159 (1985)
Berry, K.J., Mielke, P.W.: A generalization of Cohen’s kappa agreement measure to interval measurement and multiple raters. Educ. Psychol. Meas. 48, 921–933 (1988)
Berry, K.J., Mielke, P.W.: A family of multivariate measures of association for nominal independent variables. Educ. Psychol. Meas. 52, 41–55 (1992)
Berry, K.J., Mielke, P.W.: Spearman’s footrule as a measure of agreement. Psychol. Rep. 80, 839–846 (1997)
Berry, K.J., Mielke, P.W.: Extension of Spearman’s footrule to multiple rankings. Psychol. Rep. 82, 376–378 (1998)
Berry, K.J., Mielke, P.W.: Least absolute regression residuals: analyses of block designs. Psychol. Rep. 83, 923–929 (1998)
Berry, K.J., Mielke, P.W.: Least sum of absolute deviations regression: distance, leverage, and influence. Percept. Mot. Skills 86, 1063–1070 (1998)
Berry, K.J., Mielke, P.W.: Least sum of Euclidean regression residuals: estimation of effect size. Psychol. Rep. 91, 955–962 (2002)
Berry, K.J., Mielke, P.W.: Longitudinal analysis of data with multiple binary category choices. Psychol. Rep. 93, 127–131 (2003)
Berry, K.J., Martin, T.W., Olson, K.F.: Testing theoretical hypotheses: a PRE statistic. Soc. Forces 53, 190–196 (1974)
Berry, K.J., Martin, T.W., Olson, K.F.: A note on fourfold point correlation. Educ. Psychol. Meas. 34, 53–56 (1974)
Berry, K.J., Mielke, P.W., Iyer, H.K.: Factorial designs and dummy coding. Percept. Mot. Skills 87, 919–927 (1998)
Berry, K.J., Mielke, P.W., Mielke, H.W.: The Fisher–Pitman permutation test: an attractive alternative to the F test. Psychol. Rep. 90, 495–502 (2002)
Berry, K.J., Johnston, J.E., Mielke, P.W.: Exact and resampling probability values for measures associated with ordered R by C contingency tables. Psychol. Rep. 99, 231–238 (2006)
Berry, K.J., Johnston, J.E., Mielke, P.W.: An alternative measure of effect size for Cochran’s Q test for related proportions. Percept. Mot. Skills 104, 1236–1242 (2007)
Berry, K.J., Johnston, J.E., Mielke, P.W.: A Chronicle of Permutation Statistical Methods: 1920–2000 and Beyond. Springer, Cham (2014)
Bilder, C.R., Loughin, T.M.: On the first-order Rao–Scott correction of the Umesh–Loughin–Scherer statistic. Biometrics 57, 1253–1255 (2001)
Bilder, C.R., Loughin, T.M., Nettleton, D.: Multiple marginal independence-testing for pick any/c variables. Commun. Stat. Simul. Comput. 29, 1285–1316 (2000)
Biondini, M.E., Mielke, P.W., Berry, K.J.: Data-dependent permutation techniques for the analysis of ecological data. Vegetatio 75, 161–168 (1988). [The name of the journal was changed to Plant Ecology in 1997]
Blalock, H.M.: A double standard in measuring degree of association. Am. Sociol. Rev. 28, 988–989 (1963)
Blattberg, R., Sargent, T.: Regression with non-Gaussian stable disturbances. Econometrica 39, 501–510 (1971)
Borgatta, E.F.: My student, the purist: a lament. Soc. Q. 9, 29–34 (1968)
Box, G.E.P.: Science and statistics. J. Am. Stat. Assoc. 71, 791–799 (1976)
Box, J.F.: R. A. Fisher: The Life of a Scientist. Wiley, New York (1978)
Box, G.E.P.: An Accidental Statistician: The Life and Memories of George E. P. Box. Wiley, New York (2013). [Inscribed “With a little help from my friend, Judith L. Allen”]
Bradbury, I.: Analysis of variance versus randomization: a comparison. Br. J. Math. Stat. Psychol. 40, 177–187 (1987)
Bradley, J.V.: Distribution-free Statistical Tests. Prentice-Hall, Englewood Cliffs (1968)
Bradley, J.V.: A common situation conducive to bizarre distribution shapes. Am. Stat. 31, 147–150 (1977)
Brandeau, M.L., Chiu, S.S.: Parametric facility location on a tree network with an L p norm cost function. Transp. Sci. 22, 59–69 (1988)
Brennan, P.F., Hays, B.J.: The kappa statistic for establishing interrater reliability in the secondary analysis of qualitative clinical data. Res. Nurs. Heal. 15, 153–158 (1992)
Brennan, R.L., Prediger, D.J.: Coefficient kappa: some uses, misuses, and alternatives. Educ. Psychol. Meas. 41, 687–699 (1981)
Brillinger, D.R., Jones, L.V., Tukey, J.W.: The role of statistics in weather resources management. Tech. Rep. II, Weather Modification Advisory Board, United States Department of Commerce, Washington, DC (1978)
Bross, I.D.J.: Is there an increased risk? Fed. Proc. 13, 815–819 (1954)
Brown, G.W., Mood, A.M.: On median tests for linear hypotheses. In: Neyman, J. (ed.) Proceedings of the Second Berkeley Symposium on Mathematical Statistics and Probability, vol. II, pp. 159–166. University of California Press, Berkeley (1951)
Burr, E.J.: The distribution of Kendall’s score S for a pair of tied rankings. Biometrika 47, 151–171 (1960)
Burry-Stock, J.A., Laurie, D.G., Chissom, B.S.: Rater agreement indexes for performance assessment. Educ. Psychol. Meas. 56, 251–262 (1996)
Campbell, M.J., Gardner, M.J.: Calculating confidence intervals for some non-parametric analyses. Br. Med. J. 296, 1454–1456 (1988)
Capraro, R.M., Capraro, M.M.: Treatments of effect sizes and statistical significance tests in textbooks. Educ. Psychol. Meas. 62, 771–782 (2002)
Capraro, R.M., Capraro, M.M.: Exploring the APA fifth edition Publication Manual’s impact of the analytic preferences of journal editorial board members. Educ. Psychol. Meas. 63, 554–565 (2003)
Carroll, R.M., Nordholm, L.A.: Sampling characteristics of Kelley’s ε 2 and Hays’ \(\hat{\omega }^{2}\). Educ. Psychol. Meas. 35, 541–554 (1975)
Carver, R.P.: The case against statistical significance testing. Harv. Educ. Rev. 48, 378–399 (1978)
Carver, R.P.: The case against statistical significance testing, revisited. J. Exp. Educ. 61, 287–292 (1993)
Chesterton, G.K.: The Complete Father Brown Stories: “The Head of Caesar”. Star Books, Vancouver (2003)
Cochran, W.G.: The comparison of percentages in matched samples. Biometrika 37, 256–266 (1950)
Cohen, J.: A coefficient of agreement for nominal scales. Educ. Psychol. Meas. 20, 37–46 (1960)
Cohen, J.: Weighted kappa: nominal scale agreement with provision for scaled disagreement or partial credit. Psychol. Bull. 70, 213–220 (1968)
Cohen, J.: Statistical Power Analysis for the Behavioral Sciences. Academic Press, New York (1969)
Cohen, J.: Things I have learned (so far). Am. Psychol. 45, 1304–1312 (1990)
Cohen, J.: The earth is round (p < . 05). Am. Psychol. 49, 997–1003 (1994)
Cohen, J., Cohen, P.: Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences. Erlbaum, Hillsdale (1975)
Colwell, D.J., Gillett, J.R.: Spearman versus Kendall. Math. Gaz. 66, 307–309 (1982)
Conover, W.J.: Practical Nonparametric Statistics, 3rd edn. Wiley, New York (1999)
Conti, L.H., Musty, R.E.: The effects of delta-9-tetrahydrocannabinol injections to the nucleus accumbens on the locomotor activity of rats. In: Arurell, S., Dewey, W.L., Willette, R.E. (eds.) The Cannabinoids: Chemical, Pharmacologic, and Therapeutic Aspects, pp. 649–655. Academic Press, New York (1984)
Coombs, C.H.: A Theory of Data. Wiley, New York (1964)
Costner, H.L.: Criteria for measures of association. Am. Sociol. Rev. 30, 341–353 (1965)
Cramér, H.: Mathematical Methods of Statistics. Princeton University Press, Princeton (1946)
Crittenden, K.S., Montgomery, A.C.: A system of paired asymmetric measures of association for use with ordinal dependent variables. Soc. Forces 58, 1178–1194 (1980)
Cureton, E.E.: Rank-biserial correlation. Psychometrika 21, 287–290 (1956)
Cureton, E.E.: Rank-biserial correlation when ties are present. Educ. Psychol. Meas. 28, 77–79 (1968)
Curran-Everett, D.: Explorations in statistics: standard deviations and standard errors. Adv. Physiol. Educ. 32, 203–208 (2008)
Daniel, W.W.: Statistical significance versus practical significance. Sci. Educ. 61, 423–427 (1977)
Daniels, H.E.: Rank correlation and population models (with discussion). J. R. Stat. Soc. Ser. B Methodol. 12, 171–191 (1950)
Daniels, H.E.: Note on Durbin and Stuart’s formula for E(r s ). J. R. Stat. Soc. Ser. B Methodol. 13, 310 (1951)
Darwin, C.R.: The Effects of Cross and Self Fertilization in the Vegetable Kingdom. John Murray, London (1876)
David, F.N.: Review of “Rank Correlation Methods” by M. G. Kendall. Biometrika 37, 190 (1950)
de Mast, J., Akkerhuis, T., Erdmann, T.: The statistical evaluation of categorical measurements: simple scales, but treacherous complexity underneath (2014). [Originally a paper presented at the First Stu Hunter Research Conference in Heemskerk, Netherlands, March, 2013]
Decady, Y.R., Thomas, D.R.: A simple test of association for contingency tables with multiple column responses. Biometrics 56, 893–896 (2000)
Diekhoff, G.: Statistics for the Social and Behavioral Sciences: Univariate, Bivariate, Multivariate. Brown, Dubuque (1992)
Dielman, T.E.: A comparison of forecasts from least absolute and least squares regression. J. Forecast. 5, 189–195 (1986)
Dielman, T.E.: Corrections to a comparison of forecasts from least absolute and least squares regression. J. Forecast. 8, 419–420 (1989)
Dielman, T.E., Pfaffenberger, R.: Least absolute value regression: necessary sample sizes to use normal theory inference procedures. Decis. Sci. 19, 734–743 (1988)
Dielman, T.E., Rose, E.L.: Forecasting in least absolute value regression with autocorrelated errors: a small-sample study. Int. J. Forecast. 10, 539–547 (1994)
Dodd, D.H., Schultz, R.F.: Computational procedures for estimating magnitude of effects for some analysis of variance designs. Psychol. Bull. 79, 391–395 (1973)
Durbin, J., Stuart, A.: Inversions and rank correlation coefficients. J. R. Stat. Soc. Ser. B Methodol. 13, 303–309 (1951)
Dwass, M.: Modified randomization tests for nonparametric hypotheses. Ann. Math. Stat. 28, 181–187 (1957)
Dwyer, J.H.: Analysis of variance and the magnitude of effect: a general approach. Psychol. Bull. 81, 731–737 (1974)
Dyson, G.: Turing’s Cathedral: The Origins of the Digital Universe. Pantheon/Vintage, New York (2012)
Eden, T., Yates, F.: On the validity of Fisher’s z test when applied to an actual example of non-normal data. J. Agric. Sci. 23, 6–17 (1933)
Edgington, E.S.: Randomization tests. J. Psychol. 57, 445–449 (1964)
Edgington, E.S.: Statistical inference and nonrandom samples. Psychol. Bull. 66, 485–487 (1966)
Edgington, E.S.: Approximate randomization tests. J. Psychol. 72, 143–149 (1969)
Edgington, E.S.: Statistical Inference: The Distribution-Free Approach. McGraw-Hill, New York (1969)
Edgington, E.S.: Randomization Tests. Marcel Dekker, New York (1980)
Edgington, E.S., Onghena, P.: Randomization Tests, 4th edn. Chapman & Hall/CRC, Boca Raton (2007)
Edwards, D.: Exact simulation based inference: a survey, with additions. J. Stat. Comput. Simul. 22, 307–326 (1985)
Everitt, B.S.: Moments of the statistics kappa and weighted kappa. Br. J. Math. Stat. Psychol. 21, 97–103 (1968)
Ezekiel, M.J.B.: Methods of Correlation Analysis. Wiley, New York (1930)
Feinstein, A.R.: Clinical biostatistics XXIII: the role of randomization in sampling, testing, allocation, and credulous idolatry (Part 2). Clin. Pharmacol. Ther. 14, 898–915 (1973)
Feinstein, A.R.: Clinical Biostatistics. C.V. Mosby, St. Louis (1977)
Ferguson, G.A.: Statistical Analysis in Psychology and Education, 5th edn. McGraw-Hill, New York (1981)
Festinger, L.: The significance of differences between means without reference to the frequency distribution function. Psychometrika 11, 97–105 (1946)
Fidler, F., Thompson, B.: Computing correct confidence intervals for ANOVA fixed- and random-effects effect sizes. Educ. Psychol. Meas. 61, 575–604 (2001)
Fisher, R.A.: Statistical Methods for Research Workers. Oliver and Boyd, Edinburgh (1925)
Fisher, R.A.: The Design of Experiments. Oliver and Boyd, Edinburgh (1935)
Fisher, R.A.: The logic of inductive inference (with discussion). J. R. Stat. Soc. 98, 39–82 (1935)
Fisher, R.A.: Mathematics of a lady tasting tea. In: Newman, J.R. (ed.) The World of Mathematics, vol. III, section VIII, pp. 1512–1521. Simon & Schuster, New York (1956)
Fisher, R.A.: The Design of Experiments, 7th edn. Hafner, New York (1960)
Fleiss, J.L.: Estimating the magnitude of experimental effects. Psychol. Bull. 72, 273–276 (1969)
Fleiss, J.L., Cohen, J., Everitt, B.S.: Large sample standard errors of kappa and weighted kappa. Psychol. Bull. 72, 323–327 (1969)
Franklin, L.A.: Exact tables of Spearman’s footrule for n = 11(1)18 with estimate of convergence and errors for the normal approximation. Stat. Probab. Lett. 6, 399–406 (1988)
Freeman, L.C.: Elementary Applied Statistics. Wiley, New York (1965)
Frick, R.W.: Interpreting statistical testing: process and propensity, not population and random sampling. Behav. Res. Methods Instrum. Comput. 30, 527–535 (1998)
Friedman, M.: The use of ranks to avoid the assumption of normality implicit in the analysis of variance. J. Am. Stat. Assoc. 32, 675–701 (1937)
Friedman, M.: A comparison of alternative tests of significance for the problem of m rankings. Ann. Math. Stat. 11, 86–92 (1940)
Friedman, H.: Magnitude of experimental effect and a table for its rapid estimation. Psychol. Bull. 70, 245–251 (1968)
Gaebelein, J.W., Soderquist, J.A., Powers, W.A.: A note on the variance explained in the mixed analysis of variance model. Psychol. Bull. 83, 1110–1112 (1976)
Gail, M., Mantel, N.: Counting the number of r × c contingency tables with fixed margins. J. Am. Stat. Assoc. 72, 859–862 (1977)
Gardner, M.J., Altman, D.G.: Statistics with Confidence: Confidence Intervals and Statistical Guidelines. British Medical Journal, London (1989)
Geary, R.C.: Some properties of correlation and regression in a limited universe. Metron 7, 83–119 (1927)
Geary, R.C.: Testing for normality. Biometrika 34, 209–242 (1947)
Gebhard, J., Schmitz, N.: Permutation tests: a revival? I. Optimum properties. Stat. Pap. 39, 75–85 (1998)
Glass, G.V.: Note on rank-buserial correlation. Educ. Psychol. Meas. 26, 623–631 (1966)
Glass, G.V.: Primary, secondary, and meta-analysis of research. Educ. Res. 5, 3–8 (1976)
Glass, G.V.: Statistical Methods in Education and Psychology, 2nd edn. Prentice-Hall, Englewood Cliffs (1984)
Glass, G.V., Hakstian, A.R.: Measures of association in comparative experiments: their development and interpretation. Am. Educ. Res. J. 6, 403–414 (1969)
Glass, G.V., Peckham, P.D., Sanders, J.R.: Consequences of failure to meet assumptions underlying the fixed effects analysis of variance and covariance. Rev. Educ. Res. 42, 237–288 (1972)
Glass, G.V., McGraw, B., Smith, M.L.: Meta-Analysis in Social Research: Individual and Neighbourhood Reactions. Sage, Beverly Hills (1981)
Golding, S.L.: Flies in the ointment: methodological problems in the analysis of the percentage of variance due to persons and situations. Psychol. Bull. 82, 278–289 (1975)
Good, I.J.: Further comments concerning the lady tasting tea or beer: P-values and restricted randomization. J. Stat. Comput. Simul. 40, 263–267 (1992)
Good, P.I.: Permutation, Parametric and Bootstrap Tests of Hypotheses. Springer, New York (1994)
Good, P.I.: Permutation Tests: A Practical Guide to Resampling Methods for Testing Hypotheses. Springer, New York (1994)
Good, P.I.: Resampling Methods: A Practical Guide to Data Analysis. Birkhäuser, Boston (1999)
Good, P.I.: Permutation Tests: A Practical Guide to Resampling Methods for Testing Hypotheses, 2nd edn. Springer, New York (2000)
Good, P.I.: Resampling Methods: A Practical Guide to Data Analysis, 2nd edn. Birkhäuser, Boston (2001)
Good, P.I.: Extensions of the concept of exchangeability and their applications. J. Mod. Appl. Stat. Methods 1, 243–247 (2002)
Goodman, L.A., Kruskal, W.H.: Measures of association for cross classifications. J. Am. Stat. Assoc. 49, 732–764 (1954)
Goodman, L.A., Kruskal, W.H.: Measures of association for cross classifications, III: approximate sampling theory. J. Am. Stat. Assoc. 58, 310–364 (1963)
Gravetter, F.J., Wallnau, L.B.: Essentials of Statistics for the Behavioral Sciences, 8th edn. Wadsworth, Belmont (2014)
Greenhouse, S.W., Geisser, S.: On methods in the analysis of profile data. Psychometrika 24, 95–112 (1959)
Gridgeman, N.T.: The lady tasting tea, and allied topics. J. Am. Stat. Assoc. 54, 776–783 (1959)
Grier, D.A.: Statistical laboratories and the origins of computing. Chance 12, 14–20 (1999)
Grissom, R.J., Kim, J.J.: Effect Sizes for Research: A Broad Practical Approach. Lawrence Erlbaum, Mahwah (2005)
Grissom, R.J., Kim, J.J.: Effect Sizes for Research: Univariate and Multivariate Applications. Routledge, New York (2012)
Guggenmoos-Holzmann, I.: How reliable are chance-corrected measures of agreement? Stat. Med 12, 2191–2205 (1993)
Guggenmoos-Holzmann, I.: Comment on “Modeling covariate effects in observer agreement studies: the case of nominal scale agreement” by P. Graham. Stat. Med. 14, 2285–2286 (1995)
Guilford, J.P.: Fundamental Statistics in Psychology and Education. McGraw-Hill, New York (1950)
Hald, A.: History of Probability and Statistics and Their Applications Before 1750. Wiley, New York (1990)
Hald, A.: A History of Mathematical Statistics from 1750 to 1930. Wiley, New York (1998)
Haldane, J.B.S., Smith, C.A.B.: A simple exact test for birth-order effect. Ann. Eugen. 14, 117–124 (1948)
Hall, N.S.: R. A. Fisher and his advocacy of randomization. J. Hist. Biol. 40, 295–325 (2007)
Hanley, J.A.: Standard error of the kappa statistic. Psychol. Bull. 102, 315–321 (1987)
Harding, E.F.: An efficient, minimal-storage procedure for calculating the Mann–Whitney U, generalized U and similar distributions. J. R. Stat. Soc.: Ser. C: Appl. Stat. 33, 1–6 (1984)
Hayes, A.F.: Permutation test is not distribution-free: testing H 0: ρ = 0. Psychol. Methods 1, 184–198 (1996)
Hays, W.L.: Statistics. Holt, Rinehart and Winston, New York (1963)
Hedges, L.V.: Estimation of effect size from a series of independent experiments. Psychol. Bull. 92, 490–499 (1982)
Heiser, W.J.: Geometric representation of association between categories. Psychometrika 69, 513–545 (2004)
Hellman, M.: A study of some etiological factors of malocclusion. Dent. Cosmos 56, 1017–1032 (1914)
Hemelrijk, J.: Note on Wilcoxon’s two-sample test when ties are present. Ann. Math. Stat. 23, 133–135 (1952)
Henson, R.K., Smith, A.D.: State of the art in statistical significance and effect size reporting: a review of the APA task force report and current trends. J. Res. Dev. Educ. 33, 285–296 (2000)
Hess, B., Olejnik, S., Huberty, C.J.: The efficacy of two improvement-over-chance effect sizes for two-group univariate comparisons. Educ. Psychol. Meas. 61, 909–936 (2001)
Higgins, J.J.: Introduction to Modern Nonparametric Tests. Brooks/Cole, Pacific Grove (2004)
Hitchcock, D.B.: Yates and contingency tables: 75 years later. Electron. J. Hist. Probab. Stat. 5, 1–14 (2009)
Hodges, J.L., Lehmann, E.L.: Rank methods for combination of independent experiments in analysis of variance. Ann. Math. Stat. 33, 482–497 (1962)
Hodges, J.L., Lehmann, E.L.: Estimates of location based on rank tests. Ann. Math. Stat. 34, 598–611 (1963)
Hope, A.C.A.: A simplified Monte Carlo significance test procedure. J. R. Stat. Soc. Ser. B Methodol. 30, 582–598 (1968)
Hotelling, H.: The generalization of student’s ratio. Ann. Math. Stat. 2, 360–378 (1931)
Hotelling, H.: A generalized T test and measure of multivariate dispersion. In: Neyman, J. (ed.) Proceedings of the Second Berkeley Symposium on Mathematical Statistics and Probability, vol. II, pp. 23–41. University of California Press, Berkeley (1951)
Hotelling, H., Pabst, M.R.: Rank correlation and tests of significance involving no assumption of normality. Ann. Math. Stat. 7, 29–43 (1936)
Howell, D.C.: Statistical Methods for Psychology, 6th edn. Wadsworth, Belmont (2007)
Howell, D.C.: Statistical Methods for Psychology, 8th edn. Wadsworth, Belmont (2013)
Hubbard, R.: Alphabet soup: Blurring the distinctions between p’s and α’s in psychological research. Theor. Psychol. 14, 295–327 (2004)
Hubert, L.J.: A note on Freeman’s measure of association for relating an ordered to an unordered factor. Psychometrika 39, 517–520 (1974)
Hunter, A.A.: On the validity of measures of association: the nominal-nominal two-by-two case. Am. J. Sociol. 79, 99–109 (1973)
Hutchinson, T.P.: Kappa muddles together two sources of disagreement: Tetrachoric correlation is preferable. Res. Nurs. Health 16, 313–315 (1993)
Huynh, H., Feldt, L.S.: Conditions under which mean square ratios in repeated measurements designs have exact F distributions. J. Am. Stat. Assoc. 65, 1582–1589 (1970)
Irwin, J.O.: Tests of significance for differences between percentages based on small numbers. Metron 12, 83–94 (1935)
Isaacson, W.: The Innovators. Simon & Schuster, New York (2014)
Jockel, K.H.: Finite sample properties and asymptotic efficiency of Monte Carlo tests. J. Stat. Comput. Simul. 14, 336–347 (1986)
Johnston, J.E., Berry, K.J., Mielke, P.W.: A measure of effect size for experimental designs with heterogeneous variances. Percept. Mot. Skills 98, 3–18 (2004)
Johnston, J.E., Berry, K.J., Mielke, P.W.: Permutation tests: precision in estimating probability values. Percept. Mot. Skills 105, 915–920 (2007)
Jonckheere, A.R.: A distribution-free k-sample test against ordered alternatives. Biometrika 41, 133–145 (1954)
Kahaner, D., Moler, C., Nash, S.: Numerical Methods and Software. Prentice-Hall, Englewood Cliffs (1988)
Kaufman, E.H., Taylor, G.D., Mielke, P.W., Berry, K.J.: An algorithm and FORTRAN program for multivariate LAD (ℓ 1 of ℓ 2) regression. Computing 68, 275–287 (2002)
Keller-McNulty, S., Higgins, J.J.: Effect of tail weight and outliers and power and type-I error of robust permutation tests for location. Commun. Stat. Simul. Comput. 16, 17–35 (1987)
Kelley, T.L.: An unbiased correlation ratio measure. Proc. Natl. Acad. Sci. 21, 554–559 (1935)
Kempthorne, O.: The Design and Analysis of Experiments. Wiley, New York (1952)
Kempthorne, O.: The randomization theory of experimental inference. J. Am. Stat. Assoc. 50, 946–967 (1955)
Kempthorne, O.: Some aspects of experimental inference. J. Am. Stat. Assoc. 61, 11–34 (1966)
Kempthorne, O.: Why randomize? J. Stat. Plan. Inference 1, 1–25 (1977)
Kendall, M.G.: A new measure of rank correlation. Biometrika 30, 81–93 (1938)
Kendall, M.G.: The treatment of ties in ranking problems. Biometrika 33, 239–251 (1945)
Kendall, M.G.: Rank Correlation Methods. Griffin, London (1948)
Kendall, M.G.: Rank Correlation Methods, 3rd edn. Griffin, London (1962)
Kendall, M.G., Babington Smith, B.: The problem of m rankings. Ann. Math. Stat. 10, 275–287 (1939)
Kendall, M.G., Babington Smith, B.: On the method of paired comparisons. Biometrika 31, 324–345 (1940)
Kendall, M.G., Kendall, S.F.H., Babington Smith, B.: The distribution of Spearman’s coefficient of rank correlation in a universe in which all rankings occur an equal number of times. Biometrika 30, 251–273 (1939)
Kennedy, P.E.: Randomization tests in econometrics. J. Bus. Econ. Stat. 13, 85–94 (1995)
Kenny, D.A.: Statistics for the Social and Behavioral Sciences. Little Brown, Boston (1987)
Keppel, G.: Design and Analysis: A Researcher’s Handbook, 2nd edn. Prentice-Hall, Englewood Cliffs (1982)
Keppel, G., Zedeck, S.: Data Analysis for Research Designs: Analysis of Variance and Multiple Regression/Correlation Approaches. Freeman, New York (1989)
Kim, M.J., Nelson, C.R., Startz, R.: Mean revision in stock prices? a reappraisal of the empirical evidence. Rev. Econ. Stud. 58, 515–528 (1991)
Kingman, J.F.C.: Uses of exchangeability. Ann. Probab. 6, 183–197 (1978). [Abraham Wald memorial lecture delivered in Aug 1977 in Seattle, Washington]
Kirk, R.E.: Experimental Design: Procedures for the Behavioral Sciences. Brooks/Cole, Belmont (1968)
Kirk, R.E.: Practical significance: a concept whose time has come. Educ. Psychol. Meas. 56, 746–759 (1996)
Kirk, R.E.: Effect magnitude: a different focus. J. Stat. Plan. Inference 137, 1634–1646 (2006). [Keynote address delivered at the 2003 International Conference on Statistics, Combinatorics, and Related Areas, held at the University of Southern Maine]
Kraft, C.A., van Eeden, C.: A Nonparametric Introduction to Statistics. Macmillan, New York (1968)
Krause, E.F.: Taxicab Geometry. Addison-Wesley, Menlo Park (1975)
Krippendorff, K.: Bivariate agreement coefficients for reliability of data. In: Borgatta, E.G. (ed.) Sociological Methodology, pp. 139–150. Jossey-Bass, San Francisco (1970)
Kruskal, W.H.: Historical notes on the Wilcoxon unpaired two-sample test. J. Am. Stat. Assoc. 52, 356–360 (1957)
Kruskal, W.H., Wallis, W.A.: Use of ranks in one-criterion variance analysis. J. Am. Stat. Assoc. 47, 583–621 (1952). [Erratum: J. Am. Stat. Assoc. 48, 907–911 (1953)]
Lachin, J.M.: Statistical properties of randomization in clinical trials. Control. Clin. Trials 9, 289–311 (1988)
LaFleur, B.J., Greevy, R.A.: Introduction to permutation and resampling-based hypothesis tests. J. Clin. Child Adolesc. 38, 286–294 (2009)
Lance, C.E.: More statistical and methodological myths and urban legends. Organ. Res. Methods 14, 279–286 (2011)
Lange, J.: Crime as Destiny: A Study of Criminal Twins. Allen & Unwin, London (1931). [Translated by C. Haldane]
Larson, S.C.: The shrinkage of the coefficient of multiple correlation. J. Educ. Psychol. 22, 45–55 (1931)
Larson, R.C., Sadiq, G.: Facility locations with the Manhattan metric in the presence of barriers to travel. Oper. Res. 31, 652–669 (1983)
Lawley, D.N.: A generalization of Fisher’s z test. Biometrika 30, 180–187 (1938)
Lawley, D.N.: Corrections to “A generalization of Fisher’s z test”. Biometrika 30, 467–469 (1939)
Leach, C.: Introduction to Statistics: A Nonparametric Approach for the Social Sciences. Wiley, New York (1979)
Lehmann, E.L.: Parametrics vs. nonparametrics: two alternative methodologies. J. Nonparametr. Stat. 21, 397–405 (2009)
Lehmann, E.L.: Fisher, Neyman, and the Creation of Classical Statistics. Springer, New York (2011)
Lehmann, E.L., Stein, C.M.: On the theory of some non-parametric hypotheses. Ann. Math. Stat. 20, 28–45 (1949)
Levine, J.H.: Joint-space analysis of “pick-any” data: analysis of choices from an unconstrained set of alternatives. Psychometrika 44, 85–92 (1979)
Levine, T.R., Hullett, C.R.: Eta squared, partial eta squared, and misreporting of effect size in communication research. Hum. Commun. Res. 28, 612–625 (2002)
Levine, T.R., Weber, R., Hullett, C.R., Park, H.S., Massi Lindsey, L.L.: A critical assessment of null hypothesis significance testing in quantitative communication research. Hum. Commun. Res. 34, 171–187 (2008)
Levine, T.R., Weber, R., Park, H.S., Hullett, C.R.: A communication researchers’ guide to null hypothesis significance testing and alternatives. Hum. Commun. Res. 34, 188–209 (2008)
Light, R.J.: Measures of response agreement for qualitative data: some generalizations and alternatives. Psychol. Bull. 76, 365–377 (1971)
Light, R.J., Margolin, B.H.: An analysis of variance for categorical data. J. Am. Stat. Assoc. 66, 534–544 (1971)
Linn, R.L., Baker, E.L., Dunbar, S.B.: Complex performance-based assessment: expectations and validation criterion. Educ. Res. 20, 15–21 (1991)
Loether, H.J., McTavish, D.G.: Descriptive and Inferential Statistics: An Introduction, 4th edn. Allyn and Bacon, Boston (1993)
Loughin, T.M., Scherer, P.N.: Testing for association in contingency tables with multiple column responses. Biometrics 54, 630–637 (1998)
Ludbrook, J.: Advantages of permutation (randomization) tests in clinical and experimental pharmacology and physiology. Clin. Exp. Pharmacol. Physiol. 21, 673–686 (1994)
Ludbrook, J.: Issues in biomedical statistics: comparing means by computer-intensive tests. Aust. N. Z. J. Surg. 65, 812–819 (1995)
Ludbrook, J.: The Wilcoxon–Mann–Whitney test condemned. Br. J. Surg. 83, 136–137 (1996)
Ludbrook, J.: Statistical techniques for comparing measures and methods of measurement: a critical review. Clin. Exp. Pharmacol. Physiol. 29, 527–536 (2002)
Ludbrook, J.: Outlying observations and missing values: how should they be handled? Clin. Exp. Pharmacol. Physiol. 35, 670–678 (2008)
Ludbrook, J., Dudley, H.A.F.: Issues in biomedical statistics: analyzing 2 × 2 tables of frequencies. Aust. N. Z. J. Surg. 64, 780–787 (1994)
Ludbrook, J., Dudley, H.A.F.: Issues in biomedical statistics: statistical inference. Aust. N. Z. J. Surg. 64, 630–636 (1994)
Ludbrook, J., Dudley, H.A.F.: Why permutation tests are superior to t and F tests in biomedical research. Am. Stat. 52, 127–132 (1998)
Ludbrook, J., Dudley, H.A.F.: Discussion of “Why permutation tests are superior to t and F tests in biomedical research” by J. Ludbrook and H.A.F. Dudley. Am. Stat. 54, 87 (2000)
Lunneborg, C.E.: Data Analysis by Resampling: Concepts and Applications. Duxbury, Pacific Grove (2000)
Maclure, M., Willett, W.C.: Misinterpretation and misuse of the kappa statistic. Am. J. Epidemiol. 126, 161–169 (1987)
Manly, B.F.J.: Randomization and Monte Carlo Methods in Biology. Chapman & Hall, London (1991)
Manly, B.F.J.: Randomization and Monte Carlo Methods in Biology, 2nd edn. Chapman & Hall, London (1997)
Manly, B.F.J.: Randomization, Bootstrap and Monte Carlo Methods in Biology, 3rd edn. Chapman & Hall/CRC, Boca Raton (2007)
Manly, B.F.J., Francis, R.I.C.: Analysis of variance by randomization when variances are unequal. Aust. N. Z. J. Stat. 41, 411–429 (1999)
Mann, H.B., Whitney, D.R.: On a test of whether one of two random variables is stochastically larger than the other. Ann. Math. Stat. 18, 50–60 (1947)
Margolin, B.H., Light, R.J.: An analysis of variance for categorical data, II: small sample comparisons with chi square and other competitors. J. Am. Stat. Assoc. 69, 755–764 (1974)
Mathew, T., Nordström, K.: Least squares and least absolute deviation procedures in approximately linear models. Stat. Probab. Lett. 16, 153–158 (1993)
Maxim, P.S.: Quantitative Research Methods in the Social Sciences. Oxford, New York (1999)
Maxwell, S.E., Camp, C.J., Arvey, R.D.: Measures of strength of association: a comparative examination. J. Appl. Psychol. 66, 525–534 (1981)
May, R.B., Hunter, M.A.: Some advantages of permutation tests. Can. Psychol. 34, 401–407 (1993)
May, S.M.: Modelling observer agreement: an alternative to kappa. J. Clin. Epidemiol. 47, 1315–1324 (1994)
McCarthy, M.D.: On the application of the z-test to randomized blocks. Ann. Math. Stat. 10, 337–359 (1939)
McGrath, R.E., Meyer, G.J.: When effect sizes disagree: the case of r and d. Psychol. Methods 11, 386–401 (2006)
McHugh, R.B., Mielke, P.W.: Negative variance estimates and statistical dependence in nested sampling. J. Am. Stat. Assoc. 63, 1000–1003 (1968)
McLean, J.E., Ernest, J.M.: The role of statistical significance testing in educational research. J. Health Soc. Behav. 5, 15–22 (1998)
McNemar, Q.: Note on the sampling error of the differences between correlated proportions and percentages. Psychometrika 12, 153–157 (1947)
McQueen, G.: Long-horizon mean-reverting stock priced revisited. J. Financ. Quant. Anal. 27, 1–17 (1992)
Mehta, C.R., Patel, N.R.: Algorithm 643: FEXACT. A FORTRAN subroutine for Fisher’s exact test on unordered r × c contingency tables. ACM Trans. Math. Softw. 12, 154–161 (1986)
Mehta, C.R., Patel, N.R.: A hybrid algorithm for Fisher’s exact test in unordered r × c contingency tables. Commun. Stat. Theory Methods 15, 387–403 (1986)
Mehta, C.R., Patel, N.R., Gray, R.: On computing an exact confidence interval for the common odds ratio in several 2 × 2 contingency tables. J. Am. Stat. Assoc. 80, 969–973 (1985)
Metropolis, N., Ulam, S.: The Monte Carlo method. J. Am. Stat. Assoc. 44, 335–341 (1949)
Meyer, G.J.: Assessing reliability: critical corrections for a critical examination of the Rorschach comprehensive system. Psychol. Assess. 9, 480–489 (1997)
Micceri, T.: The unicorn, the normal curve, and other improbable creatures. Psychol. Bull. 105, 156–166 (1989)
Mielke, P.W.: Asymptotic behavior of two-sample tests based on powers of ranks for detecting scale and location alternatives. J. Am. Stat. Assoc. 67, 850–854 (1972)
Mielke, P.W.: Squared rank test appropriate to weather modification cross-over design. Technometrics 16, 13–16 (1974)
Mielke, P.W.: Convenient beta distribution likelihood techniques for describing and comparing meteorological data. J. Appl. Meterol. 14, 985–990 (1975)
Mielke, P.W.: Meteorological applications of permutation techniques based on distance functions. In: Krishnaiah, P.R., Sen, P.K. (eds.) Handbook of Statistics, vol. IV, pp. 813–830. North-Holland, Amsterdam (1984)
Mielke, P.W.: Geometric concerns pertaining to applications of statistical tests in the atmospheric sciences. J. Atmos. Sci. 42, 1209–1212 (1985)
Mielke, P.W.: Non-metric statistical analyses: some metric alternatives. J. Stat. Plan Inference 13, 377–387 (1986)
Mielke, P.W.: The application of multivariate permutation methods based on distance functions in the earth sciences. Earth Sci. Rev. 31, 55–71 (1991)
Mielke, P.W., Berry, K.J.: An extended class of permutation techniques for matched pairs. Commun. Stat. Theory Methods 11, 1197–1207 (1982)
Mielke, P.W., Berry, K.J.: Asymptotic clarifications, generalizations, and concerns regarding an extended class of matched pairs tests based on powers of ranks. Psychometrika 48, 483–485 (1983)
Mielke, P.W., Berry, K.J.: Cumulant methods for analyzing independence of r-way contingency tables and goodness-of-fit frequency data. Biometrika 75, 790–793 (1988)
Mielke, P.W., Berry, K.J.: Permutation tests for common locations among samples with unequal variances. J. Educ. Behav. Stat. 19, 217–236 (1994)
Mielke, P.W., Berry, K.J.: Nonasymptotic inferences based on Cochran’s Q test. Percept. Mot. Skill 81, 319–322 (1995)
Mielke, P.W., Berry, K.J.: Permutation-based multivariate regression analysis: the case for least sum of absolute deviations regression. Ann. Oper. Res. 74, 259–268 (1997)
Mielke, P.W., Berry, K.J.: Permutation covariate analyses of residuals based on Euclidean distance. Psychol. Rep. 81, 795–802 (1997)
Mielke, P.W., Berry, K.J.: Euclidean distance based permutation methods in atmospheric science. Data Min. Knowl. Disc. 4, 7–27 (2000)
Mielke, P.W., Berry, K.J.: Data-dependent analyses in psychological research. Psychol. Rep. 91, 1225–1234 (2002)
Mielke, P.W., Berry, K.J.: Permutation Methods: A Distance Function Approach, 2nd edn. Springer, New York (2007)
Mielke, P.W., Berry, K.J.: A note on Cohen’s weighted kappa coefficient of agreement with linear weights. Stat. Methodol. 6, 439–446 (2009)
Mielke, P.W., Iyer, H.K.: Permutation techniques for analyzing multi-response data from randomized block experiments. Commun. Stat. Theory Methods 11, 1427–1437 (1982)
Mielke, P.W., Berry, K.J., Johnson, E.S.: Multi-response permutation procedures for a priori classifications. Commun. Stat. Theory Methods 5, 1409–1424 (1976)
Mielke, P.W., Berry, K.J., Brier, G.W.: Application of multi-response permutation procedures for examining seasonal changes in monthly mean sea-level pressure patterns. Mon. Weather Rev. 109, 120–126 (1981)
Mielke, H.W., Anderson, J.C., Berry, K.J., Mielke, P.W., Chaney, R.L., Leech, M.: Lead concentrations in inner-city soils as a factor in the child lead problem. Am. J. Public Health 73, 1366–1369 (1983)
Mielke, P.W., Berry, K.J., Landsea, C.W., Gray, W.M.: Artificial skill and validation in meteorological forecasting. Weather Forecast. 11, 153–169 (1996)
Mielke, P.W., Berry, K.J., Neidt, C.O.: A permutation test for multivariate matched-pairs analyses: comparisons with Hotelling’s multivariate matched-pairs T 2 test. Psychol. Rep. 78, 1003–1008 (1996)
Mielke, P.W., Berry, K.J., Johnston, J.E.: A FORTRAN program for computing the exact variance of weighted kappa. Percept. Mot. Skill 101, 468–472 (2005)
Mielke, P.W., Berry, K.J., Johnston, J.E.: The exact variance of weighted kappa with multiple raters. Psychol. Rep. 101, 655–660 (2007)
Mielke, P.W., Berry, K.J., Johnston, J.E.: Resampling programs for multiway contingency tables with fixed marginal frequency totals. Psychol. Rep. 101, 18–24 (2007)
Mielke, P.W., Berry, K.J., Johnston, J.E.: Resampling probability values for weighted kappa with multiple raters. Psychol. Rep. 102, 606–613 (2008)
Mielke, P.W., Berry, K.J., Johnston, J.E.: Robustness without rank order statistics. J. Appl. Stat. 38, 207–214 (2011)
Minkowski, H.: Über die positiven quadratishen formen und über kettenbruchähnliche algorithmen. Crelle’s J (J. Reine Angew. Math.) 107, 278–297 (1891). [Also available in H. Minkowski, Gesammelte Abhandlungen, vol. 1, AMS Chelsea, New York, 1967]
Mitchell, C., Hartmann, D.P.: A cautionary note on the use of omega squared to evaluate the effectiveness of behavioral treatments. Behav. Assess. 3, 93–100 (1981)
Mood, A.M.: On the asymptotic efficiency of certain nonparametric two-sample tests. Ann. Math. Stat. 25, 514–522 (1954)
Moses, L.E.: Statistical theory and research design. Ann. Rev. Psychol. 7, 233–258 (1956)
Murphy, K.R., Cleveland, J.: Understanding Performance Appraisal: Social, Organizational, and Goal-Based Perspectives. Sage, Thousand Oaks (1995)
Myers, J.L., Well, A.D.: Research Design and Statistical Analysis. HarperCollins, New York (1991)
Nanda, D.N.: Distribution of the sum of roots of a determinantal equation. Ann. Math. Stat. 21, 432–439 (1950)
Neave, H.R., Worthington, P.L.: Distribution-Free Tests. Unwin Hyman, London (1988)
Newson, R.: Parameters behind “nonparametric” statistics: Kendall’s tau, Somers’ D and median differences. Stata J. 2, 45–64 (2002)
Neyman, J., Pearson, E.S.: On the use and interpretation of certain test criteria for purposes of statistical inference: part I. Biometrika 20A, 175–240 (1928)
Neyman, J., Pearson, E.S.: On the use and interpretation of certain test criteria for purposes of statistical inference: part II. Biometrika 20A, 263–294 (1928)
Nix, T.W., Barnette, J.J.: The data analysis dilemma: Ban or abandon. A review of null hypothesis significance testing. Res. Schools 5, 3–14 (1998)
Nix, T.W., Barnette, J.J.: A review of hypothesis testing revisited: Rejoinder to Thompson, Knapp, and Levin. Res. Schools 5, 55–57 (1998)
O’Boyle, Jr., E., Aguinis, H.: The best and the rest: revisiting the norm of normality of individual performance. Percept. Psychophys. 65, 79–119 (2012)
Okamoto, D.: Letter to the editor: does it work for coffee? Significance 10, 45–46 (2013)
Olds, E.G.: Distribution of sums of squares of rank differences for small numbers of individuals. Ann. Math. Stat. 9, 133–148 (1938)
Olejnik, S., Algina, J.: Measures of effect size for comparative studies: applications, interpretations, and limitations. Contemp. Educ. Psychol. 25, 241–286 (2000)
Olson, C.L.: On choosing a test statistic in multivariate analysis of variance. Psychol. Bull. 83, 579–586 (1976)
Olson, C.L.: Practical considerations in choosing a MANOVA test statistic: a rejoinder to Stevens. Psychol. Bull. 86, 1350–1352 (1979)
Osgood, C.E., Suci, G., Tannenbaum, P.: The Measurement of Meaning. University of Illinois Press, Urbana (1957)
Overall, J.E., Spiegel, D.K.: Concerning least squares analysis of experimental data. Psychol. Bull. 72, 311–322 (1969)
Pagano, R.R.: Understanding Statistics in the Behavioral Sciences, 6th edn. Wadsworth, Pacific Grove (2001)
Pearson, K.: Contributions to the mathematical theory of evolution. Proc. R. Soc. Lond. 54, 329–333 (1893)
Pearson, K.: Contributions to the mathematical theory of evolution, II. Skew variation in homogeneous material. Philos. Trans. R. Soc. Lond. A 186, 343–414 (1895)
Pearson, K.: Mathematical contributions to the theory of evolution, XIII. On the theory of contingency and its relation to association and normal correlation. In: Drapers’ Company Research Memoirs, Biometric Series I, pp. 1–35. Cambridge University Press, Cambridge (1904)
Pearson, E.S.: Untitled. Nature 123, 866–867 (1929). [Review by E.S. Pearson of the second edition of R.A. Fisher’s Statistical Methods for Research Workers]
Pearson, K., Heron, D.: On theories of association. Biometrika 9, 159–315 (1913)
Pfaffenberger, R., Dinkel, J.: Absolute deviations curve-fitting: an alternative to least squares. In: David, H.A. (ed.) Contributions to Survey Sampling and Applied Statistics, pp. 279–294. Academic Press, New York (1978)
Picard, R.: Randomization and design: II. In: Feinberg, S.E., Hinkley, D.V. (eds.) R. A. Fisher: An Appreciation, pp. 46–58. Springer, Heidelberg (1980)
Pillai, K.C.S.: Some new test criteria in multivariate analysis. Ann. Math. Stat. 26, 117–121 (1955)
Pitman, E.J.G.: Significance tests which may be applied to samples from any populations. Suppl. J. R. Stat. Soc. 4, 119–130 (1937)
Pitman, E.J.G.: Significance tests which may be applied to samples from any populations: II. The correlation coefficient test. Suppl. J. R. Stat. Soc. 4, 225–232 (1937)
Pitman, E.J.G.: Significance tests which may be applied to samples from any populations: III. The analysis of variance test. Biometrika 29, 322–335 (1938)
Randles, R.H., Wolfe, D.A.: Introduction to the Theory of Nonparametric Statistics. Wiley, New York (1979)
Raveh, A.: On measures of monotone association. Am. Stat. 40, 117–123 (1986)
Reinhart, A.: Statistics Done Wrong: The Woefully Complete Guide. No Starch Press, San Francisco (2015)
Rice, J., White, J.: Norms for smoothing and estimation. SIAM Rev. 6, 243–256 (1964)
Ricketts, C., Berry, J.S.: Teaching statistics through resampling. Teach. Stat. 16, 41–44 (1994)
Roberts, J.K., Henson, R.K.: Correcting for bias in estimating effect sizes. Educ. Psychol. Meas. 62, 241–253 (2002)
Robinson, W.S.: Ecological correlations and the behavior of individuals. Am. Soc. Rev. 15, 351–357 (1950). [Reprinted in Int. J. Epidemiol. 38, 337–341 (2009)]
Robinson, W.S.: The statistical measurement of agreement. Am. Sociol. Rev. 22, 17–25 (1957)
Robinson, W.S.: The geometric interpretation of agreement. Am. Sociol. Rev. 24, 338–345 (1959)
Rosenberg, B., Carlson, D.: A simple approximation of the sampling distribution of least absolute residuals regression estimates. Commun. Stat. Simul. Comput. 6, 421–438 (1977)
Rosenthal, R., Rosnow, R.L., Rubin, D.B.: Contrasts and Effect Sizes in Behavioral Research: A Correlational Approach. Cambridge University Press, Cambridge (2000)
Rouanet, H., Lépine, D.: Comparison between treatments in a repeated measures design: ANOVA and multivariate methods. Br. J. Math. Stat. Psychol. 23, 147–164 (1970)
Rousseeuw, P.J.: Least median of squares regression. J. Am. Stat. Assoc. 79, 421–438 (1984)
Routledge, R.D.: Resolving the conflict over Fisher’s exact test. Can. J. Stat. 20, 201–209 (1992)
Roy, S.N.: On a heuristic method of test construction and its use in multivariate analysis. Ann. Math. Stat. 24, 220–238 (1953)
Roy, S.N.: Some Aspects of Multivariate Analysis. Wiley, New York (1957)
Saal, F.E., Downey, R.G., Lahey, M.A.: Rating the ratings: assessing the quality of rating data. Psychol. Bull. 88, 413–428 (1980)
Salama, I.A., Quade, D.: A note on Spearman’s footrule. Commun. Stat. Simul. Comput. 19, 591–601 (1990)
Salsburg, D.: The Lady Tasting Tea: How Statistics Revolutionized Science in the Twentieth Century. Holt, New York (2001)
Särndal, C.E.: A comparative study of association measures. Psychometrika 39, 165–187 (1974)
Satterthwaite, F.E.: An approximate distribution of estimates of variance components. Biom. Bull. 2, 110–114 (1946)
Scheffé, H.: Statistical inference in the non-parametric case. Ann. Math. Stat. 14, 305–332 (1943)
Scheffé, H.: The Analysis of Variance. Wiley, New York (1959)
Schmidt, F.L., Johnson, R.H.: Effect of race on peer ratings in an industrial situation. J. Appl. Psychol. 57, 237–241 (1973)
Schuster, C.: A note on the interpretation of weighted kappa and its relations to other rater agreement statistics for metric scales. Educ. Psychol. Meas. 64, 243–253 (2004)
Scott, W.A.: Reliability of content analysis: the case of nominal scale coding. Public Opin. Q. 19, 321–325 (1955)
Senn, S.: Fisher’s game with the devil. Stat. Med. 13, 217–230 (1994). [Publication of a paper presented at the Statisticians in the Pharmaceutical Industry (PSI) annual conference held in Sept 1991 in Bristol, England]
Senn, S.: Tea for three: of infusions and inferences and milk in first. Significance 9, 30–33 (2012)
Senn, S.: Response to “Tea break” by S. Springate. Significance 10, 46 (2013)
Sheynin, O.B.: R. J. Boscovich’s work on probability. Arch. Hist. Exact Sci. 9, 306–324 (1973)
Shrout, P.E., Fleiss, J.L.: Intraclass correlations: uses in assessing rater relaibility. Psychol. Bull. 86, 420–428 (1979)
Shrout, P.E., Spitzer, R.L., Fleiss, J.L.: Quantification of agreement in psychiatric diagnosis revisited. Arch. Gen. Psychiatry 44, 172–177 (1987)
Siegel, S., Castellan, N.J.: Nonparametric Statistics for the Behavioral Sciences, 2nd edn. McGraw-Hill, New York (1988)
Siegel, S., Tukey, J.W.: A nonparametric sum of ranks procedure for relative spread in unpaired samples. J. Am. Stat. Assoc. 55, 429–445 (1960). [Corrigendum: J. Am. Stat. Assoc. 56, 1005 (1961)]
Siegfried, T.: Odds are, it’s wrong. Sci. News 177, 26–29 (2010)
Snedecor, G.W.: Calculation and Interpretation of Analysis of Variance and Covariance. Collegiate Press, Ames (1934)
Snyder, P., Lawson, S.: Evaluating results using corrected and uncorrected effect size estimates. J. Exp. Educ. 61, 334–349 (1993)
Somers, R.H.: A new asymmetric measure of association for ordinal variables. Am. Sociol. Rev. 27, 799–811 (1962)
Spearman, C.E.: The proof and measurement of association between two things. Am. J. Psychol. 15, 72–101 (1904)
Spearman, C.E.: ‘Footrule’ for measuring correlation. Br. J. Psychol. 2, 89–108 (1906)
Spitznagel, E.L., Helzer, J.E.: A proposed solution to the base rate problem in the kappa statistic. Arch. Gen. Psychiatry 42, 725–728 (1985)
Springate, S.: Tea break. Significance 10, 45–46 (2013)
Stark, R., Roberts, I.: Contemporary Social Research Methods. Micro-Case, Bellevue (1996)
Stevens, J.P.: Applied Multivariate Statistics for the Social Sciences. Erlbaum, Hillsdale (1986)
Stevens, J.P.: Intermediate Statistics: A Modern Approach. Erlbaum, Hillsdale (1990)
Still, A.W., White, A.P.: The approximate randomization test as an alternative to the F test in analysis of variance. Br. J. Math. Stat. Psychol. 34, 243–252 (1981)
Stuart, A.: The estimation and comparison of strengths of association in contingency tables. Biometrika 40, 105–110 (1953)
“Student”: The probable error of a mean. Biometrika 6, 1–25 (1908). [“Student” is a nom de plume for William Sealy Gosset]
Susskind, E.C., Howland, E.W.: Measuring effect magnitude in repeated measures ANOVA designs: implications for gerontological research. J. Gerontol. 35, 867–876 (1980)
Tabachnick, B.G., Fidell, L.S.: Using Multivariate Statistics, 5th edn. Pearson, Boston (2007)
Taha, M.A.H.: Rank test for scale parameter for asymmetrical one-sided distributions. Publ. Inst. Stat. Univ. Paris 13, 169–180 (1964)
Taylor, L.D.: Estimation by minimizing the sum of absolute errors. In: Zarembka, P. (ed.) Frontiers in Econometrics, pp. 169–190. Academic Press, New York (1974)
Tedin, O.: The influence of systematic plot arrangements upon the estimate of error in field experiments. J. Agric. Sci. 21, 191–208 (1931)
Thompson, D.W.: On Growth and Form: The Complete Revised Edition. Dover, New York (1992)
Thompson, W.L.: 402 citations questioning the indiscriminate use of null hypothesis significance tests in observational studies. http://www.warnercnr.colostate.edu/~anderson/thompson1.html (2001). Accessed 18 June 2015
Thompson, W.L.: Problems with the hypothesis testing approach. http://www.warnercnr.colostate.edu/~gwhite/fw663/testing.pdf (2001). Accessed 18 June 2015
Thompson, W.D., Walter, S.D.: A reappraisal of the kappa coefficient. J. Clin. Epidemiol. 41, 949–958 (1988)
Trafimow, D.: Editorial. Basic Appl. Soc. Psychol. 36, 1–2 (2014)
Trafimow, D., Marks, M.: Editorial. Basic Appl. Soc. Psychol. 37, 1–2 (2015)
Tschuprov, A.A.: Principles of the Mathematical Theory of Correlation. Hodge, London (1939). [Translated by M. Kantorowitsch]
Tukey, J.W.: Data analysis and behavioral science (1962). [Unpublished manuscript]
Tukey, J.W.: The future of data analysis. Ann. Math. Stat. 33, 1–67 (1962)
Tukey, J.W.: Randomization and re-randomization: the wave of the past in the future. In: Statistics in the Pharmaceutical Industry: Past, Present and Future. Philadelphia Chapter of the American Statistical Association (1988). [Presented at a Symposium in Honor of Joseph L. Ciminera held in June 1988 at Philadelphia, Pennsylvania]
Umesh, U.N.: Predicting nominal variable relationships with multiple response. J. Forecast. 14, 585–596 (1995)
Umesh, U.N., Peterson, R.A., Sauber, M.H.: Interjudge agreement and the maximum value of kappa. Educ. Psychol. Meas. 49, 835–850 (1989)
Ury, H.K., Kleinecke, D.C.: Tables of the distribution of Spearman’s footrule. J. R. Stat. Soc.: Ser. C: Appl. Stat. 28, 271–275 (1979)
van der Reyden, D.: A simple statistical significance test. Rhod. Agric. J. 49, 96–104 (1952)
Vanbelle, S., Albert, A.: A note on the linearly weighted kappa coefficient for ordinal scales. Stat. Methodol. 6, 157–163 (2008)
Vaughan, G.M., Corballis, M.C.: Beyond tests of significance: estimating strength of effects in selected ANOVA designs. Psychol. Bull. 79, 391–395 (1969)
von Eye, A., von Eye, M.: On the marginal dependency of Cohen’s κ. Eur. Pychol. 13, 305–315 (2008)
Wald, A., Wolfowitz, J.: An exact test for randomness in the non-parametric case based on serial correlation. Ann. Math. Stat. 14, 378–388 (1943)
Wallis, W.A.: The correlation ratio for ranked data. J. Am. Stat. Assoc. 34, 533–538 (1939)
Watnik, M.: Early computational statistics. J. Comput. Graph. Stat. 20, 811–817 (2011)
Watterson, I.G.: Nondimensional measures of climate model performance. Int. J. Climatol. 16, 379–391 (1996)
Welch, B.L.: The specification of rules for rejecting too variable a product, with particular reference to an electric lamp problem. Suppl. J. R. Stat. Soc. 3, 29–48 (1936)
Welch, B.L.: On the z-test in randomized blocks and Latin squares. Biometrika 29, 21–52 (1937)
Welch, B.L.: The significance of the difference between two means when the population variances are unequal. Biometrika 29, 350–362 (1938)
Welch, B.L.: On the comparison of several mean values: an alternative approach. Biometrika 38, 330–336 (1951)
Welkowitz, J., Ewen, R.B., Cohen, J.: Introductory Statistics for the Behavioral Sciences, 5th edn. Harcourt Brace, Orlando (2000)
Wherry, R.J.: A new formula for predicting the shrinkage of the coefficient of multiple correlation. Ann. Math. Stat. 2, 440–457 (1931)
Whitehurst, G.J.: Interrater agreement for journal manuscript reviews. Am. Psychol. 39, 22–28 (1984)
Whitfield, J.W.: Rank correlation between two variables, one of which is ranked, the other dichotomous. Biometrika 34, 292–296 (1947)
Wickens, T.D.: Multiway Contingency Tables Analysis for the Social Sciences. Erlbaum, Hillsdale (1989)
Wilcox, R.R.: Statistics for the Social Sciences. Academic Press, San Diego (1996)
Wilcox, R.R.: Applying Contemporary Statistical Techniques. Academic Press, San Diego (2003)
Wilcox, R.R., Muska, J.: Measuring effect size: a non-parametric analgue of \(\hat{\omega }^{2}\). Br. J. Math. Stat. Psychol. 52, 93–110 (1999)
Wilcoxon, F.: Individual comparisons by ranking methods. Biom. Bull. 1, 80–83 (1945)
Wilkinson, L.: Statistical methods in psychology journals: guidelines and explanations. Am. Psychol. 54, 594–604 (1999)
Wilks, S.S.: Certain generalizations in the analysis of variance. Biometrika 24, 471–494 (1932)
Wilson, H.G.: Least squares versus minimum absolute deviations estimation in linear models. Decis. Sci. 9, 322–325 (1978)
Yates, F.: Contingency tables involving small numbers and the χ 2 test. Suppl. J. R. Stat. Soc. 1, 217–235 (1934)
Yule, G.U.: On the association of attributes in statistics: with illustrations from the material childhood society. Philos. Trans. R. Soc. Lond. 194, 257–319 (1900)
Yule, G.U.: On the methods of measuring association between two attributes. J. R. Stat. Soc. 75, 579–652 (1912). [Originally a paper read before the Royal Statistical Society on 23 April 1912]
Zwick, R.: Another look at interrater agreement. Psychol. Bull. 103, 374–378 (1988)
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2016 Springer International Publishing Switzerland
About this chapter
Cite this chapter
Berry, K.J., Mielke, P.W., Johnston, J.E. (2016). Regression Analysis of Interval Data. In: Permutation Statistical Methods. Springer, Cham. https://doi.org/10.1007/978-3-319-28770-6_4
Download citation
DOI: https://doi.org/10.1007/978-3-319-28770-6_4
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-28768-3
Online ISBN: 978-3-319-28770-6
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)