Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

The assessment of cognitive abilities for individuals from diverse linguistic and cultural backgrounds has posed a challenge to psychologists for over a century. For example, nonverbal intelligence tests played a significant role in the psychological evaluations of recruits at the first Word War since many of them were either non-English-speaking immigrants or individuals with little or no formal schooling. They also played a significant role when evaluating the immigrants arriving at Ellis island in New York. Today, psychologists are still facing similar issues when assessing the cognitive ability of individuals in the general population who are either linguistically different from the English-speaking population and therefore are at a disadvantage when taking tests that have verbal directions, require verbal expressive skills, or exposure to formal schooling. Given the well-documented changes of the population demographics in the United States, nonverbal intelligence tests are as important today as they were a century ago.

Goals and Rationale for GAMA Development

The General Ability Measure for Adults (GAMA; Naglieri and Bardos 1997), is an instrument that is accessible to persons with a wide variety of backgrounds; it is a nonverbal test that is free of the confounding characteristics of expressive language skills and exposure to a formal English-speaking academic environment. Naglieri and Bardos (1997) stated that the GAMA “evaluates an individual’s overall general ability with items that require the application of reasoning and logic to solve problems that exclusively use abstract designs and shapes” (p. 1). In this respect, the GAMA does not follow a particular theoretical model but rather attempts to offer an alternative to the measure of general ability after recognizing many inherent problems with tests of general ability that use subtests that have been used as measures of nonverbal ability despite the fact that their administration directions include lengthy and wordy verbal directions (see for example the Wechsler series of tests).

History of Test Development

The primary goal for the development of the GAMA was to design a test that assesses general ability through a variety of nonverbal tasks that can be administered in various settings (individual or group) and with multiple formats using an online administration and scoring through the publisher’s Q-global platform (http://www.pearsonclinical.com/psychology/products/100000200/general-ability-measure-for-adults-gama.html), various test item booklets (spiral bound, paper booklet, laminated booklet) and client response answer sheets (hand-scoring, scannable). Another goal was the development of a test that was normed on a large population so that age-specific and age-sensitive norms can be derived by including a sufficient number of individuals per age group. There were indeed a number of adult intelligence tests available at the time the GAMA was designed. However, in some cases and with very known instruments, a careful examination of their norming samples revealed serious and significant limitations in terms of both the formation of the age-norming groups as well as their sample size used to derive their norms. This is especially true for the assessment of older individuals, where one notices that the normative age groups sometimes span over 20 years (i.e., 65–85) thus including within one age group individuals who are developmentally at different life-stages and quite often include small sample sizes. An additional goal of the GAMA was the reduction of the influence of motor requirements through the elimination of manipulatives as well as the reduction of the influence of speed at the item level.

Following pilot studies of approximately 200 test items in an initial item pool, the final test was selected on the basis of a series of psychometric studies. These studies included examining mean scores by age and gender, examining biased items, computing the internal consistency coefficients for each item type as well as item difficulty and item discrimination values. Correlations with traditional individually administered tests were also obtained at the pilot stages of development. Since this is a self-administered test that can be group administered as well, the amount of time necessary for test administration was examined. A 25-minute time interval was selected.

Description of GAMA

The GAMA is comprised of 66 items that are organized in four item types named Matching, Analogies, Sequences, and Construction.

The “Matching” subtest items require the subject to perceive the various shapes and color combinations, pay attention to details, and find the two shapes that are identical (see Fig. 12.1).

Fig. 12.1
figure 1

Matching sample item

Full size image

The “Analogies” subtest requires the examinee to recognize the relationship between two abstract figures in the first pair and then identify the option that completes the relationship in the second pair of designs (see Fig. 12.2).

Fig. 12.2
figure 2

Analogies sample item

Full size image

In the “Sequence” subtest the subject is required to recognize the pattern, shape, and location of a design and complete the logical sequence of the presented pattern of designs (see Fig. 12.3).

Fig. 12.3
figure 3

Sequences sample item

Full size image

Finally, in the “Construction” subtest, “items require the examinee to determine how several shapes can be combined to produce one of the designs” (p. 5; Naglieri and Bardos 1997) (see Fig. 12.4).

Fig. 12.4
figure 4

Construction sample item

Full size image

The four subtests and their scores do not represent different kinds of ability, but rather four different ways to measure general ability in nonverbal means. All items use yellow, white, black, and blue colors to enhance the presentation of the materials, making them attractive and engaging and reduce the effects of impaired color vision for some examinees.

GAMA Administration and Scoring

Administration and scoring of the GAMA is simple. The GAMA can be administered on a computer with Q-global (Pearson 2014) or using paper and pencil format. Examiners who have completed a psychological assessment course and are eligible to administer intelligence and personality tests should have no difficulty with the administration and scoring of the GAMA. In addition, with proper supervision, examiner assistants can be trained to administer the test using the guidelines provided for the various settings and with the various test materials and answer sheets. In all cases the examiner(s) must familiarize themselves with the test materials which include: the technical manual, the item booklet (spiral bound) or the group item booklet and the two types of response forms, and the self-scoring answer sheet.

Computer administration is available through Pearson’s web-based application Q-global (Pearson 2014). Q-global provides on-screen test administration, scoring, and reporting. The examiner(s) should familiarize themselves with both the Q-global platform and the GAMA manual before administration. On-screen instructions and four demonstration items are presented prior to test items and accompanying stimulus. The examinee can access the instruction at any time throughout the assessment and can mark items to review at a later time. On-screen administration of the GAMA is completed entirely online; paper and pencil testing items are not required. The minimal instructions required and items are presented in English; examinees need at least a second to third grade reading level ability to read and understand the instructions and items.

When using paper and pencil administration of the GAMA, the test directions are read out loud by the examiner who encourages the examinee to follow along in the test booklet and complete the four sample items. This affords the examiner the opportunity to teach the client the proper use of the response form. The item booklets are printed in both English and Spanish. Directions for the administration of the test are printed in the last chapter of the manual to facilitate ease of use. Examinees with a second to third grade reading ability should be able to read and understand the minimal printed instructions; however, for those who are unable to read in either English or Spanish, the examiner can pantomime the test directions. Upon completion of the samples, the examinee may begin the test with the exposure of the first item.

Scoring the GAMA is also very simple. The GAMA can be scored using the Q-global web-based scoring, Q-local desktop software, or self-scoring. A Q-global account is required for the web-based scoring. If the GAMA is administered using Q-global (Pearson 2014), a score report can be generated automatically. If the GAMA is administered using paper and pencil, and the examiner wishes to score the GAMA using the Q-Global or Q-Local options, the examinee’s responses for each item can be entered manually and a report can be generated. The report includes all scores, narrative, and profile graphs. For a sample please follow the link into the publisher’s web site. http://images.pearsonclinical.com/images/pa/pdfs/gamaprofile.pdf. Reports can be exported and saved to a computer or network drive. When a large number of examinees are assessed, such as in assessment centers, scannable answer sheets can be used for the administration of the test which thereafter can be scanned by the user or mailed to the publisher for processing. Use of a scanner is probably the most efficient method especially in settings where numerous forms need to be processed and access to the Internet might be limited or not available for various reasons (e.g., correctional facilities). If the self-scoring record form is used, all scoring steps needed are printed on the inside cover of the record form with no need to refer to the technical manual. In the self-scoring form, the examiner will find a section to score the test and obtain subtest and total raw score, calculate the subtest scale scores (mean 10, SD = 3) and the GAMA Total IQ score (mean 100, SD = 15) using the 11 norming tables which are printed on the record form. Additional scores such as comparisons between subtest scores (an optional analysis) percentile scores, confidence intervals, and classification ranges of scores can be calculated on the form.

Standardization and Psychometric Properties of GAMA

According to the 2014 Standards for Educational & Psychological Testing “validity refers to the degree to which evidence and theory support the interpretation of test scores for purposed uses of tests” (p. 11). Tests are used to answer specific questions. The evidence presented in the technical manual and the information generated in the professional literature about a test allows users to judge the quality of inferences that can be made by a test’s score(s) or stated differently, how well the test answers those specific questions. In the next few paragraphs, evidence will be presented regarding the psychometric qualities of the GAMA in support of its claim as a measure of general cognitive ability. Tests are used to answer specific questions. The evidence presented in the technical manual and the information generated in the professional literature about a test allows users to judge the quality of inferences that can be made by a test’s score(s) or stated differently, how well the test answers those specific questions. In the next few paragraphs, evidence will be presented regarding the psychometric qualities of the GAMA in support of its claim as a measure of general cognitive ability.

A normative test requires a well-designed standardization sample. The GAMA standardization sample consisted of 2360 people who ranged in age from 18–96 years and closely approximated the U.S. population according to the 1990 US census (U.S. Department of Commerce 1992, 1994) using gender, educational background, race or ethnic group, and geographic region as stratification variables. Standardization data were collected in 80 cities and 23 states across the U.S. Eleven age groups were used to collect data allowing a sufficient number of individuals to represent each age group. The sample size ranged from 219 individuals for the 70–74 age group to 310 for the 25–34 year age group. This allowed for the calculation of sensitive age-specific norms.

The GAMA offers reliable scores. The median internal consistency across 11 age groups for the GAMA Total score was 0.90 with values ranging from 0.79 for the older group (80 plus years) to 0.94 for the 35–44 year olds. Reliability coefficients greater than 0.90 were observed in seven out of the eleven age groups of the test. Average reliabilities for the four item types were 0.65, 0.66, 0.79, and 0.81 for the Construction, Matching, Sequences, and Analogies subtests, respectively. Stability coefficients were estimated with a test-retest study that included 86 adults tested across a 2–6-week test interval. With a mean test-retest interval of 25 days, the GAMA IQ score produced a stability coefficient of 0.67. The four item types produced test–retest correlations that ranged from 0.38 (Construction) to 0.74 (Sequences). Gain scores of slightly less than one-third standard deviation was consistent across the item types and GAMA IQ Score.

Multiple sources of evidence exist regarding the instrument’s validity as a measure of overall general cognitive ability. These include, the examination of developmental trends across the 11 age groups, relationships with other intelligence tests measuring similar constructs, correlations with achievement tests and performance of individuals of special populations (learning disabilities, deaf, elderly nursing home residents, individuals with traumatic brain injuries, and individuals with intellectual disabilities).

The trend of scores in the GAMA across the 11 age groups followed the expected (for adults) pattern of diminishing over time scores in visual/spatial reasoning skills. Correlations of raw scores with age ranged from −0.43 to −0.56 for the subtests and −0.59 for the GAMA IQ total score.

Mean scores and correlations with individually administered comprehensive batteries such as the Wechsler Adult Intelligence Scales (Wechsler 1981, 1997) and the Kaufman Adolescent & Adult Intelligence Scale (KAIT; Kaufman and Kaufman 1990) were also examined. The GAMA IQ scores earned were consistently similar to scores in the WAIS-R, WAIS-III, and KAIT. In addition, high and significant correlations were obtained between the GAMA and the scale scores that measure similar constructs. This was the case for subjects selected from the regular population (Naglieri and Bardos 1997), college age populations (Lassiter et al. 2000), individuals with learning disabilities (Naglieri and Bardos 1997), individuals with intellectual disabilities (Naglieri and Bardos 1997), and individuals with traumatic brain injuries (Donders 1999; Martin et al. 2000).

In studies that utilized brief intelligence tests such as the Kaufman Brief Intelligence Test (K-BIT; Kaufman and Kaufman 1990), the Shipley Institute of Living Scale (Zachary 1991) and the Wonderlic Personnel Test (Wonderlic 1992), the GAMA performed similarly to these instruments across college age populations (Leverett et al. 2001), individuals in nursing homes (Festa et al. 1999), and individuals with sudden neurological impairment (Davis et al. 2006). These studies illustrate that the GAMA offers similar cognitive ability scores to the other instruments thus leading to similar decisions.

The relationship between the GAMA and measures of academic achievement was also examined. Significant correlations were observed with the Nelson–Denny Reading Test for a sample of college students and with the Mini-Battery of Achievement for a sample derived from the general population (Festa and Bardos 2000). Significant correlations with reading subtests on the Wide Range Achievement Test-Third Edition for a sample of adults with HIV were also observed (Ryan et al. 2008). The magnitude of the coefficients observed are similar (mid 0.50) to the ones reported between other nonverbal reasoning tests and measures of achievement. The Ryan et al. (2008) study compared the GAMA and reading among a cohort of HIV patients that consisted predominantly of ethnic/racial minorities with less than an 8th grade reading level. The GAMA demonstrated a stronger correlation with global neuropsychological functioning than reading scores on the WRAT-3. In participants with less than an 8th grade reading level, the WRAT-3 was not associated with any neuropsychological domain; the GAMA was found to have strong significant correlations. Across all samples, the GAMA was associated less with participant’s education levels than the WRAT-3. The researchers concluded that “… the GAMA appears to provide valuable information for understanding individuals who may have ineffective instruction, and it should be considered in a neuropsychological evaluation when reading scores are < 80 SS” (p. 1029).

The GAMA has been used to examine the relationship between cognitive ability and job performance. Ability scores derived from the GAMA had a correlation with and predicted job performance in a population of probation and parole staff (Ogard and Karr 1998) and in a population of public and private employees in Romania (Ispas et al. 2010). These studies suggest the GAMA may be a useful tool in employee selection.

Numerous cross culture studies have utilized the GAMA as a measure of cognitive ability. The GAMA has been normed in Romania following the same procedures applied in the United States and with a successful extension of the normative population to age 16 (Iliescu and Livinti 2008). Validity evidence has been established with a sample of Romanian university students (Brazdau and Mihai 2011), and a population of public and private employees in Romania (Ispas et al. 2010). A Dutch version of the GAMA was also used for a sample of nurses from the Netherlands that work with individuals with intellectual disabilities and severe behavior problems (Gerits et al. 2004). A study with 113 individuals (mean age 20.4 years old) in Mexico was also conducted and found the GAMA test scores to be similar to those of a matched control sample from the United States (Bardos et al. 2007). Concurrent validity studies utilizing the GAMA with other instruments, such as the Wechsler Abbreviated Scale of Intelligence (WASI; Wechsler 1999) also reported similar mean performance and significant correlations between the two tests for a sample of 59 adults (mean age 32.4 years old) in Greece (Dieti and Bardos 2012) and another sample of 63 individuals (mean age 32.7 years old) who were administered the GAMA and the Kaufman Brief Intelligence Test (K-BIT2; Kaufman and Kaufman 2004). The GAMA was also administered to a matched sample of Greek and U.S. subjects (Petrogiannis et al. 1999), and a sample of adults from six regions in Greece (Spyridaki et al. 2014) with once again similar performance. In all of these cross-cultural studies, the differences between the samples on the GAMA Total scores were minimal, while the correlations obtained with other intelligence tests administered, were statistically significant, offering further support for the “easy transport” of a nonverbal test like the GAMA and its potential use across cultures.

In summary, the studies discussed above demonstrate that the GAMA offers reliable scores and has accumulated numerous evidence for its use and inferences that can be made by its scores as a measure of general cognitive ability.

GAMA Interpretation

Although the administration and scoring procedures of the GAMA are simple and can be accomplished by individuals with varying degrees of training in psychological testing and assessment, interpretation of the GAMA should always be made by individuals with formal training in psychological assessment. As always, it is best to consult the local licensing boards and/or regulatory agencies regarding test administration, test interpretation, and necessary supervisory arrangements.

The GAMA was designed to offer an estimate of a person’s overall cognitive ability measured in nonverbal means. The GAMA total score represents this effort; it is the most reliable score across all age groups; therefore it is the score to use when interpreting the test. Consistent with other intelligence and achievement tests, the GAMA total IQ score is organized with a mean of 100 and a standard deviation of 15. Examiners are strongly encouraged to report confidence intervals scores as well as classification ranges of a person’s overall ability thus enhancing the meaning of the GAMA total IQ score. Although intra-individual subtest scores can be calculated for the four item types to determine strengths and weakness, the four item types were not developed to represent separate cognitive abilities but rather they are different means of assessing general cognitive ability in nonverbal means.

When a strength or a weakness is found in a client’s profile, the score(s) should be interpreted considering both a normative as well as an intra-individual point of view. For example, a scale score of 9 might be a weakness in a person’s intra-individual profile but the score is still ranked in the average range of ability when compared to individuals the same age. This weakness will be described as a relative weakness. A score of 6 identified as a weakness in an examinee’s intra-individual profile analysis is best being described as a cognitive weakness because, it is ranked below one standard deviation when compared to individuals the same chronological age.

The GAMA can also be used as a progress monitoring tool especially for those examinees whose psychological evaluation and possible subsequent therapy is associated with a traumatic brain injury. A table was developed to assist the examiner in determining if a change in scores from one administration to the other is a reliable change. The table offers the range of scores expected for such determination. Finally, in interpreting the GAMA test scores, especially when comparing the test score with other measures of cognitive ability (e.g., WAIS-IV), the examiner should consider the unique features of the test. The GAMA requires no manipulatives, has very minimal directions requiring listening comprehension skills of the English language, and performance on the test is not affected by item-specific bonus points earned for speeded performance.

Practitioners who utilize the GAMA as part of a comprehensive evaluation might wish to compare the GAMA results to scores earned in other intelligence Wechsler Adult Intelligence Scale (WAIS-IV; Wechsler 2008) or Wechsler Abbreviated Scale of Intelligence®—Second Edition (WASI-II; Wechsler 2011) or achievement tests such as the Wide Range Achievement Test—Fourth Edition (WRAT4; Wilkinson and Robertson 2004) or the Basic Achievement Skills Inventory—Survey (BASI-Survey; Bardos 1995). Tables 12.1 and 12.2 present the values needed to consider if the performance in the GAMA is significantly different from these intelligence and achievement tests.

Table 12.1 Values needed for significance when comparing the GAMA score with the WRAT-IV or BASI-survey achievement tests
Full size table
Table 12.2 Values needed for significance when comparing the GAMA score with the WAIS-IV and WASI-II tests of intelligence
Full size table

Strengths and Weaknesses of GAMA

The GAMA with its administration format and excellent psychometric properties offers numerous advantages when an alternative instrument is needed to estimate a person’s overall cognitive ability. Those in private practice who desire an estimate of their adult clients’ cognitive ability can administer the test with minimal effort. Similarly, computer administration and group booklets can assist in the testing of small or large groups of individuals. This might be the case in correctional or personnel settings, in research studies or whenever an efficient measure of cognitive ability is desired. In addition, the GAMA can be easily adopted by researchers from other cultures who desire an adept nonverbal assessment for the testing of individuals or groups. The validity evidence from independent research studies regarding the inferences that can be made by the GAMA score as a measure of general cognitive ability is overwhelmingly supportive.

Regarding its weaknesses, the GAMA does require minimal directions requiring second grade level reading ability and/or some receptive language skills, and it only offers item booklets printed in English and Spanish. Pantomime administration of the test addresses this limitation; however, a better alternative might be translations of these brief directions in various languages. This might make the GAMA more acceptable in cultures outside the United States or for those who live in the US but do not speak English or Spanish. Computer administration is only available in English and requires a third grade reading level. Further development of on-screen administration to include pictorial instructions and instructions in various languages might increase the population of individuals who can utilize this platform.

As of the writing of this chapter, the GAMA is undergoing a national re-standardization. In addition to a more recent normative sample, the GAMA-2 will address much of the feedback we received from users in the field. For example, the GAMA-2 will:

  • include a number of easier items at the beginning of the test to increase the sensitivity of the test for individuals with intellectual disabilities.

  • eliminate written directions to the test in any language. These will be replaced with graphical representations that will communicate the demands of each task to be performed on the test.

  • include two alternate forms of the GAMA-2 to facilitate test–retest and progress monitoring of an individual’s performance.

  • allow examiners to administer and generate report with the latest technologies available (e.g., tablet administration)

Summary

The GAMA is a nonverbal test designed to evaluate an individual’s overall general cognitive ability; it neither requires expressive language nor exposure to formal English-speaking academic content to complete. It offers an efficient research supported means to estimate a person’s overall cognitive ability.