Frustration is an emotion that we experience daily in situations where the expected rewards do not correspond to those obtained. Although it is evident that not all individuals respond in the same way to the same experience, most of us react in a similar way to intense stressors. However, there is immense inter-individual variability for moderate stressors such as frustration (Cuenya et al., 2013). Studies on frustration began in the late 1920s (Tinklepaugh, 1928) and is a focus of interest today, since it is a feeling that we all experience throughout life and that can undermine the performance of coping skills and realistic assessment of tasks (DiGiuseppe et al., 2015). In psychology, frustration has been studied from two main approaches: the Dollard et al. (1939) model and the Amsel model (Amsel, 1958; Amsel, 1992). The first defines it as a stimulus (an interference) between a sequence of responses that are directed towards a goal and has been induced under laboratory conditions through different experimental procedures. The second defines it as the response of the organism (cognitive, physiological, and behavioral) to stimuli that imply a decrease, omission, or interference of reinforcers, and analyzes it through the study of the different individual reactions in situations related to the non-fulfillment of expectations.

In this context, the theoretical background of the formation of the concept of Frustration and Discomfort Intolerance (FDI) is based on Rational Emotive Behavioral Therapy (REBT, Ellis & Blau, 1998), which was the first therapy that explicitly described FDI and incorporated it into a systematic cognitive model (Harrington, 2011). The REBT established that they are not the facts or actions themselves that disturb or generate discomfort, but the judgments/beliefs about them. These beliefs can be expressed in the form of wishes or preferences (rational beliefs) or as an obligation, imperative, or requirement (irrational beliefs). Therefore, intolerance to frustration and discomfort is an irrational belief and can be observed in emotional problems related to self-esteem (Ellis, 2003).

Harrington’s (2005) Frustration Discomfort Scale (FDS) was based on REBT theory, and it was, to our best knowledge, the first instrument to measure FDI precisely. Some of the previous scales, such as the Belief Scale (Malouff & Schutte, 1986) or the Survey of Personal Beliefs (Demaria et al., 1989) measure FDI as a one-dimensional construct and provide general scores with little utility in predicting self-control problems (Kendall et al., 1995). Other scales, such as the Dysfunctional Attitude Scale (Weissman & Beck, 1978), the Multidimensional Improvement Scale (Frost et al., 1990), and the General Attitude and Belief Scale (DiGiusepee et al., 1988), evaluate FDI together with other beliefs related to self-esteem without clearly differentiating these constructs. Also, the content range is somewhat limited, and the items are very similar. However, the FDS (Harrington, 2005), whose construct validity and psychometric properties have been verified, collects the range of irrational beliefs related to FDI described in the REBT theory.

The FDS evaluates four levels: 1) discomfort intolerance (i.e., beliefs associated with intolerance to discomfort, effort, or inconvenience): represents the belief that life should be easy, comfortable, and free from hassles and efforts; 2) entitlement (i.e., a desire for justice; gratification must be satisfied immediately): it refers to frustration with other people, and represents the belief that other people should please us and not frustrate our desires; 3) emotional intolerance (i.e., beliefs about the unbearableness of thoughts and feelings that disturb us emotionally and should therefore be relieved as soon as possible or avoided): it reflects the belief that uncertainty and the thoughts and feelings associated with the emotional distress it causes should be avoided, and 4) achievement frustration (i.e., perfectionist beliefs that should not be frustrated): perfectionist achievement beliefs related to a task (Harrington, 2005).

The factor analysis supported the hypothesis that the FDI is a multidimensional construct showing high internal consistency. The FDS was designed for a Scottish population and has already been validated in countries such as Serbia (Stanković & Vukosavljević-Gvozden, 2011), whose results support the multidimensional model of intolerance to frustration, as well as Turkey (Ozer et al., 2012) and Argentina (Medrano et al., 2018). In Italy (Tripaldi et al., 2018), the multifactorial structure was confirmed, although with a different number of factors compared with the original scale.

Furthermore, the Italian validation examined the factorial structure of one factor, three factors, and four factors of the original FDS. Emotional intolerance was ruled out as a factor. These results are likely to be related to a limited range of scores. The authors concluded that this fact considerably influenced the factor structure of the scale (Tripaldi et al., 2018). On the other hand, the French validation data suggested a one-dimensional structure (Chamayou et al., 2016). For the Turkish version of the FDS, the authors removed four problematic items due to low correlation. The poor loadings of these items might reflect the particular characteristics of the Turkish sample population (Ozer et al., 2012). Additionally, the Argentinian validation of FDS (Medrano et al., 2018) provided the most extensive set of fit with a four-factor CFA first order and one-factor CFA second order.

However, it is challenging to establish further differences between the validations of the scale since the FDI has not been related to the same variables, except for self-esteem. For example, the Italian validation measured depression (Tripaldi et al., 2018), the Serbian validation measured depression, anger, and anxiety (Stanković & Vukosavljević-Gvozden, 2011), the French validation measured emotional regulation and perceived stress (Chamayou et al., 2016), and the Turkish, procrastination (Ozer et al., 2012). On the other hand, the relationship between FDI and unhealthy emotions and assertive behavior (Filippello et al., 2014) and emotional problems (Jibeen, 2013) has been studied, along with addiction to the internet (Kewalramani & Pandey, 2020), tendency to drug abuse (Adroom et al., 2016) or emotional regulation, and stress (Chamayou et al., 2016). Intolerance to the inevitable annoyances associated with goal-seeking or adjustments to goals implies a decline in life satisfaction; that is, discomfort beliefs impair life satisfaction because they reduce commitment to long-term goals, which require more effort (DiGiuseppe, 1991). Frustration generates negative emotions in such a way that its expression, magnitude, and duration may be related to the processes of emotion regulation (Gross, 2002). Therefore, a significant relationship could be found between the FDS scale and emotional skills (i.e., emotional intelligence). With another frustration scale, this relation has been shown (Kumari & Gupta, 2015). However, the majority of the previous FDS validations did not study the relationships with personal resources (such as emotional intelligence) and positive life events (such as flourishing) that have been shown to be protectors of discomfort and facilitators of well-being and happiness (Cordeiro et al., 2016; Tejada-Gallardo et al., 2020).

To the best of our knowledge, there are no studies on the FDI in Chile, nor are there studies that have used other scales to measure constructs similar to the FDI. Only a few studies that use the Basic Psychological Needs Satisfaction Frustration Scale (BPNSFS; Chen et al., 2015) have been conducted in Chile. This instrument measures psychological needs (autonomy, competence, and relatedness) defined by the self-determination theory and understands frustration as the opposite of satisfaction when experiencing negative feelings for not meeting those needs. Validation of the instrument was carried out with Chilean university students (Del Valle et al., 2018). Also studied, with the BPNSFS, was the relationship between the satisfaction of needs and materialism, attitudes, and well-being in Chilean workers (Unanue et al., 2017).

Given that there are no instruments to specifically study FDI, it is crucial to have a validated scale to evaluate it. The purpose of the present study was to validate the FDS using a Chilean adult population and to examine relationships between these construct and emotional variables that could eventually have a protector role against frustration and discomfort.

Method

Participants

From our analysis, we determined that to achieve power = .95 to detect a medium effect size (Faul et al., 2007), we would need a sample of 220 participants. Even so, the sample of participants for this non-probabilistic study was composed of 1187 Chilean university students (whose personal information was removed for blind review), their relatives and acquaintances, and they were surveyed applying snowball sampling. They came from different regions of Chile, although the regions were not homogeneously represented. Eighty-nine percent of the sample was from the Bío Bío region, which is one of the most densely populated in the country. The socioeconomic status was self-reported and was based on the last family income statement declared in the national CASEN survey (Ministerio de Desarrollo Social y Familia, 2017). It defines the following socioeconomic status: low (family income from 626,021 to 1,252,041 Chilean pesos for a household of four people), medium (from 1,252,041 to 1,878,062 Chilean pesos), and high (from 1,252,041 to 2,504,083 Chilean pesos). Also, marital status, age, and educational demographics were diverse. The age range was 18 to 96 years. Participants sociodemographic data are listed in Table 1.

Table 1 Sociodemographic characteristics
Full size table

Instruments

Sociodemographic data. A brief questionnaire collected information on age, sex, marital status, educational level, and economic status.

The Frustration Discomfort Scale (FDS; Harrington, 2005) consists of 28 items with four subscales of seven items each: discomfort intolerance, entitlement, emotional intolerance, and achievement. The items are written in terms of intolerance to frustration (“I cannot stand/tolerate/endure”). Respondents estimate the strength with which they feel distressed or frustrated in a particular situation using a 5-point Likert scale (1 = absent, 5 = very strong) (Appendix 1). Higher scores indicate a greater intolerance to frustration. The lowest and highest scores were 28 and 140. The instrument had acceptable subscale reliabilities (as is shown below), with Cronbach’s alpha coefficient of .88 to .92 for the subscales. In addition, its factorial or construct validity was corroborated in other validation studies and scale adaptation to other languages (Chamayou et al., 2016; Medrano et al., 2018; Tripaldi et al., 2018).

The Rosenberg Self-Esteem Scale consists of 10 items using a 4-point Likert scale (1 = strongly agree, 4 = strongly disagree), with no subscales. Total scores ranged from 10 to 40, where the higher the score, the higher the level of self-esteem. The original scale showed good internal reliability (α = 0.87) (Rosenberg, 1989). The Chilean version of the scale has good internal reliability (Mestre & Guil, 2003). For this study, the Cronbach alpha scale is good (α = .81).

The Trait Meta-Mood Scale (TMMS-24; Fernandez-Berrocal et al., 2004; validated in Chile by Espinoza et al., 2015), developed by Salovey et al. (1995), measures emotional intelligence. It consists of 24 items and evaluates three dimensions (eight items per factor): attention to feelings, emotional clarity, and emotional repair. These are scored on a 5-point frequency response scale (1 = never, 5 = very often). The instrument assesses how people reflect on their moods. Scores for each dimension ranged from 10 to 40. The original version obtained alphas ranged from .86 to .90 (Fernandez-Berrocal et al., 2004). The Chilean version has shown acceptable internal consistency and satisfactory test-retest reliability (Espinoza et al., 2015). For this study, the Cronbach alpha scale range from .89 to .87.

The Flourishing Scale (FS; Diener et al., 2009; adapted to Spanish by Ramírez-Maestre et al., 2017) was created to measure the degree of perceived psychological well-being. It consists of eight items on a 5-point Likert scale (1 = totally disagree, 5 = totally agree) with no subscales. Scores can range from 8 (strong disagreement with all items) to 56 (strong agreement with all items). The Cronbach alpha of the original scale is good (α = .87). The Spanish version showed adequate internal consistency (α = 0.89) (Ramírez-Maestre et al., 2017). For this study, the Cronbach alpha scale is good (α = .86).

Procedure

The authors used Google Forms for data collection through a shared link, either sent by the researcher himself to university students or by the students to their acquaintances and/or relatives. Thus, a snowball sampling technique was implemented for data collection. These questionnaires were answered in the presence of the researcher if they were students. The questionnaires included a description of the objectives of the study. It was predicted that the questionnaire survey would take approximately 40 min to be completed. Data were collected anonymously. Only participants who signed informed consent were included in the study. They were no incentives given for participation.

The adaptation of the FDS to the Chilean population was carried out following the guidelines of the International Test Commission (Bartram & Muniz, 2016). The FDS was translated by a bilingual expert in REBT and by an English translator. The differences were then discussed by a panel of three psychologists who are experts in Chilean REBT. They verified the adequacy of the final translation and its appropriateness to the Chilean culture. Finally, a pilot test of the scale was conducted amongst a sample of 50 university students who were asked to answer the questions and comment on their understanding of the items. They confirmed that questions were clear and easy to understand, so no modifications were required. Appendix 1 lists the items (in Spanish). The survey was then carried out and the results were analyzed.

Data Analysis

The statistical package SPSS Version 25.0 was used to calculate descriptive statistics. Regression and correlation analyses were performed to test for convergent and divergent validity. Internal consistency was calculated using Cronbach’s α coefficient (Taber, 2018), and the omega coefficient (Green & Yang, 2015) was used to observe factor reliability. To test the factor structure of the scale, a confirmatory factor analysis (CFA) was performed with AMOS 22. Maximum likelihood estimation was used for analysis. To test model fit, a chi-squared test, degrees of freedom (df), the comparative fit index (CFI), Tucker-Lewis index (TLI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMSR) were used following the criteria for the good fit: CFI > .90, TLI > .90 RMSEA < .08, and SRMSR < .05 (Marsh et al., 2020). To examine the factorial invariance across gender groups, a multiple-group confirmatory factor analysis (MG-CFA) was conducted through configural, metric, and scalar factorial invariance (Ariely & Davidov, 2012). This step invariance process compares each model to the previous one. For this, the RMSEA and CFI are used, and variations in the fit of the models are compared and are quantified through ΔCFI and ΔRMSEA. Given a well-fitting configural model, differences in CFI greater than .01 and differences in RMSEA and SRMR greater than .015 are suggested to be an indication of substantial differences between models (Chen, 2007).

Results

Reliabilities and Associations between FDS and Related Variables

Cronbach’s alpha value for the subscales showed satisfactory internal consistency: discomfort intolerance (α = .76), entitlement (α = .80), emotional intolerance (α = .80), and achievement (α = .75). The correlations between FDS total scores and subscale scores were positive. Construct validity was tested using Pearson product-moment correlations. Pearson correlation values between each item and the total score were significant with a rank between .35 and .66. All correlation indexes were significant (2-tailed) (p < .01).

The relationship between dimensions and total FDS with self-esteem, emotional intelligence dimensions, and flourishing scores were examined using Pearson’s correlation coefficient. As Table 2 indicates, the present results revealed a negative correlation (r = −.133) between FDS and self-esteem, which was weaker but in the expected direction. The correlations between the FDS score and the score in the attention to feelings dimension were positive (r = .165). Negatively significant correlations were also found between the discomfort intolerance and emotional intolerance with flourishing (r = −.101; r = −.113). To further explore the convergent and divergent validity, multiple regression analyses were performed using FDS total and subscales as predictor variables, and the self-esteem, attention to feelings, emotional clarity, emotional repair, and flourishing as outcome. Results highlight total FDS predictor power over self-esteem (β = −.133) and attention to feelings (β = .165). Also, discomfort intolerance (β = −.101) and emotional intolerance (β = −.113) showed a significant negative predictor effect over flourishing. Lastly, emotional intolerance showed a significant negative predictor effect over emotional clarity (β = −.095) and emotional repair (β = −.080).

Table 2 Internal consistencies, correlations and regression between the dimensions of the FDS and other variables
Full size table

Gender Differences

Results shows that women scored slightly higher than men in three subscales (discomfort intolerance, entitlement, and emotional intolerance) and in the total FDS score, but the differences were only significant in the emotional intolerance dimension (t = −2.43, p = .015), with a small effect size (d = .14). The women’s average scores were more significant than the men’s (Table 3).

Table 3 Comparative coefficients for male and female sample
Full size table

Confirmatory Factor Analysis (CFA)

According to previous studies, a series of CFA models were compared. The results showed that Model 1 (one-factor model with “items” residual uncorrelated) did not fit the data well (see Table 4). In the same way, unidimensional Model 2 (one factor with “items” residual correlated) exhibited a questionable fit. Similarly, Model 3 (four-factor uncorrelated with “items” residual uncorrelated) and Model 4 (four-factor correlated with “items” residual correlated) demonstrated an inappropriate fit. In contrast, Model 5 (four-factor correlated with “items” residual correlated) structure showed a good model fit.

Table 4 Goodness-of-fit statistics
Full size table

Note: Model 1: One factor, “items” residual uncorrelated; Model 2: One factor, “items” residual correlated; Model 3: Four factors uncorrelated and “items” residual uncorrelated; Model 4: Four factors correlated and “items” residual uncorrelated; Model 5: Four factors correlated and “items” residual correlated.

As shown in Fig. 1, the target factor loadings for the items ranged from .38 to .83, and all were statistically significant. The omega coefficient reliability indicated high factorial reliability for the full scale (ω = .92) and adequate internal consistency (α = .92). The results are in keeping with those found in the original study and other validation studies of the FDS.

Fig. 1
figure 1

Standardized factor loadings estimated of CFA model for FDS scale validation

Full size image

Note: All significant < .001. Di = Discomfort intolerance; E = Entitlement; Ei = Emotional intolerance; A = Achievement. Items 1–28 can be seen in Annex 1.

Multi-Group Confirmatory Factor Analysis (MG-CFA)

The MG-CFA Model 1 examined configural invariance by four-factor structure across gender groups (Table 5). The results showed that Model 1 configural invariance fitted very well, and all factor loadings were significant (p < .05). Next, a metric invariance model was tested by factor loadings constrained to be equal across groups. The results showed a good Model 2 fit, and no significant changes were found when compared to a configural invariance model. This indicated that factor loadings were invariant across gender groups. Finally, intercepts and factor loadings were constrained to be equal across gender groups to examine the scalar invariance. The results of the scalar model invariance also showed a good fit of Model 3. When Model 2 was compared, no significant changes occurred. Therefore, intercepts and factor loadings were invariant across gender groups. The results of the invariance test revealed that gender differences in the model were not significant. Consequently, the model was consistent across the sexes.

Table 5 Results of the tests of measurement and structural invariance for the multi-group model of FDS across gender
Full size table

Discussion

This is the first study to examine the validity and reliability of the FDS amongst Chilean adults. The Chilean FDS has similar psychometric properties to the original (Harrington, 2005), is theoretically consistent with the original, and reaches comparable psychometric standards. The results presented suggest that the scale can be applied to the Chilean population. The total FDS score and the subscale scores showed adequate internal consistency, which is consistent with other versions (Chamayou et al., 2016; Filippello et al., 2014; Medrano et al., 2018; Ozer et al., 2012; Stanković & Vukosavljević-Gvozden, 2011; Tripaldi et al., 2018); it also showed that the four-factor structure kept the original distribution (Harrington, 2005), and measurement invariance across gender groups was confirmed.

The correlation results revealed associations between total FDI and other variables. Negative correlations were obtained between FDS scores and self-esteem, in line with the British, Turkish, Argentinian, and Serbian validations (Harrington, 2005; Medrano et al., 2018; Ozer et al., 2012; Stanković & Vukosavljević-Gvozden, 2011). This may be because people who get frustrated easily often have evasive personalities with low levels of personal worth (Harrington, 2005). Regarding the relation FDI-EI, we found a positive association between attention to feelings (TMMS-24; Fernandez-Berrocal et al., 2004) and FDI. Consequently, the role of EI as a protector of discomfort could not be observed in our population. This conclusion is in line with the study of Chamayou et al. (2016), where in their validation, the FDI was modeled as a unitary concept. They suggested that the feeling of frustration would be an independent concept of the adapted management of emotions (emotional regulation). On the other hand, we did find a negative correlation between two dimensions of FDS (emotional intolerance and discomfort intolerance) and flourishing. There are examples in the literature of this relation. For instance, it was observed that need satisfaction and need frustration would have different effects, the first over the well-being and the latter over the ill-being (Cordeiro et al., 2016).

Concerning the gender score, the present study has found significant sex differences in emotional intolerance scores of the FDS. Females showed high intolerance in the emotional aspect similar to the findings of the Argentinian and the Turkish validation (Medrano et al., 2018; Ozer et al., 2012). This result may be explained by the fact that females in certain cultures could be more sensitive to emotional information processing (Craig et al., 2009).

The CFA revealed that the Chilean FDS shares the four-factor structure with the original version, which demonstrates its psychometric strength and indicates that the factors represent an underlying multidimensional FDI construct. Finally, the factors corresponding to the dimensions of the FDS were moderately interrelated, supporting the assumption that they are correlated, measuring different aspects of the same construct. The evaluation of the translation indicated that the Spanish language version of the FDS is equivalent to the original. Our four-factor structure results are also consistent with the Turkish version of the FDS (Ozer et al., 2012) as well as with the Argentinian (Medrano et al., 2018). The Serbian and Italian validations of FDS (Stanković & Vukosavljević-Gvozden, 2011; Tripaldi et al., 2018) also present the multifactorial structure. Only the validation results from France showed a one-dimensional structure (Chamayou et al., 2016).

The analysis of the multi-group invariance CFA model revealed that there were no differences in most respects; no significant differences were found between male and female, as occurred in the original scale, Argentinian and Italian validation (Harrington, 2005; Medrano et al., 2018; Tripaldi et al., 2018). This suggests that the FDS can be reliably applied to Chilean female and male adults.

The present study has several limitations. Firstly, the findings should be confirmed since a method of sampling for convenience (snowball sampling) was used. Therefore, representativeness can be questioned, and results are sensitive to sampling bias. Secondly, we used a sample that was not geographically homogenous. It would be interesting to use a random sampling method for generalizing our outcomes. However, since Chile is a country that is culturally homogenous except for small areas with a greater percentage of the indigenous population, we believe that the results could be extended to the whole country.

Despite these limitations, the present study has important practical implications.

The Chilean validation of the FDS offers additional support for the strength of the FDI construct among language groups (in this case, Spanish), and allows researchers to further study frustration in this specific population. The use of this easily administrable scale, together with other reliable and valid scales, will allow checking the predictive power of the FDI and its relationship with other variables. Moreover, the Chilean version of the FDS that we have obtained also contributes to carrying out cross-cultural studies in which theoretical models and training programs can be validated to improve tolerance to frustration in different cultures, especially in Spanish-speaking countries. On the other hand, Chilean professionals in the clinical and educational field can also count on a reliable and valid evaluation instrument that helps to make accurate diagnoses, as well as verify the effectiveness of their therapies and training in both adult women and men. For the first time, as far as we know, a relationship has been found between FDI and the dimension “attention to feelings” of the EI. Furthermore, a negative correlation was observed for flourishing and FDI, indicating that it would be a protector of discomfort and facilitator of well-being. In conclusion, the present study provides promising evidence that the FDS is a reliable and valid instrument for evaluating FDI within the Chilean population.