Psychometric Properties of the Generalized Pliance Questionnaire in Spain

Context and Objectives

Rule-governed behavior (RGB) constitutes a central functional-analytic concept for explaining complex human behavior. Within this framework, generalized pliance represents a pattern of rule-following that has been theoretically associated with psychopathology because it prevents verbal construction of values and contact with natural consequences of one's own behavior. The Generalized Pliance Questionnaire (GPQ) was recently developed as a self-report measure to evaluate this construct. Two versions exist: an 18-item version (GPQ-18) and a 9-item version (GPQ-9).

Although the GPQ has demonstrated adequate psychometric properties in Colombian, English, and Irish samples, its analysis in Spanish-speaking contexts such as Spain remains limited. Therefore, the objective of this study was to analyze the factor structure, measurement invariance across gender, internal consistency, and convergent construct validity of both GPQ versions in an online Spanish sample of 846 participants.

Method

Participants

The sample consisted of 846 participants (78.7% female) with ages ranging from 18 to 72 years (M = 35.38, SD = 11.12). Participants responded to an anonymous online survey distributed through social media. All were Spanish nationals. Relative educational level was: 0.1% no studies, 3.3% primary education (compulsory education), 29.7% mid-level study graduates (high school or vocational training), and 65.6% were undergraduate or college graduates (1.3% did not indicate educational level).

Focal Instrument

The Generalized Pliance Questionnaire (GPQ) is an 18-item measure responded to on a 7-point Likert scale (1 = never true; 7 = always true). The short 9-item version (GPQ-9) was developed for survey research applications. Both versions were originally validated in Colombia with excellent psychometric properties.

Other Instruments

To establish convergent construct validity, five self-report measures were utilized:

1. Depression, Anxiety and Stress Scales-21 (DASS-21): Evaluates emotional symptoms; 21 items on a 4-point scale. The Spanish version demonstrated adequate internal consistency and convergent validity.

2. Acceptance and Action Questionnaire-II (AAQ-II): Measures experiential avoidance; 7 items on a 7-point scale. The Spanish version showed one-factor structure and adequate internal consistency.

3. Cognitive Fusion Questionnaire (CFQ): Measures cognitive fusion; 7 items on a 7-point scale. The Spanish version demonstrated excellent internal consistency and measurement invariance across gender.

4. Valuing Questionnaire (VQ): Assesses behavioral consistency with one's values; 10 items on a 7-point scale divided into two subscales (Progress and Obstruction). The Spanish version showed convergent validity and measurement invariance.

5. Satisfaction With Life Scale (SWLS): Measures self-perceived well-being; 5 items on a 7-point scale. Demonstrated one-factor structure and convergent validity.

Analyses

Confirmatory factor analysis (CFA) was conducted using Robust DWLS estimation with polychoric correlations. Model fit for the one-factor model was evaluated using goodness-of-fit indices: RMSEA, CFI, NNFI, and SRMR. Metric and scalar invariance analyses across gender were conducted by comparing progressively more restrictive nested models. Cronbach's alpha with 95% confidence intervals was calculated for internal consistency, corrected item-total correlations were computed, and Pearson correlations assessed convergent validity. Correlations were interpreted as small (.10–.20), medium (.21–.36), or strong (>.36).

Results

Factor Structure

GPQ-18: The one-factor model showed good fit according to CFI (.97), NNFI (.97), and SRMR (0.053), although fit was marginal according to RMSEA (0.100). After allowing error terms to correlate between three item pairs following modification indices, fit improved substantially (RMSEA = 0.084; CFI = .98; NNFI = .98; SRMR = 0.047). Overall model fit is considered acceptable.

GPQ-9: The one-factor model was good according to CFI (.96), NNFI (.95), and SRMR (0.063), but showed poor fit according to RMSEA (0.146). Modification indices did not significantly improve RMSEA. Consequently, GPQ-9 fit was unacceptable and subsequent analyses were conducted only with the GPQ-18.

Measurement Invariance

The GPQ-18 demonstrated both metric and scalar invariance across gender. Changes in RMSEA, CFI, and NNFI were less than .01, meeting all invariance criteria. This indicates the instrument functions equivalently across men and women.

Internal Consistency

The GPQ-18 showed excellent internal consistency (α = .95; 95% CI [.95, .96]). Corrected item-total correlations were high for all items, ranging from .51 (Item 17) to .85 (Item 13). The sample mean score on the GPQ-18 was 58.04 (SD = 20.21). There were no statistically significant differences in scores between genders (men: M = 57.12, SD = 19.47; women: M = 58.52, SD = 20.27; t(811) = 0.82, p = .41).

Convergent Construct Validity

The GPQ-18 showed strong positive correlations with experiential avoidance (r = .53), cognitive fusion (r = .51), and obstruction in values (r = .49). It showed strong negative correlations with progress in values (r = −.37) and life satisfaction (r = −.36). It showed moderate to strong positive correlations with total emotional symptoms (r = .42), depression (r = .42), anxiety (r = .33), and stress (r = .39). All correlations were statistically significant (p < .001).

Discussion and Conclusions

The study demonstrated that the GPQ-18 is a valid measure of generalized pliance in Spanish samples. Specifically, the instrument exhibited a clear one-factor structure, measurement invariance across gender, excellent internal consistency, and convergent construct validity. Findings are coherent with previous research in other populations, replicating positive correlations with experiential avoidance, cognitive fusion, values obstruction, and emotional symptoms, and negative correlations with values progress and life satisfaction.

A unique contribution of this study was the finding of metric and scalar invariance of the GPQ-18 across gender. This is an important contribution because it establishes that comparing scores between men and women is valid. A notable finding is that, in contrast to previous studies with child versions that found gender differences (girls scoring higher than boys), the GPQ-18 showed no significant differences between adult men and women, suggesting that generalized pliance levels might equalize during adulthood.

However, the GPQ-9 showed unacceptable fit in this Spanish sample, contrary to previous validations in Colombian and English samples. This divergence is likely due to methodological deficiencies commonly found in short-form development.

The study presents certain limitations: (a) convenience online sample with greater female representation, (b) absence of psychometric analyses with clinical samples, (c) correlations only with self-report measures without including behavioral measures such as the Wisconsin Card Sorting Test, and (d) mean scores higher than in Colombian samples, requiring cross-cultural invariance analysis. It is recommended that future research analyze criterion validity of the GPQ-18 in Spanish samples and measurement invariance across different Spanish-speaking cultures.

Significance and contribution

This study contributes to the field of clinical psychology by providing evidence of GPQ-18 validity in Spanish-speaking contexts. The instrument demonstrated clear factor structure, measurement invariance across gender enabling valid group comparisons, and theoretically coherent relationships with psychological inflexibility constructs, providing a valid and reliable tool for assessing rule-governed behavior in research and clinical practice.

---