Arabic Leary Impostorism Scale evaluated in university students

A seven-item Arabic version of the Leary Impostorism Scale showed high internal consistency and promising but incomplete evidence of validity in an initial cross-sectional study of 185 Arabic-speaking university students at King Saud University in Riyadh. The instrument, designed to assess feelings of intellectual fraudulence, self-doubt and fear of exposure despite evidence of competence, produced a Cronbach’s alpha of 0.93 and a McDonald’s omega of 0.93. Its strongest results came from correlations with other self-report measures of impostorism, but the researchers described the evidence as preliminary because confirmatory model fit was mixed, test–retest reliability was estimated from only 14 students, and most participants were medical students.

The study was conducted from May to July 2026 among students enrolled in health and non-health colleges at King Saud University. Participants were at least 18 years old, spoke Arabic, were regular students and provided informed consent. The final psychometric sample comprised 185 respondents with complete answers to all seven Arabic Leary Impostorism Scale, or LIS, items. Medical students represented 143 of them, or 77.3%, while nursing accounted for 19, allied medical fields for 11 and non-health specialties for 11; one participant was studying pharmacy. Data were collected through an electronic questionnaire that also included established measures of impostorism, general self-efficacy and perfectionism. The institutional review board of King Saud University’s College of Medicine approved the protocol, and all participants consented electronically.

The Arabic instrument was developed through two independent forward translations, followed by harmonization, independent back-translation into English and review by two native English speakers from the United Kingdom. The original scale developer could not be consulted. After discrepancies were discussed and resolved, the translated questionnaire underwent a face-validity assessment, and the resulting feedback informed the final Arabic version. The original LIS contains seven items rated on a five-point scale, from “not at all characteristic” to “extremely characteristic.” Total scores range from 7 to 35, with higher scores indicating more impostorism. The seven questions target doubt about ability, difficulty internalizing accomplishments, attribution of success to factors such as luck or effort, and fear that others will discover a perceived lack of competence. Its short format is intended to reduce respondent burden when the construct is examined alongside other educational or psychological variables.

Internal consistency was excellent: both Cronbach’s alpha and McDonald’s omega were 0.93, and the mean correlation among items was 0.67. That high inter-item association indicates that the questions formed a coherent scale, although it also raises the possibility that some items may be redundant, an issue the researchers said should be tested through item-reduction studies. Temporal stability was more modest. Among 14 participants who completed the scale again after one to four weeks, the Pearson correlation and intraclass correlation coefficient were both 0.65. A Bland–Altman analysis found little average drift between administrations, with a mean difference of 0.05, but the limits of agreement ranged from −1.21 to 1.32 on the five-point response metric. Because the retest sample was small, these estimates provide only preliminary evidence about the scale’s stability over time.

Factor analyses generally favored the LIS as a single-scale instrument, although not without caveats. The Kaiser–Meyer–Olkin measure of sample suitability was 0.90, and Bartlett’s test of sphericity was significant, χ²(21) = 1536.59, p < 0.001. Exploratory factor analysis based on polychoric correlations, an approach suited to ordinal items, produced loadings from 0.81 to 0.94. The first factor had an eigenvalue of 5.45 and accounted for 96.58% of the estimated common variance in the principal-factor solution. A scree plot also showed a clear break after the first factor. All seven items had positive polychoric correlations. Confirmatory analysis found significant freely estimated loadings for every item, with standardized loadings ranging from 0.78 to 1.00, but the strength of those loadings did not translate into uniformly good fit for the overall model.

Global fit statistics in the one-factor confirmatory model were mixed. The comparative fit index was 0.91 and the standardized root mean square residual was 0.06, offering some support for the model. By contrast, the Tucker–Lewis index was 0.87 and the root mean square error of approximation was 0.19, with a 90% confidence interval of 0.16 to 0.22, indicating suboptimal fit under commonly used criteria. The model’s χ² statistic was also significant, χ²(14) = 105.39, p < 0.001. Because exploratory and confirmatory analyses used the same 185-person sample, the confirmatory results did not independently cross-validate the factor structure. The researchers also noted that the sample remained modest for precise estimation and that replication in larger groups is required. The elevated RMSEA and lower TLI mean that the one-factor solution should not be treated as conclusively established despite the strong standardized loadings and the exploratory result.

Validity evidence was strongest when the Arabic LIS was compared with other instruments targeting the same construct. Scores correlated at 0.76 with the 20-item Clance Impostor Phenomenon Scale, with a 95% confidence interval of 0.69–0.81, and at 0.77 with its 10-item version, with a 95% confidence interval of 0.70–0.82. The LIS also had a correlation of 0.66 with the three-item Impostor Phenomenon Short Scale, with a 95% confidence interval of 0.57–0.73. All three associations were statistically significant at p < 0.001. These results suggest that the brief Arabic LIS captures meaningful overlap with established self-report measures of impostorism, although the analysis cannot separate construct-specific variance from shared response style, social-desirability effects or other forms of method variance because no criterion standard for impostorism exists. The two scales are not interchangeable: the CIPS covers a wider range of impostorism-related content across 20 items, whereas the LIS concentrates on the core experience in seven questions.

Relationships with self-efficacy and perfectionism were less conclusive. The LIS showed a weak positive association with general self-efficacy, r = 0.20, 95% CI 0.05–0.33, rather than the negative relationship observed in a validation of the Arabic CIPS. Its correlations with rigid, self-critical and narcissistic perfectionism were also weak, ranging from 0.16 to 0.18, and all were positive. The perfectionism results were directionally consistent with previous research, but the authors did not prespecify the discriminant hypotheses and did not test the explanations offered afterward. The self-efficacy result is therefore better treated as an open question than as evidence that the LIS has established discriminant validity. Comparisons involving self-efficacy and perfectionism were based on 81 respondents, increasing uncertainty around the estimates. The investigators noted that differences in item content, the restricted score range and the smaller comparison sample could contribute to the discrepancy, while emphasizing that replication is needed to clarify it.

Descriptive scores were relatively low: the mean item response was 1.85, with a standard deviation of 0.95, on the five-point scale. Women had a mean of 1.92 compared with 1.72 for men, but the study did not test measurement invariance across gender, so the difference cannot establish that the scale measured a different underlying construct for the two groups. The results also cannot show that impostorism is declining over time, changing across disciplines or increasing with educational experience. Cross-sectional data, the uneven distribution of academic majors and the concentration of medical students limit those interpretations. The authors identify broader multicenter testing, balanced recruitment across health and non-health fields, and prospective measurement during clinical rotations as next steps for evaluating generalizability, responsiveness and predictive utility.

Overall, the findings support the Arabic LIS as a promising brief measure for survey research involving Arabic-speaking university students, particularly when impostorism is one construct among several. The scale demonstrated excellent internal consistency, a predominantly single-factor structure in exploratory analysis and strong-to-moderate convergence with three established impostorism instruments. Yet its test–retest estimate was based on only 14 students, its confirmatory fit was uneven, and its relationship with general self-efficacy diverged from expectations. The study therefore establishes an initial psychometric foundation rather than a fully validated instrument. Larger and more diverse Arabic-speaking samples, independent factor-analytic replication, formal invariance testing and longitudinal studies of responsiveness are needed before the scale can be regarded as fully established or used to track change over time.

Subject of Research: Initial psychometric evaluation of the Arabic Leary Impostorism Scale among Saudi university students

Article Title: The Arabic Leary Impostorism Scale: initial psychometric evaluation in a university student population

Article References: AbuDujain, N. M., Alswailem, A., Alotaibi, S. K., Alhumud, M. H., Almutairi, J. S., AlSaif, H. I., Hakami, S. A., Jansen, M. P., & Almigbal, T. H. (2026). The Arabic Leary Impostorism Scale: initial psychometric evaluation in a university student population. Frontiers in Psychology, 17, Article 1973835. https://doi.org/10.3389/fpsyg.2026.1973835

Image Credits: AI Generated

DOI: 10.3389/fpsyg.2026.1973835

Keywords: Leary Impostorism Scale, Arabic validation, impostorism, psychometrics, Saudi Arabia, university students, cross-sectional study, self-report measures, factor analysis, measurement reliability, Arabic, Leary