Foreign Language Enjoyment Scale Shows Variable Reliability Across Contexts

A reliability generalization meta-analysis published in Current Psychology has examined the consistency of the Foreign Language Enjoyment (FLE) scale, a widely used instrument in second language acquisition research. The study, conducted by Gizem Mutlu-Gülbak and Cem Oktay Güzeller, systematically evaluated the reliability properties of the scale developed by Dewaele and MacIntyre in 2014.

The meta-analysis included 192 independent datasets obtained from the Web of Science and Scopus databases. The researchers followed the REGEMA guidelines established by Sánchez-Meca and colleagues in 2021 to conduct a systematic literature review and data extraction. The analysis applied a random-effects model based on restricted maximum likelihood estimation, incorporating the Knapp–Hartung adjustment to account for the variability between studies.

The primary finding of the study is that the pooled Cronbach’s alpha coefficient for the FLE scale was 0.874. The 95% confidence interval for this estimate ranged from 0.867 to 0.881, while the 95% prediction interval was broader, spanning from 0.779 to 0.969.

Substantial heterogeneity was observed in the data, with an I-squared value of 98.06. This high level of heterogeneity indicates that reliability varies considerably across different samples and contexts. To understand the sources of this variation, the researchers conducted moderator analyses. They found no statistically significant difference in reliability between studies published in the recent period of 2022 to 2025, which had a mean of 0.876, and those published between 2015 and 2021, which had a mean of 0.867.

Sample size also did not emerge as a significant moderator of reliability. The analysis showed no significant effect of sample size, with a Qm statistic of 1.48 and a p-value of .478. However, the form of the scale used did have a significant impact on reliability. The form type was a significant moderator, with a Qm statistic of 11.12 and a p-value of .004. The meta-analytic pooled estimate was highest for adapted forms of the scale, with a mean of 0.884, and lowest for the long form, which had a mean of 0.857.

Additionally, a weak negative association was observed between the number of items in the scale and its reliability, with a correlation coefficient of r = -0.15. Diagnostic tests of the data indicated funnel-plot asymmetry, which is consistent with possible small-study effects.

The study concludes that while the Foreign Language Enjoyment scale demonstrates satisfactory internal consistency on average, the substantial heterogeneity observed indicates that reliability cannot be assumed to be constant. Researchers are advised to evaluate the scale separately in each new sample to ensure its appropriateness for their specific context. By documenting both the average reliability and the contextual variability of this widely used enjoyment measure, the study provides positive psychology research in second language acquisition with an empirically grounded basis for evaluating the scale in future applications.

Subject of Research: Reliability generalization meta-analysis of the Foreign Language Enjoyment scale

Article Title: Foreign language enjoyment (FLE) Scale: A reliability generalization meta-analysis

Article References: Mutlu-Gülbak, G., & Güzeller, C. O. (2026). Foreign language enjoyment (FLE) Scale: A reliability generalization meta-analysis. Current Psychology, 45(22), Article 1768. https://doi.org/10.1007/s12144-026-10339-7

Image Credits: AI Generated

DOI: 10.1007/s12144-026-10339-7

Keywords: Foreign Language Enjoyment, Reliability Generalization, Meta-Analysis, Cronbach's Alpha, Second Language Acquisition, Psychometric Properties, Heterogeneity, Moderator Analysis, Internal Consistency, Scale Validation, Positive Psychology, Educational Measurement, Dewaele and MacIntyre, Random-Effects Model, Prediction Interval, Small-Study Effects, Form Type, Sample Size, Statistical Heterogeneity, Research Methodology