Perceptions of Kuwaiti pupils’ financial literacy and behaviour surveys

About the study

This report documents the 2024-2025 phase of the evaluation of the Bankee programme, funded by the National Bank of Kuwait, focusing on the development, piloting, and validation of survey instruments designed to support ongoing monitoring and future impact evaluation in Kuwaiti primary schools. Rather than assessing programme impact, the study concentrates on establishing robust, reliable, and context-appropriate measurement tools that can be used repeatedly as Bankee scales.

Led by NatCen International in collaboration with NatCen’s Centre for Evaluation and the Questionnaire Development Team, as well as Creative Confidence Consulting Company (CCCC), the study followed a co-production and bilingual design approach. Three survey instruments were developed and tested: a Financial Literacy and Behaviour (FL/FB) survey for teachers, a Classroom Behaviour (CB) survey for teachers, and a parallel CB survey for school principals. Together, these tools aim to capture teachers’ and school principals’ perceptions of pupils’ financial understanding and everyday classroom behaviours aligned with Bankee’s Theory of Change.

The instruments were piloted across 12 Bankee schools, with participation from teachers and principals who completed the surveys in Arabic or English. Quantitative psychometric testing assessed factor structure, reliability, and overall measurement performance, while an embedded feedback module collected user perspectives on clarity, length, and contextual fit. This dual focus allowed the study to refine the instruments while minimising respondent burden.

The findings demonstrate that the surveys are psychometrically sound and suitable for use in future pre-post evaluations, with targeted refinements identified for subsequent rounds. By providing validated, bilingual tools tailored to the Kuwaiti educational context, this study lays the groundwork for generating consistent, decision-useful evidence on pupils’ financial literacy and classroom behaviour as the Bankee Programme continues to expand.

Findings

The findings demonstrate that the Bankee survey instruments are psychometrically sound, contextually appropriate, and ready for use in future pre-post evaluations. Based on a pilot conducted across 12 Bankee schools, with participation from 293 teachers and 12 principals, the study established a robust empirical foundation for validation and refinement. Psychometric testing confirmed that the instruments capture coherent constructs aligned with the Bankee Theory of Change, while remaining practical for use in school settings. Iterative factor and reliability analyses led to targeted refinements, including item pruning, reallocation of overlapping content, and clarification of wording, resulting in improved model fit and strong internal consistency across instruments. Descriptive findings from the pilot provide useful contextual signals about perceived pupil financial literacy and classroom behaviour, particularly differences between public and private schools, but are presented strictly as baseline patterns rather than evidence of impact. User feedback indicated general acceptability of the surveys, alongside clear guidance for future improvement, including reducing length, improving clarity in some Arabic wording, and strengthening digital delivery. Taken together, the findings confirm that the instruments are fit for purpose, with clear priorities identified for strengthening future waves through larger leadership samples, formal cognitive interviewing, and alignment with ongoing programme developments.

Key findings

• The pilot achieved a strong teacher sample and a sufficient principal sample for reliability testing, establishing a solid baseline for validation.
• The teacher FL/FB instrument demonstrated strong internal consistency (α≈0.95) following refinement.
• Exploratory and confirmatory factor analyses improved model fit by resolving overlap and simplifying structure.
• The revised FL/FB structure better reflects how teachers judge pupils’ financial understanding and behaviour.
• The teacher Classroom Behaviour instrument showed high reliability (α≈0.94) across five intended domains.
• Engagement and Collaboration emerged as closely related but distinct constructs in classroom practice.
• The principal Classroom Behaviour instrument showed acceptable reliability, with further testing needed at scale.
• Embedded feedback suggested overall acceptability, with calls to shorten instruments and streamline delivery.
• Descriptive results reflect implementation context rather than programme impact.
• The instruments are ready for integration into powered, stratified pre-post evaluations with minor refinements.

Methodology

This study focused on the development, piloting, and validation of bilingual survey instruments to support future pre-post evaluation of the Bankee Programme in Kuwaiti primary schools. Instrument development followed an iterative, co-production approach involving NatCen’s Questionnaire Development Centre (QDT), Evaluation, and International teams, working closely with Creative Confidence Consultancy (CCCC). Early technical workshops established survey architecture, including question flow, routing, length limits, and online programming requirements, alongside expectations for bilingual drafting and translation.

Survey items were developed separately for (i) a Teacher Financial Literacy and Behaviour (FL/FB) survey and (ii) Classroom Behaviour (CB) surveys for teachers and principals. Item development was informed by the Bankee Theory of Change, findings from the 2023-24 evaluation, programme documentation (including bonuses and fines lists), and relevant academic literature and validated tools. Drafts were reviewed iteratively across teams and with CCCC to ensure conceptual coverage, clarity, and reduced respondent burden.

Psychometric testing included reliability testing using Cronbach’s alpha and item-total correlations, alongside confirmatory and exploratory factor analyses for teacher surveys. These analyses informed targeted item removal, reassignment, and restructuring to improve model fit and internal consistency. Due to small sample size, factor analysis was not conducted for the principal survey, though reliability was assessed.

As formal cognitive interviewing was not feasible, an embedded feedback module captured structured quantitative and qualitative feedback on survey length, clarity, flow, and relevance. This feedback was analysed alongside psychometric results to inform final instrument refinement.