Supplementary Instruments and De-identified Mixed-Methods Data on GenAI Literacy and Lesson Design among Egyptian EFL Pre-service Teachers

This repository contains de-identified research instruments, quantitative score tables, qualitative extracts, and manuscript-support files associated with a mixed-methods study of generative artificial intelligence literacy and GenAI-supported lesson design among Egyptian pre-service EFL teachers.

The study examined participants’ conceptual understanding of generative AI, prompt-engineering competence, pedagogical integration, linguistic and pragmatic evaluation, ethical awareness, and capacity to exercise informed human oversight when designing EFL lessons. Participants were drawn from Beni-Suef University (BSU) and Menoufia University (MU) and included third- and fourth-year pre-service teachers.

The deposited materials are intended to support methodological transparency, secondary examination of the reported variables, instrument adaptation, and replication in comparable teacher-education settings.

Repository Contents

Research instruments

• appendix_galq.tex
  Contains the 20-item GenAI Literacy Questionnaire (GALQ). The instrument uses a five-point agreement scale and comprises five four-item domains:
  1. Conceptual understanding
  2. Prompt engineering for EFL pedagogy
  3. Pedagogical integration and task design
  4. Linguistic and pragmatic evaluation
  5. Ethics, privacy, and national guidelines

• appendix_rubric.tex
  Contains the instructions for the GenAI-supported EFL lesson-design task and its four-criterion analytic rubric. Participants were required to prepare a 45-minute lesson plan, submit a complete prompt log, and provide a 500-word critical commentary explaining how they evaluated, modified, rejected, or retained GenAI-generated material.

  The rubric assesses:

  • Pedagogical alignment
  • Prompt quality
  • Linguistic evaluation
  • Ethical awareness

  Each criterion is scored from 1 to 4, producing a total score from 4 to 16.

• appendix_protocols.tex
  Contains the reflective-journal prompts and semi-structured interview guide. The protocols address participants’ experiences of GenAI output quality and error, interpretations of academic integrity and national AI guidance, rejection or adaptation of generated materials, changes in AI use, the metaphor of the teacher as a filter, and barriers anticipated during school practicum.

Quantitative data tables

• raw_n300.tex
  Contains a de-identified participant-level table for 300 pre-service teachers. Variables include participant ID, university, academic year, pre-test scores for the five GALQ domains, post-test scores for the same domains, and the average lesson-design rubric score.

  Variable groups are:

  • PrA–PrE: pre-test scores for the five GALQ domains
  • PoA–PoE: post-test scores for the five GALQ domains
  • RubAvg: average analytic-rubric score
  • Uni: institutional site
  • Yr: academic year

  The table contains 150 participants from BSU and 150 from MU.

• raw_n142.tex
  Contains de-identified pre-test and post-test analytic-rubric scores for 142 lesson-design records. The table reports criterion-level and total scores.

  Variable abbreviations are:

  • PA: pedagogical alignment
  • PQ: prompt quality
  • LE: linguistic evaluation
  • EA: ethical awareness
  • Tot: total rubric score

  The dataset includes 72 BSU records and 70 MU records.

Qualitative materials

• qual_journals.tex
  Contains 50 de-identified representative reflective-journal extracts, with 25 extracts from each university. Each extract is classified according to a primary analytic focus, including human agency, contextual and cultural adaptation, professional identity, ethical integrity, and implementation challenges.
• qual_interviews.tex
  Contains 47 de-identified representative interview quotations categorized by institutional site and thematic focus. The thematic categories include:
  • Changes in GenAI use
  • Recognition and correction of hallucinations
  • Human agency and teacher oversight
  • Professional credibility and disclosure
  • Practicum and implementation barriers

The qualitative files contain selected representative extracts rather than complete interview transcripts or complete reflective journals. They are provided to document the evidentiary basis of the thematic interpretation while limiting the disclosure of potentially identifying contextual information.

Manuscript-support files

• diff_original.tex
• diff_revised.tex

These files are LaTeX compilation drivers for the original and revised manuscript versions. They require their associated preamble and manuscript-body source files to compile and do not independently contain the manuscript text.

Data Structure and Confidentiality

All participant records use coded identifiers. Direct personal identifiers, including participant names, email addresses, student numbers, and contact information, are not included.

The quantitative files provide domain-level or criterion-level scores rather than item-level survey responses. The qualitative files provide selected de-identified extracts rather than complete transcripts. Users should therefore limit secondary analysis and interpretation to the variables and textual evidence contained in the deposited files.

Institutional abbreviations are used as follows:

• BSU: Beni-Suef University
• MU: Menoufia University

Recommended Use

These materials may be useful for:

• Replication or adaptation of the GALQ
• Evaluation of GenAI literacy in teacher education
• Design of GenAI-supported EFL lesson-planning interventions
• Development of human-in-the-loop assessment rubrics
• Mixed-methods research on responsible GenAI integration
• Comparative research across teacher-education institutions
• Examination of cultural, ethical, and contextual evaluation of GenAI-generated pedagogical content

Researchers adapting the questionnaire, rubric, or qualitative protocols should undertake appropriate contextual validation and obtain the necessary institutional ethical approval before collecting new data.

Citation

Users of these materials should cite both the Zenodo record and the associated research article. The Zenodo DOI should be inserted into the article’s data-availability statement once the repository record has been published (10.5281/zenodo.21478100).