Research
Human-centred educational research
The research programme connects empirical work in language education and assessment with a broader concern: how educational systems can remain technically defensible while responding to the human beings affected by them.
Established research base
The programme grows from work in applied linguistics, English-language education, learner proficiency and fluency, language assessment, curriculum and materials evaluation, mixed-methods inquiry and technology-enhanced learning. The doctoral research examined cultural and pedagogical fit in EFL materials, while earlier work investigated relationships among lexical fluency, temporal fluency and language proficiency.
Materials, assessment and educational evidence
Current public research work includes materials and curriculum evaluation, educational assessment, mixed-methods inquiry, learner needs and the interpretation of evidence in educational decision-making. Projects that remain under confidential development are intentionally not described on this public site.
From research questions to defensible educational claims
Educational evidence does not become meaningful simply because it is quantified. Its value depends on the question being asked, the conditions under which observations were produced, and the decisions that the evidence is expected to support. A learner’s performance on a particular task, for example, may reveal something important about present attainment without establishing a stable limit on future development. Interpreting that distinction requires attention to opportunity, instructional context, measurement limitations and plausible alternative explanations.
This perspective also informs Dr. Rezazadeh Yazdeli’s approach to the relationship between language education, assessment, and learner development. Empirical findings, theoretical arguments and professional observations should be brought into dialogue without being treated as interchangeable forms of proof. Where a proposed framework remains under development, its conceptual rationale can be explained while claims of predictive accuracy, effectiveness or psychometric validity are reserved for appropriately designed investigations.
Golden Talent
The talent-development programme is designed around multi-source evidence rather than one-test labelling. Its longer-term infrastructure is being built to distinguish student self-report, parent information, teacher observation, academic records and assessment-derived evidence, while preserving instrument versions and research provenance for later validation studies.
Academic context and interpretation
Evidence is valuable when the limits of its interpretation remain visible
In educational research, identifying a pattern is not the same as explaining why that pattern emerged or establishing that it will appear in another setting. Materials, learner histories, classroom interaction and institutional conditions can influence the meaning of the evidence that a study produces. It is therefore important to connect methodological decisions with the scope of the claims being made, and to distinguish what is observed from what remains an interpretation requiring further examination.
The developing Golden Talent framework draws on this concern by separating evidence sources and by treating pathway decisions as hypotheses that can be reviewed through experience and feedback. Such an approach may provide a rationale for future inquiry, but the framework's proposed instruments require their own studies of content, use, reliability and validity. Conceptual coherence is a starting point for research rather than proof of an educational effect.
For a fuller account of the proposed framework and its limitations, see Golden Talent and the methodology.
The separate Journal of Human-Centred Education, Learning and Assessment (JHELA) is developing an English-language scholarly forum in related areas. It is currently inviting qualified reviewers to express interest. Its editorial operations and inaugural issue remain in preparation, and the journal should not be confused with the verified personal publication record presented elsewhere on this site.
Research integrity · AI-assisted academic writing
Using AI in academic writing without transferring authorship or source responsibility
AI-assisted drafting can make some academic tasks more manageable, including the organisation of a working outline, comparison of alternative phrasing and critical examination of the clarity of an argument. Yet these uses do not alter the central responsibility of the researcher: to determine the research question, justify methodological choices, interpret evidence and remain accountable for every published claim. A fluent paragraph is not evidence that its assertions are true, that a cited study exists or that its interpretation is warranted.
Source verification requires returning to the original scholarly record. For each proposed citation, the author should establish that the work exists, confirm its bibliographic metadata through an authoritative publication or identifier record, read the relevant passage in context and check whether it supports the precise proposition being made. AI-generated titles, quotations, page numbers, DOIs and apparently plausible bibliographies should not enter a manuscript merely because they look scholarly. Where the underlying source is inaccessible, the claim must be qualified, independently supported or withheld.
Responsible practice also requires a traceable account of assistance. Researchers should observe the disclosure requirements of their institution, publisher or target journal; document consequential AI use where required; protect confidential data, unpublished submissions and personal information; and check that any permitted automation has not displaced human review. Authorship and contributor roles must follow the applicable scholarly policies rather than being assigned to a software system. These safeguards help distinguish linguistic support from intellectual work and technical convenience from evidence-based scholarship.
For related research context, consult the verified publication record, the academic profile and the existing academic services overview. These links document the scope of the work presented here; they do not certify third-party AI output or guarantee the acceptance of any manuscript.