Assessing grammatical and lexical competence among Honduran lower-secondary students: preliminary validity evidence from the EGLCA-9

Assessing grammatical and lexical competence remains challenging in educational systems where locally validated and context-sensitive EFL assessments are scarce. This study reports the development and preliminary validation of the English Grammatical and Lexical Competence Assessment (EGLCA-9) for ninth-grade students in three public schools in Tegucigalpa, Honduras. The pilot pool comprised 35 four-option multiple-choice items. After preliminary item screening, 29 items were subjected to exploratory factor analysis (EFA) and 22 were retained for subsequent analyses. Parallel analysis supported a correlated two-factor representation comprising a 15-item grammatical factor and a 7-item lexical factor ( φ = .66, 95% CI [.56,.70]). The retained PAF solution yielded reported variance proportions of 34.87% and 9.72% for the two factors, respectively (44.6% cumulatively). Internal consistency estimates based on observed binary responses were ω = .847 for the 22-item total score; .820 for the grammatical subtest, and .671 for the lexical subtest. Separate unidimensional IRT calibrations showed adequate absolute fit for both subtests. The 2PL model was retained as the primary descriptive calibration model, although comparisons with the 1PL model were mixed, particularly for the lexical subtest; 3PL models did not converge. Measurement precision was greatest near θ = .65 for the grammatical subtest and θ = −.02 for the lexical subtest. No local-dependence violations or sex DIF were detected under the applied procedures. Therefore, the findings provide preliminary evidence for interpreting EGLCA-9 scores as indicators of grammatical and lexical competence in this development sample. Nevertheless, they do not establish national norms, formal CEFR alignment, or evidence for admission or placement decisions.