Immersive virtual reality for qualitative fieldwork training in doctoral Education: A quasi-experimental mixed-methods study

This mixed-methods study compared qualitative observation performance in two intact doctoral-student groups in Education (48 participants; 24 per condition), without individual randomisation. The immersive virtual reality (IVR) programme comprised ten 50-min sessions developed in Unity and addressed thick description, contextual attention, nonverbal recording, interpretive inference, and observer reflexivity. Data comprised pretest–posttest rubric scores, field-diary and interview coding, and a participation checklist. Exploratory participant-level Mann–Whitney tests detected no baseline difference (U = 261.00, p = .576), but higher IVR posttest scores (U = 540.50, p < .001, rrb = .877) and gains (U = 572.00, p < .001, rrb = .986). An ANCOVA with HC3 standard errors yielded an adjusted group contrast of 3.49 points (95% CI [2.91, 4.06]). All five dimension-level comparisons remained below .05 after Holm correction under participant-level assumptions. Coded qualitative records described attention to contextual and nonverbal evidence and the distinction between observation and interpretation. These results indicate a preliminary association with the IVR training package. One academic group per condition, an undocumented group-allocation procedure, modest rubric internal consistency, and limited qualitative source verification constrain inference. HC3 adjustment does not account for within-group dependence, and the study cannot separate intervention effects from academic-group differences or establish transfer to authentic fieldwork.