Unanticipated Baseline Differences: Evidence of Expectancy Bias in an Adaptive Mobile Learning Intervention in Sub-Saharan Africa

This study presents empirical evidence of expectancy bias emerging before an educational technology intervention within a Sub-Saharan African university context. As part of a larger quasi-experimental investigation evaluating the effectiveness of an Adaptive Mobile Learning System (AMLS) for a PhD degree, a pre-intervention baseline assessment (Survey 2) was conducted to establish group equivalence. However, due to procedural disclosures rather than intentional priming, the baseline assessment revealed that the experimental group, having prior knowledge of their forthcoming use of the AMLS in phase 3, exhibited statistically significant differences, and achieving considerably higher scores than the control group across multiple constructs. This included cognitive learning outcomes, engagement, and adoption perceptions, despite both groups using identical, non-digital study materials at the time. Subsequently, independent samples t-tests and Cohen’s d analyses indicated consistent small-to-moderate effect sizes (0.35 to 0.45), indicating a pervasive influence of expectancy bias present even before the intervention began. The observed bias is interpreted through complementary psychological frameworks like the Hawthorne effect (enhanced performance through awareness of observation), placebo effects in learning (performance shaped by perceived novelty), and demand characteristics (responses aligned with anticipated expectations). Together, these perspectives illustrate a form of psychological priming that may elevate learners’ motivation, self-efficacy, and perceived competence in anticipation of innovative learning technologies. These findings underscore critical methodological implications for educational technology research, particularly in low-resource settings, where participants’ expectations may inadvertently distort baseline assessments. Recommendations are provided to enhance experimental rigor and ensure a more valid interpretation of intervention effects in future studies.