Item analysis after online exams is essential for maintaining assessment validity, evaluating question quality, and ensuring fairness across candidate cohorts. Rather than judging a summative test solely by aggregate pass rates, post-exam statistical evaluation scrutinises response patterns for each individual question. This empirical review highlights whether specific assessment items accurately measured candidate competence or introduced bias through ambiguous phrasing, miskeyed answers, or flawed response choices.
The two primary metrics utilised during this process are item difficulty and item discrimination. The difficulty index, or facility value, reflects the proportion of candidates who answered an item correctly, immediately revealing questions that were unexpectedly arduous or uncharacteristically trivial. Simultaneously, the discrimination index evaluates how effectively an item distinguishes between high-performing and lower-performing test-takers, verifying that candidates with stronger overall subject mastery were genuinely more likely to secure the mark.
Distractor analysis forms another vital component of post-assessment reviews for objective question formats. By examining how frequently candidates selected each incorrect option in multiple-choice questions, examiners can identify non-functioning distractors that attracted zero responses, as well as plausible distractors that disproportionately trapped high achievers due to confusing wording. These insights enable awarding bodies to adjust provisional scoring where necessary before final results are published.
Systematic post-test reviews provide the structured evidence required to calibrate and maintain long-term question banks. Retiring poor items and refining borderline prompts over time ensures that future assessment forms remain balanced, reliable, and legally defensible. OnlineTestPlus provides educators and awarding organisations with automated assessment reporting and item-level response analytics, simplifying post-exam reviews and strengthening the integrity of digital testing programmes.