Statistical Validation

Exam Statistics Unique to formative assessment systems, our EdifyStat module automatically generates meaningful item-level information that provides in-depth interpretations of district and school-wide performances and indicate the validity of the items themselves. EdifyStat reports Classical Test Theory (CTT) statistics such as n-count, point-biserial correlations, proportion correct, proportion responding (distractors), discrimination indices, low-endorsement, high-endorsement, and p-values. In addition to reporting the statistics, edifyStat provides an online interpretation that helps the district official determine the quality of the item for use in valid assessment score assumptions. Item StatisticsThat is, edifyStat analyzes the assessment information statistics and reports a summary of whether the item is of sufficient quality to continue using within the assessment. In addition to the item level statistics and interpretations, edifyStat also reports assessment-wide statistics: n-count, mean, variance, standard deviation, minimum score (overall, low group, high group), maximum score (overall, low group, high group), alpha coefficient, and the standard error of measurement. Lastly, to assist in score interpretation, edifyStat reports a Score Distribution graphic that provides the frequency, cumulative frequency, percentile rank, and percent of number of items answered correctly by all sampled examinees involved in the assessment event.