Assessment, Research, and Evaluation
Assessment essentials
Know reliability (consistency) and validity (accuracy for an intended use). Reliability is necessary but not sufficient for validity. Distinguish norm-referenced tests (compare to a norm group) from criterion-referenced tests (compare to a standard). Interpret standard scores, z-scores, T-scores, stanines, and percentile ranks correctly. Percentile rank is the percentage of the norm group scoring at or below a score-not percent correct. Watch norm-group mismatch, language/cultural bias, accommodations, and informed consent. Research concepts include correlation versus causation, sampling, action research, needs assessment, and program evaluation. Process data describe services; perception data describe attitudes; outcome data describe results.
Traps
A common trap says a high reliability coefficient proves validity. Another equates percentile rank with percentage correct or treats a standard score as a raw score. Correlation questions may imply causation. Also do not call a test “biased” merely because groups differ; examine item and norm evidence.
Worked example
A student scores 85 on a test with mean 100 and SD 15. z = (85 − 100)/15 = −1. T = 50 + 10(−1) = 40. Percentile rank is about 16. The score is one SD below the mean; it is not “85% correct” and not automatically “failing” without a criterion. Rule of thumb: ask what the score means for this student, in this context, for this decision.