Assessment
Assessment
Assessment serves different purposes: screening (who may need help), diagnosis (what the specific needs are), progress monitoring (is the intervention working), and summative evaluation (what has been learned). Norm-referenced tests compare a student with peers and report standard scores (mean 100, SD 15), percentile ranks, and stanines; criterion-referenced tests compare performance with a mastery standard. Grade equivalents are easily misread - a 2.0 grade equivalent does not mean the student belongs in second grade.
Distinguish reliability (consistency) from validity (measuring what it claims). A test can be reliable but invalid for a given purpose; validity requires reliability. Standard error of measurement reminds you to report confidence intervals rather than treating a score as exact. Watch for floor and ceiling effects when a test is far too easy or too hard.
Progress monitoring requires brief, repeatable measures sensitive to short-term growth - curriculum-based measurement such as words correct per minute or digits correct - graphed against an aim line with decision rules (for example, four data points below the aim line trigger an instructional change). MTSS/RTI tiers use universal screening and increasingly intensive, research-based interventions, though RTI may never be required as a precondition for evaluation.
Traps: reading percentile rank as percent correct; assuming a standard score of 85 is 'average'; using a norm-referenced achievement test for weekly progress monitoring; and overlooking the difference between testing accommodations and alternate assessments, which are reserved for students with the most significant cognitive disabilities.
Worked example: A standard score of 85 is one standard deviation below the mean - roughly the 16th percentile - low average, not a severe discrepancy by itself.
Rule of thumb: match the tool to the question being asked.