Practice Set 1 — Measurement & Evaluation

A is true: with outliers in open-ended project scoring, the median summarises typical performance more fairly than the mean. R is also true and is precisely the mechanism — the median is a positional statistic fixed by the rank of the middle case, so extreme magnitudes cannot pull it. (b) is wrong because R is not an unrelated truth; it states the exact reason the median resists distortion; (c) misjudges R, which correctly describes the median as positional rather than magnitude-based; (d) denies A, yet the superiority of the median under outliers is standard practice in project assessment.

Statements I, II and III are correct: standardized instructions, uniform testing conditions and objective scoring keys each remove a source of random error, and reducing error variance is what raises reliability. Statement IV works the other way — a very short test samples the domain thinly, so chance weighs more heavily and reliability falls. (b) includes statement IV alongside two valid measures; (c) likewise pairs the shortening error with instructions and scoring keys; (d) repeats the same mistake with instructions and uniform conditions — each fails on statement IV alone.

Both statements are true and R explains A: tests yield quantifiable evidence of attainment, that evidence exposes where instruction has fallen short, and it is precisely this diagnostic value that makes testing indispensable to a Principal's quality assurance and academic supervision. (b) is wrong because the evidence-based supervision described in R is the very ground of the claim in A; (c) is wrong because R accurately states what test data contribute to supervision; (d) is wrong because the usefulness of tests for institutional quality assurance in A is well established.