VS flashcards

32 cards. Read the term and the definition here. The deck at the top of this page is optional.

Blueprint weight sum

The published domain weights of a complete exam blueprint must total 100 percent; any other total means one or more weights is missing or mis-transcribed and must be re-verified.

Primary source rule

A blueprint claim counts as verified only when two independent primary sources agree; secondary summaries, forums, and unofficial study guides serve as leads rather than evidence.

Null with review flag

When a fact cannot be confirmed, store null and mark the record needs_review instead of guessing; an explicit gap is safer than a confident error.

Content outline crosswalk

A crosswalk table maps each official domain to every internal topic, exposing topics that no domain covers and topics claimed by more than one domain.

Domain weight sampling check

Multiply each domain weight by the total item count to obtain expected items per domain; observed counts far from expectation signal blueprint drift or mapping errors.

Version and retrieval date

Every blueprint-derived artifact records the blueprint's version identifier and the date it was retrieved, so stale content can be detected and retired.

Placeholder domain

A provisional domain label used before the official outline is available; it must be renamed and re-mapped once the real blueprint arrives.

Scope boundary test

A topic belongs to a domain only if the official outline names it there; topical similarity alone is not sufficient evidence of correct placement.

Difficulty p-value

The p-value is the proportion of examinees answering correctly, ranging from 0 to 1; higher values mean easier items, and values near 0.5 allow maximum discrimination.

Discrimination index D

D equals the proportion correct in the upper scoring group minus the proportion correct in the lower group, typically the top and bottom 27 percent of examinees.

Distractor analysis rule

A distractor chosen by fewer than about five percent of examinees, or chosen more often by high scorers than low scorers, is non-functioning and should be revised.

Standard error of measurement

SEM equals the standard deviation multiplied by the square root of one minus reliability; it estimates the typical spread of a single observed score around the true score.

KR-20 reliability

An internal-consistency estimate for dichotomously scored items; values near 1 indicate consistent measurement, and longer, more homogeneous tests tend to produce higher values.

Guessing correction formula

Corrected score equals rights minus wrongs divided by options-minus-one; it removes the expected credit that random guessing would otherwise contribute to a multiple-choice total.

Item-total point-biserial

The correlation between an item's score and the total test score; negative or near-zero values flag items that measure something other than the intended construct.

Item revision decision rule

Revise an item when difficulty is extreme and discrimination is low; retire it when neither revision nor re-analysis restores acceptable difficulty and discrimination statistics.

Spaced retrieval schedule

Review intervals that expand over time, such as one day, three days, one week, and one month, counter the forgetting curve better than repeated massed reading.

Testing effect

Retrieving information from memory strengthens later recall more than rereading the same material for an equal amount of time, even though rereading feels easier.

Interleaved practice

Mixing different problem types within a single session improves discrimination among problem types and transfer, though it feels harder and slower than blocked practice.

Working memory limit

Working memory holds roughly four novel chunks of information at once, so new material should be introduced in small, connected pieces.

Forgetting curve

Retention of unrehearsed material decays approximately exponentially with time; the steepest loss occurs within the first day after initial learning.

Deliberate practice

Practice focused on a specific weakness, with immediate feedback and repeated correction, produces larger gains than undirected repetition of already-easy tasks.

Mastery criterion

A predefined performance threshold, such as several consecutive correct attempts, that ends study of a topic and triggers a scheduled review later.

Cognitive load split

Extraneous load arises from poor presentation, such as split sources or redundant text; reducing it frees capacity for the intrinsic difficulty of the material.

Time per item budget

Divide total testing minutes by item count to obtain average time per item; track cumulative pace at checkpoints and adjust early rather than at the end.

Two-pass strategy

The first pass answers only confident items and flags the rest; the second pass returns to flagged items so easy items are not lost to time pressure.

Guessing threshold

With no penalty for wrong answers, answer every item; with a penalty of one over options-minus-one, guess only when eliminated options raise the correct probability above that threshold.

Expected value of guessing

Expected score equals probability correct times the credit for a right answer minus probability wrong times the penalty for a wrong answer.

Answer change rule

Change an answer only when a specific reason supports it, such as a misread qualifier or a recalled fact; unexamined second-guessing tends to reduce accuracy.

Dimension check

Verify that units match on both sides of a calculation; a dimensional mismatch signals an algebra or conversion error before any numbers are evaluated.

Stem before options

Read the question stem and predict the answer before reading the choices, which reduces attraction to plausible but incorrect options.

Rounding discipline

Carry extra digits through all intermediate steps and round only the final answer to the precision the question explicitly requests.