SPSC Screening Test key terms

A preliminary selection assessment used to judge whether a candidate meets the minimum standard for a role or program before advancing to fuller evaluation. It usually tests aptitude, reasoning, or job-relevant knowledge under timed conditions.

The minimum score a candidate must reach to pass a screening stage and proceed. It is set in advance, often from expected vacancy numbers or a required standard, and scores below it are rejected regardless of ranking.

A standardized instrument that measures mental abilities, personality traits, or aptitudes under controlled conditions. Results are compared against a norm group so that individual scores can be interpreted consistently across candidates.

The consistency of a test in producing stable results across repeated administrations, parallel forms, or different raters. A reliable test is a precondition for valid interpretation, but consistency alone does not prove the test measures the right thing.

The degree to which a test measures what it claims to measure and supports the inferences drawn from scores. Evidence may come from content coverage, correlation with job performance, or agreement with established measures.

Evidence that a test actually reflects the underlying trait or ability it is meant to capture, such as numerical reasoning or verbal comprehension. It is supported by correlations with related measures and by absence of unexpected relationships.

The extent to which test scores predict an external outcome, such as later job performance or training success. It is usually reported as a correlation coefficient between the test and the criterion measure.

A form of criterion-related evidence in which scores collected before selection are correlated with performance measured later on the job. It is the strongest practical evidence that a screening test forecasts real outcomes.

Criterion-related evidence obtained when the test and the outcome measure are administered at roughly the same time, often to current employees. It is quicker to gather than predictive evidence but can be distorted by selection effects.

A consistency estimate obtained by administering the same test twice to the same people and correlating the two sets of scores. High values suggest scores are stable rather than reflecting chance or momentary states.

The degree to which items within a test measure the same underlying attribute and produce coherent scores. It is commonly estimated with statistics such as Cronbach's alpha or split-half correlations.

The consistency of scores assigned by different assessors judging the same performance or response. It matters most for interviews, written exercises, and other assessments where human judgment rather than a scoring key is used.

The reference sample whose scores define how a test result is interpreted. A candidate's raw score is converted into a relative standing by comparison with this group's distribution.

Scoring that interprets a candidate's result relative to the performance of others rather than against an absolute standard. Selection decisions then depend on rank order within the candidate pool.

Scoring that compares performance against a fixed standard of competence rather than against other candidates. A candidate passes by demonstrating required knowledge or skill, regardless of how peers performed.

The percentage of people in a norm group who scored at or below a given score. A percentile of 80 means the candidate outperformed about 80 percent of the comparison group.

A measure of how widely scores spread around the average. In screening, it helps interpret whether a difference between two scores is meaningful or within normal fluctuation.

A symmetrical, bell-shaped spread of scores in which most results cluster near the mean and become rarer toward the extremes. Many ability tests are designed so that scores approximate this shape in a representative population.

Administering a test under uniform conditions, with fixed instructions, timing, and scoring rules, and establishing norms from a representative sample. It allows results from different candidates and sittings to be compared fairly.

The controlled process of delivering a test, including reading instructions, managing time, securing materials, and handling irregularities. Consistent administration protects fairness and preserves the reliability of scores.

Supervision of candidates during a test to prevent cheating and ensure rules are followed. Invigilators verify identity, control access to materials and devices, and record any irregularities observed.

Measures that protect test content and results from unauthorized access, leakage, or tampering. It covers item storage, controlled distribution, secure platforms, and confidentiality of candidate data.

Deliberate rule-breaking during an assessment, such as impersonation, copying answers, using prohibited devices, or leaking items. It typically leads to score cancellation and may bar the candidate from future sittings.

A stored, organized collection of test questions with metadata such as difficulty, topic, and usage history. Items are drawn from it to build forms while controlling exposure and maintaining comparable difficulty across sittings.

Statistical review of individual questions to assess difficulty and how well each discriminates between stronger and weaker candidates. Poor items may be revised or removed to protect test quality.

The proportion of candidates who answer a question correctly, often called the facility value. Very easy or very hard items provide little information for distinguishing among candidates.

A statistic showing how well a question separates high-scoring candidates from low-scoring ones. Items that strong candidates miss more often than weak candidates are usually flawed and need review.

A statistical procedure that adjusts scores from different test forms or sittings so they are comparable in difficulty. It prevents candidates from being advantaged or disadvantaged by which version they sat.

The unadjusted count of correct answers or points earned before any conversion. It is usually transformed into a scaled score, percentile, or pass decision for interpretation.

A converted score placed on a common reporting scale so results from different forms or sections can be compared. Scaling does not change the underlying performance, only how it is expressed.

The unavoidable imprecision in any score, arising from factors such as guessing, fatigue, or item sampling. Confidence intervals around scores acknowledge that an observed result is an estimate, not an exact truth.

An estimate of how much a candidate's observed score is likely to differ from their true score because of measurement error. It is used to build confidence bands around reported results.

An assessment of reasoning skills such as verbal, numerical, and abstract thinking, often used to predict learning speed and job performance. It is typically timed and scored objectively.

The ability to interpret and manipulate numbers, tables, and graphs to reach correct conclusions under time pressure. Questions usually require calculating rates, percentages, or comparisons rather than recalling formulas.

The ability to understand written passages and judge whether statements are logically supported by them. Correct answers follow only from the given text, not from outside knowledge or personal opinion.

The ability to detect patterns and rules in shapes, symbols, or sequences without relying on language or numbers. It is often used as a culture-reduced measure of general reasoning.

The ability to draw valid conclusions from premises, identify assumptions, and spot flawed arguments. Questions may ask what must be true, what weakens a claim, or what follows necessarily.

Evaluating the strength and structure of arguments, including assumptions, evidence, and conclusions. It requires separating claims that are supported from those that merely sound persuasive.

A test presenting realistic work scenarios where candidates choose or rank the most and least effective responses. It assesses judgement and behavioural fit rather than raw knowledge, and scoring reflects expert consensus.

A defined pattern of behaviour linked to effective performance, such as teamwork, resilience, or communication. Screening may assess these through situational judgement, interviews, or biographical questions.

An interview using predetermined questions and scoring criteria applied consistently to every candidate. It improves comparability and reduces bias compared with free-flowing conversation.

A multi-exercise selection process combining tests, group activities, presentations, and interviews, observed by trained assessors. Judgements are pooled against defined competencies to produce an overall recommendation.