Measurement Reliability and Validity psych 209

Reliability

the consistency or stability of a measure

True score

the actual value of a variable without error

Measurement error

the difference between observed value and true score

Pearson product moment correlation coefficient

measures the strength and direction of the linear relationship between two continuous variables. It ranges from +1 to -1 where

+1 indicates a perfect positive linear relationship, -1 a perfect negative relationship, and 0 no linear correlation.

Test retest reliability

measures the stability and consistency of a test or instrument over time by administering it twice to the same group and calculating the correlation between scores

Alternate forms reliability

measures consistency by administering two different but equivalent versions of a test (Form A and Form B) to the same group

Internal consistency reliability

How well different items on a test or a survey that are intended to measure the same construct produce similar results

Split half reliability

How consistent different sections of the

same test are . . . Do different sections or halves of the same test

yield similar scores?...If a single test measuring one construct is

split into parts, then participants should score similarly on each

part.

Cronbachs alpha

measures internal consistency—how closely related a set of survey questions or test items are as a group. It determines if questions meant to measure the same concept

Item total correlations

tool measuring the relationship between a single survey question (item) and the overall test score

Interrater reliability

agreement between different observers

Reliability versus accuracy

reliability is consistency while accuracy refers to correctness

Face validity

subjective, superficial assessment of whether a measurement tool or test appears to measure what it claims to measure "at face value". It is used as a quick, initial check to ensure relevance

Content validity

extent to which a measurement tool (like a test or survey) represents all facets of a given construct, ensuring it comprehensively covers the subject matter. does it truly measure the concept its supposed to

Predictive validity

extent to which a score on a scale or test predicts future performance or behavior, determined by a high correlation between the test results

Concurrent validity

a measure of how well a new test or assessment compares to a previously validated, "gold standard" measure administered at the same time. It is a form of criterion validity used to ensure a new method produces similar results to established ones

Convergent validity

a subtype of construct validity that assesses whether tests or measures that are supposed to measure the same construct are, in fact, strongly correlated. test is similar to other tests and measures the same thing

Discriminant validity

ensures that a test or measure intended for a specific construct does not falsely correlate with measures of unrelated or distinct constructs.Confirming that a tool measures a unique concept rather than overlapping with others

Reactivity

when awareness of being measured changes behavior

Nonreactive

measurement that does not influence behavior

Nominal scales

used to classify or label data into distinct, unordered categories (gender, eye color)

Ordinal scales

ranks data in a specific, meaningful order, but does not define the exact numerical distance between the values (dislike-like)

Interval scales

quantitative measurement scale that features ordered values, consistent and equal distances between adjacent points, it lacks a "true zero (temp)

Ratio scales

a quantitative measurement scale featuring ordered, equally spaced intervals and a true, meaningful zero point (weight)