Chapter 8 Psychological Tests and Measurement Scales Flashcards
Likert
Scale using a response format where respondents select responses to an item from an ordered range and a ranked score is given to the response
Semantic differential
Scale measuring meaning of an object for the respondent by having them place it between the extremes of several bi-polar adjectives
Thurstone
Scale in which raters assess the relative strength of each item and respondents agreeing with that item receive the average ‘scale value’ for it
Visual analogue
Scale where respondents mark their position on a line between two polar opposites and the distance of their make from one extreme is measured and becomes a score
Diagnostic item
Scale item not obviously or directly connected to the attitude objects, yet which correlates well with overall scores and therefore has discriminatory power and predictive power
Discriminatory power
Extent to which an item, or the test as a whole, separates people along the scoring dimension
Disguise
Feature of questioning approach that keeps respondents ignorant of the aims of the questioning
Factor analysis
Statistical technique, using patterns of test or sub-test correlations, that provides support for theoretical constructs by locating correlational ‘clusters’ and identifying explanatory factors
Psychometric test
Test the attempts to quantify through measurement psychological constructs such as skills, abilities, character etc.
Psychometrist/psychometrician
Person who creates and is a specialist with psychometric tests
Psychometry
The technology of test creation for the quantification of psychological constructs
Reliability
Consistency of a psychological scale
Cronbach’s alpha
A measure of scale reliability using the variance of respondents’ scores on each item in relation to the overall variance on the scale
Item analysis
Checking each item in a scale by comparing its relationship with total scores on the scale
Kuder-Richardson
A special form of Cronbach’s alpha performed on a test with dichotomous items
Split-half
Correlation between scores on two equal parts of a test
External
Consistency of a test with itself when administered more than once
Test-retest
Testing of the same group of respondents twice on separate occasions in order to estimate external reliability
Internal
Consistency between the items of a scale or test. Tendency for people to score at the same strength on similar items.
Response (acquiescence) set
Tendency for people to agree with test items as a habitual response
Scale value
On a Thurstone scale, the average of judges’ ratings of an item; respondent is given this score if they agree with it
Spearman-Brown correlation
In split-half reliability testing, provides an estimate of the true split-half reliability value from the correlation between two test halves, recognising that the raw split-half correlation is based on a set of items only half the length of the actual scale
Standardisation
Setting up of measurement norms for the populations for whom a psychometric test is intended
Test norms
Test statistics for known and identifiable groups who have taken the test. These can be used to make a fair comparison for individual test takers.
Validity
Extent to which a test measures the construct that is was intended to measure
Concurrent validity
Extent to which test results conform with those on another test assumed to measure the same construct and taken at the same time
Construct validity
Extent to which the existence of a construct is established through an interlinked set of diverse research findings. The theoretical establishment of a psychological construct through concerted and logically related psychological research.
Content validity
Extent to which test covers the whole of the relevant topic area, as assessed by experts
Criterion validity
Extent to which test scores can predict phenomena such as difference between groups
Face validity
Extent to which the purpose of a test is self-evident
Known groups
Test of criterion validity involving groups between whom scores on the test should differ
Predictive validity
Extent to which test scores can be used to make a specific prediction on some other variable
