Test Theory
1
,Lecture 1 - Introduction
Psychological test
● Psychological test = a systematic procedure for comparing the behavior of two or more people
○ Behavioral samples
○ Samples must be collected in a systematic way
○ Purpose → compare behavior, identify and qualify differences
■ inter-individual
■ intra-individual
● Item scores → test scores (usually sum) → interpretation
○ item - indicator of a construct
Statistics basics - check füzet
● Z-scores (standardized)
○ mean=0
○ SD=1
○ for comparison of individual scores
● Variance-covariance matrix
○ variance- diagonal
○ covariance - off-diagonal
● Correlation matrix
○ 1 - diagonal
○ correlations - off-diagonal
2
, Lecture 2 - Properties of tests and items
Crucial properties of a psychological test
● Observable- aimed at measuring behavior (manifestation of unobservable construct)
● Objective- systematic
● Comparative- comparison of diff people / people over time
Types of tests
Maximum performance vs typical performance
● Maximum performance
○ skills / aptitude
○ 2 types
■ Power tests
● without time pressure
● nr. of correct answers
■ Speed tests
● time pressure
● similar question difficulty
● eg.: Bourdon-dot concentration test (for hiring train drivers)
● Typical performance
○ measuring characteristics
■ personality traits
■ attitudes
■ disorders
Norm-referenced vs criterion-referenced
● Norm-referenced
○ compare people to the rest of the population
○ good norm-data is crucial
● Criterion-referenced
○ compare people to an absolute standard
○ criterion people have to live up to
What do tests contain
● Test material - to give to participants
● Test forms - to register results
● Test manual - all important information
○ precise instructions
○ score-processing procedure
○ norm-tables ( → how scores should be coded/interpreted)
○ discussion of (scientific) qualities
Item scores are determined in a way that they are indicative of the constructwe want to measure
→ higher item scores =’higher’ on that attribute
3
1
,Lecture 1 - Introduction
Psychological test
● Psychological test = a systematic procedure for comparing the behavior of two or more people
○ Behavioral samples
○ Samples must be collected in a systematic way
○ Purpose → compare behavior, identify and qualify differences
■ inter-individual
■ intra-individual
● Item scores → test scores (usually sum) → interpretation
○ item - indicator of a construct
Statistics basics - check füzet
● Z-scores (standardized)
○ mean=0
○ SD=1
○ for comparison of individual scores
● Variance-covariance matrix
○ variance- diagonal
○ covariance - off-diagonal
● Correlation matrix
○ 1 - diagonal
○ correlations - off-diagonal
2
, Lecture 2 - Properties of tests and items
Crucial properties of a psychological test
● Observable- aimed at measuring behavior (manifestation of unobservable construct)
● Objective- systematic
● Comparative- comparison of diff people / people over time
Types of tests
Maximum performance vs typical performance
● Maximum performance
○ skills / aptitude
○ 2 types
■ Power tests
● without time pressure
● nr. of correct answers
■ Speed tests
● time pressure
● similar question difficulty
● eg.: Bourdon-dot concentration test (for hiring train drivers)
● Typical performance
○ measuring characteristics
■ personality traits
■ attitudes
■ disorders
Norm-referenced vs criterion-referenced
● Norm-referenced
○ compare people to the rest of the population
○ good norm-data is crucial
● Criterion-referenced
○ compare people to an absolute standard
○ criterion people have to live up to
What do tests contain
● Test material - to give to participants
● Test forms - to register results
● Test manual - all important information
○ precise instructions
○ score-processing procedure
○ norm-tables ( → how scores should be coded/interpreted)
○ discussion of (scientific) qualities
Item scores are determined in a way that they are indicative of the constructwe want to measure
→ higher item scores =’higher’ on that attribute
3
