Please enable JavaScript.
Coggle requires JavaScript to display documents.
Psychometric Properties of Assessment - Coggle Diagram
Psychometric Properties of Assessment
Direct observation approaches
Behaviour you are attempting to explain: specific, clearly defined, observable, measurable, reliable definition
what do you want to record?
Prospective/retrospective?
diary/questionnaire/video/audio?
who records? where does the recording take place?
Consequences - what reductions occured? What stopped? What increases occured?
Multiple baseline methods of tracking a problem
Reliability
Test-retest reliability
: does a subsequent completion result in the same score if all other variables stay the same?
Item-total correlations
: do the items on each scale correlate with the total score for each scale?
Intraclass correlation:
how strongly are scores related within different groups of people?
Internal consistency (Cronbach's Alpha)
: how closely a related a set of items are as a group - acceptable range is above 0.70
Validity
Construct validity
Convergent validity
: the tool correlates well with tools that measure the same/similar constructs
Divergent validity
: the tool correlates poorly with tools that measure unrelated constructs
Criterion Validity:
do scores correlate with an external 'gold standard' criterion eg criminal charges
Factor structure
Factorial analysis
Exploratory Factor Analysis:
focus on establishing an underlying factor structure
Confirmatory Factor Analysis
: used to verity a hypothesised factor structure
Testing Fit: how well the proposed model accounts for the correlation of all items
Root Mean Square Error of Approximation
Comparative Fit Index
Chi-square
Fit of the model
Standardised root mean square residual (SRMR)
Factorial invariance
: a good measure will show similar factor structures across different populations - do the psychometric properties stay the same across different groups