Please enable JavaScript.
Coggle requires JavaScript to display documents.
PSY204 - Week 3 Reliability (Classical Test Theory (CCT) (Methods for…
PSY204 - Week 3 Reliability
Item Response Theory (IRT)
Treats for Internal consistency only
Classical Test Theory (CCT)
X = T + E
Observed Score True Score (T) Unsystematic Error (E)
Normal Distribution Curve
CCT assumes a normal distribution of data around the true score
Major Sources of Unrealiability
Test Contruction
Test Admininstration
Test Takers Variables
Test Scoring and Interpretation
Variance Model: Classical Test Theory can be expressed in terms of variances
Methods for Determining Reliabilty
Inter-Test Intervals
Test-Retest
Alternate Forms
Inter-Scorer Reliability
Internal Consistency
Split Half Method: Spearman-Brown
Kuder-Richardson
KR20
and KR21 (not really used)
Cronbachs Alpha
The Reliability Coefficient
What proportion is NON error
Generalisation Theory (GT)
Assesses many sources of unreliability simultaneously
ANNOVA (Analysis of Variance)
Systematic Error
Non Systematic Error
Special Issues in Reliability
Homogeneous / Heterogeneous
Dynamic or Static
Test scores restricted or not
Speed Test or Power Test
Test is Criterion Referenced or not