ØEmpirical strategies
- ØRely on data collection & statistical analyses
- Øto determine the meaning of a test response
- ØOr the nature of personality and psychopathology.
- ØLike deductive strategies- self report feature – item responses were their own views, opinions, and feelings.
- Øuse experimental research to determine empirically the meaning of a test response.
- ØIt comprise
- A.The criterion-group
- B.The factor analysis method.
The Criterion-Group Strategy
A basis for comparison= criteria (Criterion)
Contrasted group method/ external strategy / Empirical strategy /criterion keying method.
test designers choose items to distinguish a group of individuals with certain characteristics, the criterion group, from a control group;
Test constructors select and administer a group of items to all the people in this criterion group as well as to a control group.
Constructors then attempt to locate items that distinguish the criterion and control groups, or how the two groups contrast.
the approach attempts to determine which items discriminate the criterion and control groups.
Little importance to content or face validity.
The content of the items is of little consequence. What matters is that aggressive individuals marked “True” to these items, thereby discriminating the aggressive individuals from the control group.
1st determine the item of ‘bewade log’ (criterion group) then cross validation of scale by checking how well it distinguishes an independent criterion sample.
After cross validation develop norms to obtain Z.
After a scale has been constructed and cross-validated, the third step in the criterion approach is to conduct additional research to ascertain empirically.