Please enable JavaScript.

Coggle requires JavaScript to display documents.

Test Construction (6 interwined stages for valid tests (constructing the…

- - - - Expert rankings
        glasgow coma scale
        
        experts assess the situation
      - Method of equal-appearing intervals
        Example: scale of attitudes toward physical exercise
        
        large number of true-false statements reflecting attitudes
        
        represent statements to group of judgjes who sort them into 11 categories
        
        average favourability and SD are calculated for each item -> items with high SD are taken out (unreliable)
        
        20-30 items are chosen to cover the range of one dimension
      - Method of absolute scaling
        
        obtaining a measure of absolute item difficulty based on results for different groups of test takers
        
        set of common items is administered to 2+ groups -> relative difficulty of the items between the groups serves as basis for compari
      - Likert scales: 5 responses ranging from very good to very bad
      - Guttman scales: if a subject agrees with one statement, he shall also agree with a milder statement of the same underlying continuum
      - Method of empirical keying: based on empirical evidence, not judgement
        
        test items selected for a scale based entirely on how well they contrast a criterion group from a control group
      - Rational scale construction (internal consistency): all scale items correlate positively with each other and with total score for the scale
        example: personality self-report measures
        
        find items possibly related to construct in question