PSYC 3250 Chapter Notes - Chapter 11: Cultural Bias, Takers, Factor Analysis
Document Summary
Measure a wider and richer array of human abiliies and skills than those of muliple-choice tests: muliple-choice tests. Are assumed to measure narrow facets of performance: research doesn"t support that constructed responses provide more informaion than muliple-choice. In fact muliple-choice is believed to be the beter opion. Diicult to have reliable scoring for constructed responses. Any method for scoring constructed responses is likely, laborious, expensive and slow. Muliple-choice are easy to develop and score, and can easily be made long and detailed. Item wriing: must generate an item pool, common issues. Item length: examinees might get confused and miss the point. Draw on a theory and then translate it into test items. Problem is that items are oten transparent: atheoreical approach. Used to see what correlates highest with 1 of 2 disinct groups. Are much less transparent as they oten hold no theoreical signiicance.