There are two main types of scales: rating and rankng
Rating is used to elicit responses with regard to the object,event or person studied e.g. Dichotomous scale - to elicit a Yes or No answer; Likert Scale - to examine how strongly subjects agree or disagree with statements on a 5-point scale; Stapel Scale - measures both the direction and intensity of the attitude toward the items under study
Ranking is used to make comparisons between or among objects, event or persons and elicit the preferred choices and ranking among them
e.g. paired Comparison – used to assess preferences; Forced Choice - enables respondents to rank objects relative to one another among the alternative provided;
Comparative Scale - provides a benchmark or a point of reference to assess attitudes toward the current object, event or situation under study
Validity (certainty that what is to be measured is what is measured)
Content - how well the dimensions and elements of a concept have been delineated
Criterion-Related - when the measure differentiates individuals on acriterion it is expected to predict; Construct - how well the results fit the theories around which the test is designed
Reliability (the extent to which it is without bias -error free)
Test-retest and parallel-form
Test-Retest-the correlation between results obtained at different times from same set on the same questionnaire.
The higher the better the test-retest reliability
Parallel-Form-correlation between responses on 2 comparable set that differ only in sequence or wordings
No comments:
Post a Comment