In: Statistics and Probability
second scenario: A teacher wants to use an existing test that measures student attitudes about community involvement. she wants to be sure the instrument is reliable. The instrument contains 20 statements individuals rate from 1 to 5. the test was designed to be homogeneous. use one of these strategies(test-retest, alternate-form method, or internal consistency method}, explain why the method used is appropriate, and describe the steps used to assess reliability for the particular strategy method used??
We can use the test retest method to assess reliability. This test is appropriate when we are measuing attributes that are relatively stable , i.e, do not fluctuate much over time. Student attitudes about community involvement is one such attribute. Also this test does not require much practice and is economic. The procedure for conducting this test is as follows:
1) Conduct the test on a group of students. Let these scores be denoted by X.
2) After sufficient amount of time (not too short, nor too long) , conduct the same test on the same group of students (Retest) . Let these scores be denoted by Y.
3) The reliability of the test is give by the correlation coefficient between these two set of scores. In this case the score is the number marked for each statement by each student. Suppose for statement 1, student (a) marks 1 in test and 3 in retest. Then x=1 and y=3 . This should be done for each student to get total scores for statement 1.
4) Repeat step 3 for all statements in the test.
5) This correlation coefficient can either be Karl Pearson's correlation coefficient given by
or It can be Spearman's rank correlation coefficient given by