Construction of valid tests on interpersonal relations for office occupations

The purpose of this study was to construct valid tests with equated alternate forms in the area of interpersonal relations (communicating with others) in office occupations.

An outline consisting of l00 social role criteria factors and three communication categories was constructed using various office procedures textbooks and other references. The outline was sent to a jury of Certified Professional Secretaries and teacher educators who rated each factor on its importance for successful communications in the three categories listed. The refined outline was used to construct 299 multiple choice test items. The test items were reviewed by experts to provide the "best" and "worst" answer for each item, revised, prepared in four sets of 40 items each, and administered to college students enrolled in shorthand and/or office procedures classes.

Test papers were returned from nine states for analyses and totaled l09 each for Test 1 and 3, and 123 each for Test 2 and 4. Tests were scored. Difficulty and discrimination indices were calculated for each item. Questions were then divided into Form A: Tests Al, A2, and A3, each consisting of 30 items. Further refinement of the test questions was conducted and additional testing was completed. Sixty-four test papers were returned from two additional colleges for analyses. Test Form B: Tests Bl and B2 resulted.

Major findings of the study were:

1. The criterion for difficulty was met by 93 percent of the 90 test items in Form A. All of the items in Form B met the criterion.

2. The criterion for discrimination was met by all the items in Tests Al, A2, and A3. The items in Test Bl met the criterion by 97 percent, and the items in Test B2 met the criterion by 93 percent.

3. All items in Form A met the criteria for the connnunication categories. Tests Bl and B2 exceeded the criteria for the categories of worker to superior and worker to subordinate. Test Bl had 17 percent of the test items in the category of worker to peer and was slightly below the 20 percent level preferred for this category. Test B2 met the criterion for the category of worker to peer.

4. Both Form A and Form B met the criterion for test length by l00 percent with 30 items on all forms.

5. The difference in the means of Tests Bl and B2 was not significant at the 10 percent level.

Recommendations for use of the information resulting from the study by the office occupations teacher included the following:

1. Tests Bl and B2 can be used as pre- and post-tests. They were found to be highly reliable and predictable.

2. The communication categories and the difficulty indices can be used by the office occupations teacher to construct somewhat equivalent tests to meet specific classroom needs.