The problem of this study was the lack of external validation of final examinations generated from test item banks written for Georgia Department of Technical and Adult Education (DTAE) courses. The purpose was to establish content validity for a sample 75-item final examination generated from a test item bank written for Practical Nursing course NSG 111, Nursing Process I.

Test item banks contain test items written to measure student achievement of specified competency areas. The item banks are stored on a computer, which permits final examinations to be randomly generated. Weights assigned to competency areas determine the percent of test items which are generated for each competency area.

A modified Delphi technique was used to conduct this study. Forty Practical Nursing instructors at 28 DTAE institutes rated the extent to which individual test items, groups of test items assigned to seven competency areas, and the entire set of test items measured student achievement. Each member of the panel of experts was recommended by the Vice President for Instructional Services and had been assigned to teach and develop instructional materials for the course. A consensus (80%) was reached after two Delphi rounds.

The results indicated that 71 of the 75 test items were appropriate for measuring student achievement in assigned competency areas. Seventy of the test items were rated appropriate on the first round.

None of the groups of test items assigned to measure student achievement of competency areas were rated appropriate. While individual test items were appropriate, the number of items in the group was not adequate for the competency area identified.

Likewise, 75 test items were not rated adequate to measure all competency areas of the course. Participants believed additional test items were needed.