Development of a valid and reliable first aid and emergency care test for college and university students

This investigation developed a valid, reliable, and objective basic first aid and emergency care knowledge test for use at the college and/or university level. A list of potential subject content areas was developed from a review of the basic first aid and emergency care textbooks of the American Red Cross and the National Safety Council. The subject content areas were submitted to a national panel of experts which weighted each specific subject content area in relation to the number of test items that should be developed from each subject content area. The weightings provided by the members of the jury panel served as the table of specifications which provided direction in the development of a seventy-five multiple-choice item test. Content validity of the instrument was established through the use of the American Red Cross, and National Safety Council textbooks in the development of the preliminary item pool. A total of 225 preliminary test items were reviewed by the members of the jury panel. Each test item was rated indispensable, good, acceptable, delete or revise, or delete. Following review by the jury panel, two parallel forms of instrument were developed. The preliminary forms of the instrument were administered to one two-year community and/or junior college, and one four-year college and/or university from each of the six geographical districts of the American Alliance for Health, Physical Education, Recreation, and Dance. Preliminary Form A was administered to 156 students, and Form B was administered to 157 students. The data analysis, which included item discrimination, item difficulty, and reliability coefficient facilitated the refinement/development of the final instrument which consisted of two parallel forms, each consisting of seventy-five multiple choice items. The final forms of the instrument were administered to a sample of 927 college students. Form A was administered to 477 students, and Form B was administered to 450. The data were analyzed to determine item difficulty, item discrimination, and reliability coefficient. The Kuder-Richardson reliability was calculated to be .82 for both Form A and Form B.