The Scientific Attitude Inventory, TSAI, was developed in an attempt to measure the Scientific Attitude of elementary teachers. Form D, the form used with the study population, was developed through a refinement technique involving three pre-test forms, Forms A, B, and C.From an original pool of items, fifty-three items were selected to constitute Form A. Form A was constructed and administered to a pre-test population for the purpose of refinement of individual items from the item pool. On the basis of an item analysis of the responses by the pre-test population, several of the items were revised.The revised items from Form A, together with newly written items constituted Form B. Form B was administered to a second pre-test population for the purpose of refinement of individual items. Several items of Form B were revised on the basis of an item analysis.The set of items which resulted from the pre-testing of Forms A and B, revised as appropriate, were submitted to a panel of judges in order to establish a response key and content validity of the items for the purpose of measuring the Scientific Attitude.Those items which were judged to have content validity constituted Form C. Form C was administered to a third pretest population in order to identify the set of items to be used in Form D, the form used with the study population.The identification of the set of items from Form C to be used in Form D was accomplished through a series of reductions in the number of items in Form C. On the basis of an item analysis, the items with the lowest item validity indices were removed from the instrument. The resultant version was scored and an item analysis made unisg the new instrument as the criterion measure. This procedure was repeated until diminishing returns were noticed in the split-halves reliability. The 45 items of Form C which produced the highest split-halves reliability were used in Form D and constituted Version 45 C.Inspection of the 45 items of Form C which were used in Form D revealed that with but one exception, the items which produced the highest item validity indices were items keyed "disagree." Therefore, in the construction of Form D, 25 additional items were uted keyed "agree" in order to achieve an apparent balance in the response key.A 70 item instrument, Form d, was administered to the study population of 224 elementary teachers in graduate study at the masters level in the Elementary Education Department of Ball State University, Muncie, Indiana, during the Summer of 1969. The administration of Form D provided data and Norms of Performance for Form D, Version 45 C. This version of TSAI yielded a split-halves reliability of 0.72 with the study population.The complete 70 item instrument was also scored and analysed. Further, a series of reductions in the number of items, refining the form as an internal criterion measure of validity, was effected in the same manner as was used with Form C. Diminishing returns in split-halves reliability was seen beyond the 45 item set, identified as Form D, Version 45. This version yielded a split-halves reliability of 0.80. Version 65 of Form D, and each subsequent version in the reduction series yielded split-halves reliabilities equal to, or exceeding, 0.70, the minimum level of reliability specified in the design of the study. Norms of performance were compiled from the administration of Form D to the study population for each of the versions which yielded a reliability equal to, or greater than, 0.70.
| Identifer | oai:union.ndltd.org:BSU/oai:cardinalscholar.bsu.edu:handle/181826 |
| Date | January 1970 |
| Creators | Weinhold, John D. |
| Contributors | Mascho, George L. |
| Source Sets | Ball State University |
| Detected Language | English |
| Format | vi, 154 leaves ; 28 cm. |
| Source | Virtual Press |
Page generated in 0.0024 seconds