Date of Award
Education Specialist (Ed.S)
Advanced Education Programs
The research was designed to investigate variables affecting the consistency between rater s using the Kansas Internship Assessment Instrument. The 92 subjects were volunteer teachers and administrators field testing the 1987-1988 Kansas Internship Program. The composite null hypotheses related to independent variables of gender, age, race, types of raters, size of training groups, number of sites for training, team as opposed to Individual training, team training at one and two sites, full compared to partial training and consistency of ratings due to time (practice) A statistically significant result indicated a lack of consistency in ratings; therefore, retaining a null hypothesis Indicated consistency. The null hypotheses were in two groups. Nine null hypotheses were tested employing c hi -square and one used the Z test. Twenty-eight of 40 null hypotheses were rejected. The results appeared to support the following generalizations: 1. many findings were inconclusive. 2. The type of rater (peer-supervisor) on the assessment team was not associated with the consistency of ratings: therefore, ensuring that a team composed of teachers and administrators is not necessary to have consistent ratings. 3. That if team members do not train together or at the same site, it will not affect consistency. 4. Ratings will not be affected according to whether teams train as a unit or separately. The results appeared to support the following implications: 1. the size of training groups should be kept small and definitely under 25 per group. This finding is consistent with research as quoted in the literature surveyed. 2. The results should be considered in the development of all assessment/assistance teams and the roles of each team member studied carefully. 3. As the model continues to evolve, the Kansas State Department of Education should be able to use the results to improve the training module and to help in the development of assessment and assistance teams. 4. Teacher training institutions can use these results to help develop courses for evaluators and for first year teachers. 5. Local school systems may use these results whenever an assessment/assistance team is created and consequently trained.
Copyright 1988 Herbert P. Weissenfels
Weissenfels, Herbert, "Effects of Some Variables on Inter-Rater Reliability Using the Kansas Internship Assessment Instrument" (1988). Master's Theses. 2089.