Masters Theses

Date of Award

8-1976

Degree Type

Thesis

Degree Name

Master of Science

Major

Agricultural Extension

Major Professor

Cecil E. Carter Jr.

Committee Members

Robert S. Dotson, John Brower

Abstract

This analytical study was concerned with determining the reliability of data received from the Tennessee Extension Management Information System, particularly weekly activity reports. The study was divided into the following three problem areas: 1) Extension Leaders' consistency in recoding three Extension activities for weekly activity reports, 2) Extension Leaders' knowledge of TEMIS codes and the extent to which they could select correct codes for reporting hypothetical activities, and 3) the extent Extension Leaders correctly coded selected inservice training meetings on their weekly activity reports. For Problem Areas I and II data were secured from interviews with 30 Tennessee Extension Leaders located across the state. Data for Problem Area III were obtained from the state office files containing the weekly activity reports of the 54 leaders attending the beef inservice training meetings and the 55 leaders attending the swine inservice training meetings. For Problem Areas I and II, the frequency of consistency of the leaders' coding in each weekly activity report field was computed. The findings were shown in numbers and proportions. A coefficient of agreement was computed for each Extension Leader and for each weekly activity report field. For Problem Area III, the data were analyzed by counting the frequency of use for each TEMIS code and presenting the findings in numbers and percentages. Major findings of the study are briefly stated as follows. In recoding three activities, the leaders were highly consistent in weekly activity report fields, audience and personal location, each having 94 per cent of the codes consistent. Low coding consistency was shown in number in audience and time expended fields, having 63 per cent and 69 per cent respectively of the entries coded consistent. Both of these fields required actual numbers to be reported rather than TEMIS codes. Overall the Extension Leaders showed a moderate level of coding consistency having a 0.68 coefficient of agreement. Leaders were more consistent in receding the five required weekly activity report fields as shown by a 0.75 coefficient of agreement. The coefficient of agreement was 0.58 for the three optional fields. Extension Leaders under reported numbers recorded in the number in audience. There was a net decrease of 19 persons reported and in time expended a net decrease of 21 hours was reported on the receding of three activities. The Extension Leaders showed very high levels of coding correctness in reporting two hypothetical activities. Weekly activity report field number in audience was reported 100 per cent correct. Audience and time expended fields each had 98 per cent of the codes correctly reported. Lowest coding correctness was in the subject field having 51 per cent of the codes correctly reported. When all the fields were considered coding correctness for the hypothetical activities was very high shown by a 0.89 coefficient of agreement. Coding correctness remained high in the required fields, 0.89, and in the optional fields, 0.88. The leaders demonstrated that they could correctly code hypothetical activities when given adequate information and instructions (i.e. the reporting system can produce very consistent data). The numerous codes used by the leaders in each weekly activity report field when reporting inservice training meetings (i.e. 9 purpose codes, 23 subject codes, 8 audience codes, 3 personal location codes, and 10 teaching method codes) showed the various individual interpretations made by the leaders. The use of three purpose codes and five subject codes inappropriate for reporting inservice training meetings which pertain to a specific program area revealed a lack of knowledge concerning the proper codes to use in reporting inservice training meetings. In the weekly activity report fields, audience, personal location, and teaching method, which have set codes for reporting inservice training meetings, the use of seven audience codes, one personal location code, and nine teaching method codes which were incorrect revealed a lack of knowledge of TEMIS codes for reporting inservice training meetings. The personal location field had the most entries correctly receded (2.60 entries per leader). The audience field had only 1.23 entries per leader correctly coded. Recommendations were also included.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS