| A method of program evaluation was developed which was non-preordinate, used visual observation of motor behavior as the primary means of data collection, used methods from ethology (e.g. ethogram construction, one-zero sampling, focal animal sampling) and ecology (e.g. diversity, catastrophe, stress), and was an attempted formative-outcome evaluation of an exemplar. The method used a statistic, the diversity index, never before reported in the social sciences, including evaluation. The exemplar was an evaluation of an automotive, farm and diesel mechanic training program at a community college. The report of the exemplar was well-received by the community college faculty who were able to understand the natural language varialbes, discuss the implications of the findings and implement the recommendations.;Both a two-group and a three-group comparison were performed in the exemplar. The two-group comparison was of program graduates and was between the Non-Completer (N-C) group and the Cohort group.The three-group comparison was between the cohort graduates, the current community college class labs (PC Lab) and community shop mechanics (STS). Data was divided into tool-use and non-tool use activities. The N-C versus Cohort comparison yielded significant differences on three activity varialbes (drive, watch, inspect) and no tool-use variables. The Cohort versus PC Lab versus STS comparison yielded one significant tool-use difference (ratchets) and three significant differences on activity variables (search, talk, and watch). Staffing theory (Wicker, 1973/1979) was used to explain the data and make recommendations. |
