A model to predict elementary school teachers' use of computerized educational technology to teach health and safety topics.

다국어 초록 (Multilingual Abstract)

This study developed three models that were designed to predict and explain elementary school teachers' engagement in three criterion variables. Model I was used to predict and explain engagement in the behavior of teaching health and safety topics. Model II was used to predict and explain teachers' engagement in the general behavior of using computerized educational technology (CET). And, Model III was use to predict and explain teachers' engagement in the specific behavior of using CET to teach health and safety topics. Each engagement criterion variable was considered to have two components: a self-rating of teaching behavior component and a position on the Stages-of-Change scale component. All three models hypothesized that teachers' engagement in the activity would be a function of: (a) their perceived importance of the activity; (b) their self-efficacy in performing the activity; (c) their positive beliefs and values about the activity; (d) the availability of resources for the activity; and (e) their pre-service and in-service educational backgrounds in the activity. While it was hypothesized that the availability and the condition of the resources was important for both teaching health and safety topics and for using computerized educational technology (CET), this variable was only assessed in the general use of the computerized educational technology prediction model.
A total of 204 credentialed teachers completed a specially constructed 74-item online instrument designed to validate the three prediction models. Teachers in the sample taught in all grade levels from pre-kindergarten to grade six. The majority of the participants were female (87.3%, n=178) with 19 or more years of teaching experience (31.5%, n=58) and over 50 years of age (37.3%, n=65).
Coefficient alpha measures of internal consistency for the different criterion and prediction scales ranged for .62 to .96. Multiple regression analysis was used to validate each of the models. All three models had squared multiple correlation coefficients that were statistically significant at p<.01. This was interpreted as supporting each of the prediction models. The relative importance of each of the predictor variables, implications of the research and suggestions for further research are also discussed.