|Asia-Pacific Forum on Science Learning and Teaching, Volume 8, Issue 1, Article
13 (June, 2007)
Meral HAKVERDI, Berna GÜCÜM & Hünkar KORKMAZ
Factors Influencing Pre-service Science Teachers’ Perception of Computer Self-efficacy
A multiple regression analysis was conducted to examine the degree of association between the outcome variables (perception of personal self-efficacy in teaching with computers; perceptions of outcome expectancy) and the explanatory variables ( personal use of computers, educational use of computers, level of computer use, grade level, school year, number of computer-related courses taken, age and gender). Two regression models were tested to investigate the influence of explanatory variables on each of the outcome variables. Analyses were performed by use of SPSS REGRESSION. Results of the evaluation of the assumption for linear regression analysis led to deletion of the variable “school year” to reduce the multicollinearity. Twenty-seven cases with missing data were deleted from the regression analysis, n=278 for each analysis.
The first regression model consisted of eight explanatory variables and the outcome variable “perceptions of outcome expectancy”. Results showed that R2 was not statistically significant at the 0.05 level. The second regression model consisted of eight explanatory variables and the outcome variable “perception of personal self-efficacy in teaching with computers”. Results showed that R2 was statistically significant, F (8, 278) =23,844, p = .000. This model indicates that the explanatory variables are jointly associated with 41% of the pre-service science teachers’ perception of personal self-efficacy in teaching with computers.
The regression formula for the second research question is as follows:
Y = α + ß1X1 + ß2X2 + ß3X3 + ß4X4 + ß5X5 + ß6X6 + ß7X7 + ß8X8 + ß9X9 + ε
When the eight explanatory variables are placed in the regression model, the following formula results:
Perception of personal self-efficacy in teaching with computers = α (constant) + ß1*(outcome expectancy) + ß2*(personal use of computers) + ß3*(educational use of computers) + ß4*(level of computer use) + ß5*(grade level) + ß6*(number of computer- related courses taken) + ß7*(age) + ß8*(gender) + ε (error).
Table 6 shows the unstandardized regression coefficient (b), the standardized regression coefficient (ß), and the observed t-values (t). Two of the eight variables were statistically significant at 0.05 levels: educational use of computers and the level of computer use.
Table 6 indicates that level of computer use and educational use of computers are highly related to the outcome measure of pre-service science teachers' personal self-efficacy in teaching with computers (p < .000, and p < .006 respectively). In this regression equation, no other variable was significant at the p < .05 level. This observation is interpreted to mean that as the perception of pre-service science teachers’ level of computer use decreased, it is likely that personal self-efficacy in teaching with computers increased.
The final regression equation, built from information in the Table 6, is as below:
Perception of personal self-efficacy in teaching with computers = 49.797 +
(1.561E-03)*(outcome expectancy) + 0.416*(personal use of computers) +
(-0.856)*(educational use of computers) + (-6.539)*(level of computer use) +
(-1.034)*(grade level) + 0.605*(number of computer-related courses taken) + 0.402*(age) + (-0.427)*(gender) + ε (error).
Table 6. Regression Analysis Summary for Pre-service Science Teachers' Personal Self-efficacy in Teaching with Computers
Personal use of computers
Educational use of computers
Level of computer use
Numbers of computer- related courses
Note: R2 = .407 (n= 278, p = .000)
Copyright (C) 2007 HKIEd APFSLT. Volume 8, Issue 1, Article 13 (June, 2007). All Rights Reserved.