23

Space (took class online)

.746

.371

.071 2.009 .

045

Location (took class at JALC)

-.314

.254

-.045 -1.237

.

217

F = 74.775, p <.01; Adjusted R

2

= .422.

Table 4 presents findings from the multiple regression analysis of the post confidence in

government scores. The adjusted R

2

of .422 indicates that this model explains about 42% of the

variance in the post internal political efficacy. Furthermore, the probability of the F value of

74.775 for this model is less than the alpha level of .01. Therefore, the null hypothesis that the

combined effects of prior confidence in government, cognitive pre score, time, space, and

location is equal to zero must be rejected. A significant amount of variation in post confidence

in government is explained by this model.

Table 4 further illustrates that post confidence in government will increase .635 for each

unit change in prior confidence in government, holding constant all other variables in this model.

Upon examining the significance associated with obtaining the t value (18.895) for this

coefficient, the null hypothesis that the effect of prior confidence in government on post

confidence in government is equal to zero must be rejected. Time seems to be a valid influence

on post confidence in government since the significance associated with obtaining the t value

(2.978) for this coefficient is significant at .000. Therefore, the null hypothesis that the effect of

when the course is taken on post confidence in government is equal to zero must be rejected.

More specifically, when a student takes American Government 101 in the Fall semester post

confidence in government increases by .678. Additionally, when the student is in an online