faculty member have lower total engagement scores. The model had an adjusted R2 of .34

signifying that 34% of the variance of student engagement in the class was from these

five variables (p<.05). See Table 4-78 for the unstandardized coefficient (B), intercept

(Constant), and standardized coefficient (0).

Table 4-78. Class A Backward Stepwise Multiple Regression Explaining Student Total
 Engagement (n=56)
 Model
 Construct B SE Beta t. Sign. F Sign.
 (Constant) 5.64 9.45 0.60 .55 6.62 .00
 Sufficiency of originality
 gap -0.16 0.10 -.19 -1.57 .12
 Total Motivation 0.53 0.25 .25 2.16 .04
 Total Stress 0.27 0.08 .39 3.39 .00
 Number of similar
 courses 3.07 1.40 .26 2.20 .03
 College classification 2.88 1.10 .30 2.61 .01
Note. Adjusted R2=.34

Class B

 Considering Class B, backward stepwise regression was used to explain student

engagement with independent variables of cognitive style gap, total stress, total

motivation and student demographic variables. The best fitting model with the most

explanation of the dependent variable left two variables including rules/group conformity

cognitive style gap (P=.33) and total motivation (P=.38). The most important independent

variable in this model was total motivation, however the focus of this study was to

examine the relationship between dissimilar cognitive style and student engagement. For

students in Class B with an innovative 5-point rules/group conformity gap have an

average 1.80 points higher engagement score than students with no rules/group

conformity gap controlling for student motivation. Note that the scale used to measure of

total engagement used in this study comprised 72 points. The data suggests that students