only 41.06% of the total respondents were seniors. It may be that freshman, sophomore

and junior students lowered student engagement mean scores in each of the classes since

college freshman tend to have lower levels of engagement than college seniors (Kuh et

al.). A one-way ANOVA was conducted to examine if significant differences of student

engagement scores existed between college freshman, sophomores, juniors and seniors.

Significant differences were found among the four groups of students (F=2.75, p=.04).

However the assumption of homogeneity of variance was not met as Levene's Test of

Equality of Error Variances indicated to reject the null hypothesis, (F=3.00, p=.03) and

conclude that the four groups of students were not equal regarding variance. Because

variances among the four groups of students were not homogenous and sample sizes

differed among the four groups, a Games-Howell post-hoc test was conducted to account

for these differences. In this analysis, college senior student engagement scores were not

significantly different from junior student engagement scores (p=.07), sophomore student

engagement scores (p=.99) or freshman student engagement scores (p=.87). The

researcher concluded that low scores of student engagement found in this study were not

attributed to college classification.

Summary of Findings for Objective Two

 In determining students' cognitive style seven classes had total score means that

were slightly adaptive (below the mean of 95), one class had total cognitive style score

means that were slightly innovative (above the mean of 95), and one class had a cognitive

style score approximating 95. Although self-selected courses may have a tendency to be

more adaptive or more innovative, the total mean cognitive style scores of these classes

were indicative of the percentage of females in the course. Kirton (1999) states that the

total cognitive style mean score for females is 90, which is five points lower than the