Honors College Students' Adjustment Factors and Academic Success: Advising Implications

Participants and Procedures

This study used a nonexperimental ex post facto design with student data archived between 2007 and 2010 at a single institution. The midsized, urban, research-intensive university enrolls more than 20,000 students annually. Incoming students are, on average, 18 years old with grade-point averages (GPAs) higher than 3.10 (on a 4.00 scale) and average SAT scores higher than 1030. Archival data included the Transition to College Inventory (TCI), an assessment instrument administered by the institution to first-year students in the summer prior to matriculation to identify incoming learners at risk for academic challenges (Pickering & Calliotte, 1996; Pickering, Calliotte, Macera, & Zerwas, 2005). Some TCI data were used to assess participants' expected academic, social, personal–emotional, and institutional adjustments. Additional academic data were obtained from university records.

To form comparison samples for use in the study, data from a group of 200 randomly selected first-year honors college students and a group of 200 randomly selected students outside the honors college were created using a medium effect size of 0.05 at power equal to .80. We determined that an absolute minimum of 128 students with complete scores was needed for the analysis, and we then set out to exceed the minimum requirement; on the basis of Cohen's (1988) longstanding discussion of power analysis, we selected a midlevel effect size between Cohen's recommended high and low extremes. The nonhonors student sample comprised eligible first-year students who were invited to enter the Honors College but declined to participate. We eliminated from the study a few students for whom only partial TCI or incomplete institutional data were available. Of the 393 students for which data were obtained, 197 (50.1%) were enrolled in the Honors College and 196 (49.9%) were not. Among the students making up the data for the random sample, 110 (28.0%) first enrolled in Fall 2007, 117 (29.8%) first enrolled in Fall 2008, 101 (25.7%) first enrolled in Fall 2009, and 65 (16.5%) came from the entering Fall 2010 cohort. Regarding gender, 56.7% identified as female and 43.3% identified as male; 63.1% indicated ethnicity as White, 16.5% as African American, 3.3% as Latina/o, and 3.1% responded as other. These proportions reflected the university student composition; however, institutional reporting of the aggregate institutional data did not allow us to describe the overall sample in greater detail or the two groups along the lines of common demographic variables.

Honors College Experience

The institution's Honors College was designated according to the National College Collegiate Honors Council best-practice characteristics for a fully developed honors college (National Collegiate Honors Council, 2014). The Honors College is directed by a college dean and is based on a learning-community design.

Participants in the Honors College can apply to live in honors housing. They complete a designated curriculum and take a required number of designated honors classes, which are not necessarily more difficult than regular classes on the same topic but are taught by selected mentor faculty. The program encourages Honors College students to learn material together. The Honors College students may receive small amounts of financial support for undergraduate research projects, conference travel, and other scholarly activities. Typically, the institution's Honors College participants ranked in the top 10% of their high school class and earned a high school GPA of 3.5 or higher (transfer students are expected to have at least a 3.8 college GPA).

Measures: College Adjustment Factors

Demographic and academic information was obtained from university records for students in and out of the Honors College, including fall GPA and retention status. To measure adjustment factors, we used the TCI, which is a well-established noncognitive measure, designed by Old Dominion University researchers to enhance predictions about first-year college academic performance (Pickering et al., 2005). It is used to evaluate student self-reports of their own attitudes, personality characteristics, pre-college academic behaviors, and expected collegiate performance and engagement. Details about the instrument are found in Manual for the Transition to College Inventory (Pickering et al., 2005). More specifically, the measure is used to assess reasons for attending college and for choosing a particular institution, high school senior year experiences, self-ratings of traits and abilities, attitudes about being a college student, and self-predictions of college academic success and involvement.

The TCI has been proven a reliable and valid instrument pertaining to first-year success (Pickering et al., 2005). Furthermore, the TCI has been specifically effective for assessing a student's potential risk for academic difficulty with the institutional population we studied; in other words, the measure was appropriate for the campus we targeted (McGrath & Braunstein, 1997).

The TCI features 115 self-report items. Respondents indicate the level to which they agree that the item describes their situation. Some items featured dichotomous choices such as applies or does not apply to me. Others featured Likert-type scales such as very good chance, some chance, or no chance of applying to me after matriculation such that higher scores indicate higher predicted academic success (i.e., lower expected academic risk).

Factors

An exploratory factor analysis with the principle axis method and varimax rotation (with a minimum factor loading threshold of 0.40) yielded a factorial model for explaining factors that might be related to academic performance. We were interested in eight of the instrument factors: college involvement, influences on college choice, student role commitment, athletic orientation, personal and academic concerns, self-confidence, institutional commitment, and socializing orientation. The criterion-related validity of the factors was shown, with the TCI being predictive of first-year student challenge, academic difficulty at the end of the first semester, attrition, and graduation—with increasing rates of academic difficulty and negative academic outcomes corresponding to increased risk as identified by TCI scores (Pickering et al., 2005). The definitions of the eight factors examined in this study are presented in Table 1. The data on GPAs and retention rates for both groups of participants are shown in Table 2.

Table 1 Means and standard deviations for TCI factors

View Fullscreen

Table 2 Means and standard deviations for first-semester GPA and percent retention

View Fullscreen

Between-Group Comparison of First-Year Adjustment Factors

Answers to RQ1 were used to compare first-year adjustment factors expressed by students in and out of the Honors College. A one-way multivariate analysis of variance (MANOVA) was conducted to assess group differences regarding college adjustment factors. Group membership was the independent variable and the eight TCI factors were the dependent variables. Descriptive statistics are presented in Table 1.

As a preliminary step to answering the research questions through the MANOVA data, which were based on the assumption that the two groups would show balanced, equal variances, we conducted analyses to test the homogeneity of variance (as per Mertler & Vannatta, 2005). Specifically, using Levene's test of the equality of the error variances, we found that the assumption of equality of variances was violated. Therefore, according to Tabachnick and Fidell (2007), we selected a more conservative alpha level to determine significance. When a significance value of .01 was used, Levene's test indicated that the assumption of the equality of the variable was not violated. Furthermore, for prudence, we also conducted Box's test of the equality of the covariance to additionally test for homogeneity of variance. Using Box's M assumptions, this test also confirmed that the homogeneity was not violated. The multivariate analysis of covariance (MANCOVA) indicated a statistically significant difference between the reports of students in and out of the Honors College on adjustment factors: Wilk's λ = .839, F (8, 384), p = 0.000, partial η² = .16. The effect size indicated that 16.1% of the variance in the self-reports of adjustment (Factors 1–8) could be explained by whether or not a student participates in the Honors College.

The results of the analysis of the dependent variables were taken into consideration separately using a Bonferroni-adjusted α of 0.001. Univariate statistics on data of students in and out of the Honors College showed statistically significant differences for two variables: Factors 2 and 6 (Table 3).

Table 3 TCI variables predictive of students in the honors college (n = 197) and not in the honors college (n = 196)

View Fullscreen

Factor 2 refers to the importance a student places on various external factors, people, and institutional characteristics in making decisions about college attendance and college selection. It was significant: F (1, 391) = 13.05, p = 0.000; partial η² = 0.032. The effect size indicated that 3.2% of the variance in the scores was explained by group membership (in or out of the Honors College). Therefore, generally speaking, Honors College students placed greater importance on external factors than did their peers not in the honors program when making their college-choice decisions (Table 1).

Factor 6 refers to the level of self-confidence a student expresses regarding academic and personal skills and abilities associated with college success. We also found that Factor 6 showed statistical significance: F (1, 391) = 49.86, p = 0.000; partial η² = .113. The effect size indicated that 11.3% of the variance in scores was explained by group membership (in or out of the Honors College). In an interestingly finding, Honors College students expressed less academic and personal self-confidence that did their peers outside the Honors College (Table 1).

Adjustment Factors as Predictors of Between-Group Differences in First-Semester Grades

We used responses from RQ2 to examined eight college adjustment factors as predictors of between-group differences in actual first-semester academic success. A logistic regression was used to measure the extent to which levels on the eight adjustment factors predicted differences in actual first-semester academic success, with group membership and scores on the eight adjustment factors as independent variables and first-semester GPA as the dependent variable. The full model with all predictors was statistically significant: χ² (9, n = 393) = 86.61, p < 0.001. This results indicates that the model, using the independent variables, can predict first-semester academic success.

A combination of indices of the model's overall fit was examined (as per Mertler & Vannatta, 2005). First, we estimated the proportion of variability in the dependent variable, which was taken into account by all the predictor variables in the equation (including the Cox and Snell R² and the Nagelkerke R²). The model explained between 19.8% (Cox and Snell R²) and 26.8% (Nagelkerke R²) of the variance in first-semester GPA. It also correctly classified 71.5% of cases (where success was defined as a GPA of 3.0 or higher).

Second, we used a goodness-of-fit measure to compare actual and predicted cases for the dependent variable. The Hosmer-Lemeshow goodness-of-fit test yielded a χ² value that indicated support for the model: χ² (8, N = 393) = 8.68, p = 0.37.

Finally, we used the Wald test to determine whether the predictor variables contributed significantly to the predictive ability of the model. The Wald test indicated that Factor 2, influences on college choice (p < 0.05), and group membership (p = 0.000), whether in the Honors College or not, reliably predicted differences in first-semester academic success. Summarized logistical regression data on the factors predicting academic success differences are presented in Table 4 (nonsignificant variables are not shown).

Table 4 Variables predicting between-group differences in first-semester GPA

View Fullscreen

Adjustment Factors as Predictors of Between-Group Differences in Retention

We used RQ3 to examine eight college adjustment factors as predictors of between-group differences in retention. We used another logistic regression to measure the extent to which the eight adjustment factors predicted differences in actual first-semester academic success, with group membership and scores on the eight adjustment factors as the independent variables and actual first-year spring-semester retention serving as the dependent variable. We found that the relationship pattern analyzed by the model was not a statistically significant predictor of retention differences: χ² (9, N = 393) = 11.103, p > 0.05.

Self-confidence and Influences on College Choice: Myth vs. Reality

Honors College participants in our study expressed less self-confidence in the various academic and personal skills and abilities required for college success than did their peers outside the Honors College. In contrast, they placed greater importance on external factors concerning their college choice, including input from important others such parents and family, high school counselors, and college admissions officers as well as descriptive characteristics of the institution, than did their peers. Taken together, these findings seem to provide additional support for previous research indicating a relationship between college expectations, realities, and success.

Baker and Siryk (1984) and Baker et al. (1985) first reported on a phenomenon now commonly referred to as myth versus reality pertaining to college adjustment. Specifically, they found that college and university students tend to make such unrealistically high self-evaluations and set such unreasonably high expectations for their collegiate experience at their institution of choice that they often overestimate their academic and personal abilities to adjust to the campus environment; therefore, they fail to adequately adjust and do not pursue the academic and developmental tasks needed for success. In turn, the students who expressed greater overestimation of skills and abilities and very high institutional expectations experienced more first-year challenges than those with a more realistic view of themselves and their college.

Along these lines, in a series of studies with a variety campus populations (e.g., students attending university orientation classes, mental health clients at academic risk, African American or peer mentor program participants, college women with eating concerns, and students in service learning), led by the senior researcher (Switzer) of this research, revealed that maintaining accurate self-appraisals and accurate assessments of institutional supports was a critical factor in student success (see Schwitzer, 2005, for a summary). In each case, either an intensive advising or mental health counseling support program was implemented to ameliorate successfully the risks to which the different populations were most susceptible.

In fact, Jackson, Pancer, Pratt, and Hunsberger (2006) found that learners with accurately prepared expectations tended to demonstrate better personal adjustment and academic achievement than did students with either overly optimistic or profoundly pessimistic expectations of self and the college experience. Smith and Wertlieb (2005) also reported that new students with unrealistically high academic and social–personal expectations had lower first-year GPAs than students with either average or below average expectations. Furthermore, Nadelson, Semmelroth, Martinez, Featherstone, Fuhriman, and Sell (2013) suggested that as their academic achievement increases, incoming students express more reasonable expectations for some aspects of the college experience. Consistent with these previous studies, our findings suggest that honors college students may set more realistic self-expectations for their academic adjustment, and therefore, may be better prepared to utilize various skills and abilities when appropriate and to seek supports when needed to be academically successful. Schwitzer (2005) referred to maintaining such intact self-expectations as self-cohesion and reported this self-assessment as an important buffer against adjustment difficulties (p. 32).

Our findings also suggest that honors college learners may set more realistic institutional expectations by relying on rational input from parents and other authorities more than on internal impressions because they tend to choose institutions and honors college participation by following the advice of high school counselors, using rational information from web sites, and seeking the advice of other important people in their lives (Kampfe, Chasek, & Falconer, 2016). Likewise, Hébert and McBee (2007) found that many gifted high school students seek academic environments that meet their intellectual needs. They pay particular attention to the match between their scholarly abilities and the opportunities for valuable classroom experiences and deeper learning (Nichols & Chang, 2013). Overall, our findings suggest that reliance on nuanced, complex, and accurate expectations of self and the institution contributes to the academic adjustment of honors college students to a greater extent than it does for their peers.

Implications for Advising Practice

Our findings provide evidence confirming the myth versus reality theoretical construct (Baker & Siryk, 1984). Therefore, they can inform advising students both in and out of honors colleges. As Schwitzer (2005) reported, advising and mental health counseling appear to have positive, but somewhat different, effects at the ends of the at-risk continuum.