Vous êtes sur la page 1sur 5

Available online at www.sciencedirect.


Learning and Individual Differences 18 (2008) 492 – 496


The role of personality and motivation in predicting early college academic

success in non-traditional students at a Hispanic-serving institution
James C. Kaufman a,⁎, Mark D. Agars a , Muriel C. Lopez-Wagner b
Learning Research Institute, California State University, San Bernardino, CA, United States
California State University, San Bernardino, CA, United States
Received 19 May 2006; received in revised form 20 November 2007; accepted 22 November 2007


Non-cognitive factors represent a chance to learn more about how to help students succeed in early college experiences. This study examined
personality and motivation as predictors of first-quarter GPA in a sample of 315 non-traditional undergraduates at a Hispanic-serving institution.
Our results provide support for the importance of high levels of conscientiousness, intrinsic motivation, and low levels of extrinsic motivation in
first-quarter school success. Implications and possible interventions are discussed.
© 2008 Elsevier Inc. All rights reserved.

Keywords: Non-cognitive; School achievement; College; Hispanic; Motivation; Personality

Almost 15 years ago, Mouw and Khanna (1993) noted that ment is conscientiousness (e.g., Higgins, Peterson, & Rihl, 2007;
there was considerable variance in college GPA that could not Noftle & Robins, 2007). Other studies have occasionally found
be explained by the traditional pre-college predictors (e.g., high evidence to support a different factor (for example, Lounsbury,
school GPA, SAT and ACT scores). Since then, Robbins et al. Welsh, & Gibson, 2005, found evidence for openness to expe-
(2004) identified three types of predictors: traditional (e.g., SAT rience), but nothing close to the strong pattern arguing for the
scores), demographic (e.g., gender), and psychosocial factors importance of conscientiousness.
(e.g., personality, motivation). Robbins, Allen, Casillas, Peter- The relationship between motivation and school success is
son, and Le (2006) found that several psychosocial factors were also well established (e.g., Pintrich & Schunk, 1996), if less
related to academic performance and retention. clearly. Several studies have looked at academic success and
The important factors, which they dub “Student Readiness” motivational orientation. Intrinsic motivation and its related
indicators, included Academic Discipline, Social Activity, goals have been found to relate to classroom success (e.g.,
Emotional Control, Commitment to College, and Social Con- Church, Elliott, & Gable, 2001). The research is not as consistent
nection. Indeed, one can argue, as did Robbins et al. (2004), that as it is with personality; for example, Baker (2004) found that
such a descriptive approach is a valuable one. Such approaches motivation was not related to academic achievement.
integrate more traditional concepts, such as the five-factor per- Studies that have examined both personality and motivation
sonality model or the intrinsic–extrinsic theory of motivation, have found highly conflicting results (e.g., McKenzie, Gow, &
allowing past research to be utilized. Schweiter, 2004; Phillips, Abraham, & Bond, 2003). In this
Indeed, there have been many studies of personality and paper, we study personality and motivation variables on a unique
academic success, mostly using the Big-Five model that argues population of non-traditional students at a Hispanic-serving
five basic factors underlie traditional personality assessment. The institution to determine if past results are replicated. Most of the
personality factor most repeatedly linked to academic achieve- previously discussed studies have been conducted on traditional
populations. Yet non-traditional students can often have dif-
⁎ Corresponding author. ferent needs, demands on their time, and reliance on different
E-mail address: jkaufman@csusb.edu (J.C. Kaufman). types of cognitive abilities (e.g., Chao & Good, 2004). Self-
1041-6080/$ - see front matter © 2008 Elsevier Inc. All rights reserved.
J.C. Kaufman et al. / Learning and Individual Differences 18 (2008) 492–496 493

beliefs in Hispanic populations have been studied (e.g., Gloria, initial weeks of the Fall quarter of 2003. Each form of
Castellanos, & Lopez, 2005); in contrast, variables such as per- communication included a description of the study and
sonality and motivation have not been studied in depth. Given information on how to participate. In order to assure
the extensive studies showing the importance of personality and confidentiality, each interested student was assigned a
motivation, this oversight should be rectified. Based on past unique password which served as their access code to the
results, we anticipate that, in replication of past work, con- on-line survey. The survey included one hundred ninety-
scientiousness and intrinsic motivation will be related to aca- three Likert-type items, and a single page of demographics.
demic achievement in a population of non-traditional students at One hundred ten items were used in the present study. The
a Hispanic-serving institution. remaining items measured similar individual academic
Experiences during the early stages of college have been characteristics, but were not germane to the current research.
found to be critical to student adjustment and, ultimately, to Participation required approximately 45 min, and each
long-term academic success (Harackiewicz, Barron, Tauer, & participant was entered into a drawing for $5.00 gift
Elliot, 2002). In fact, researchers have suggested that even the certificates. Academic information, coded by participant
first two weeks of school can have long-term impact on college password, was obtained from the University's institutional
student experiences (Woosley, 2003). This study will examine research office and combined with survey data. Personality,
the relationship of personality and motivation on first-quarter motivation, and demographic data were collected from
academic success. participants during the first several weeks of the Fall quarter.
Subsequent academic performance data (i.e., participants'
1. Method Fall Quarter GPA) were gathered at the beginning of the
Winter quarter (i.e., roughly three months after survey data
1.1. Educational institution were gathered).

The host University is mid-sized (approximately 14,000 1.4. Measures

students), and serves a student body that comprises a substantial
portion of non-traditional students. Specifically, the mean age of the 1.4.1. Motivation orientation
student body is 25.9 years. The student population is ethnically The Kaufman–Agars Motivation-orientation Scale
diverse (32.75% Hispanic, 11.74, African American, 4% Asian- (KAMS: Kaufman & Agars, in press) was used to measure
Pacific Islander), and the University is recognized by the US participant motivation. The KAMS comprises nine scenarios
Federal Department of Education as a Hispanic-serving institution. (such as choosing a college class) and 60 items. Participants
Further, the University draws from a population that is substantially respond using a 10-point Likert-type scale to indicate the
economically disadvantaged, evidenced in part by the fact that the extent to which each statement represents his or her likely
most recent data (2005–2006) reveals 57% of the student behavior or preference. The scale includes intrinsic and
population received the need-based Pell grant. In addition, more extrinsic factors, with 30 items representing each. The
than 50% of the student population is first-generation college intrinsic factor represents individual preferences for activ-
student, and nearly 40% of the student population has dependents. ities that are worthwhile, provide opportunity for growth,
and suggest that intelligence is malleable. The extrinsic
1.2. Participants factor represents individual preferences for tasks that
achievement, rewards and recognition, and the belief that
Participants (n = 315) were undergraduate students complet- intelligence is fixed. In the present study, coefficient alphas
ing their first quarter of study at a four-year university in the were .91 (intrinsic) and .93 (extrinsic).
southwestern United States. The sample was predominantly
women (245; 79%) and represented a mix of racial/ethnic back- 1.4.2. Personality
grounds including European American/Caucasian (46%), His- The five-factor model of personality was measured using the
panic American/Latino (27%), African American/Black (10%), 50-item version of the International Personality Item Pool (IPIP;
and Asian American (9%), with the remaining 8% indicating Goldberg, 1999; International Personality Item Pool, 2001). The
“other.” The mean age of participants (M = 23.53, SD = 8.55) was IPIP comprises 10 Likert-type items to measure each of five
slightly lower than the overall University population (25.9), but personality dimensions including extroversion, agreeableness,
is consistent with the mean age (23.33) of undergraduates. This conscientiousness, emotional stability, and openness to experi-
difference was expected given the sample comprises new ence. Coefficient alphas for each dimension were: extroversion
students. The mean GPA in the present sample (M = 3.11, SD = (α = .87), agreeableness (α = .76), conscientiousness (α = .78),
.66) was slightly higher than the GPA for the overall University emotional stability (α = .85), and openness to experience
population (3.01). (α = .76).

1.3. Procedure 1.4.3. Academic success

Academic success was measured using first-quarter
Participants were recruited through in-class announce- grades. The University's Institutional Research Office
ments, flyers, and an invitation sent via email during the provided academic information which was combined with
494 J.C. Kaufman et al. / Learning and Individual Differences 18 (2008) 492–496

Table 1
Means, standard deviations, and inter-correlations of all study variables
Variable Mean (SD) 1 2 3 4 5 6 7 8 9 10 11
1. GPA 3.11 (.66) na
2. HSGPA 3.31 (.47) .28 na
3. Mother education 4.51 (1.69) .16 .13 na
4. Father education 4.62 (1.77) .17 .16 .60 na
5. KAMS-I 7.50 (1.21) .08 .07 .06 .05 .91
6. KAMS-E 5.23 (1.27) −.05 .14 .04 −.03 .56 .93
7. Extraversion 3.15 (.80) .03 .10 .02 .05 .17 .10 .87
8. Agreeableness 4.09 (.55) .06 .12 .01 .10 .30 − .05 .30 .76
9. Conscientiousness 3.71 (.61) .18 .26 −.05 .07 .16 .09 .01 .22 .78
10. Emotional stability 3.20 (.78) .07 −.04 .01 .03 .04 − .13 .20 .16 .16 .85
11. Openness 3.66 (.57) .12 .08 .14 .10 .22 − .01 .31 .30 .18 .08 .76
Note: Listwise N = 315. Scale alphas are italicized and presented diagonally where appropriate.

survey data. As a result, success data were represented by for subsequent analyses. Means, standard deviations, and biva-
actual GPA values. riate correlations for all study variables are presented in Table 1.
In order to examine the relationships between personal charac-
1.4.4. Control variables teristics and first-quarter performance, hierarchical multiple
High school GPA and parental education levels were mea- regression in SPSS was used. Variables were entered in two
sured and included as control variables in subsequent analyses. steps and examined for the prediction of first-quarter GPA.
Each participant provided a self-report high school GPA (using As the primary intention of the study was to examine the
a 1.0–4.0 scale). Parental education level was measured using predictive impact of personality and motivation orientation be-
two separate single-item responses to “Mother's Education” and yond that explained by traditional measures of success, we
“Father's Education.” Responses were coded a value from one entered High School GPA and parental education level in Step 1.
to seven, with higher values indicating higher levels of edu- In Step 2, the seven individual difference variables were entered
cation (e.g., “1” = “Grade 5 or below”, “3” = “Some high school simultaneously, including intrinsic and extrinsic motivation
but didn't finish”, and “7” = “Graduate degree”. orientation and measures of the five personality dimensions
(extroversion, agreeableness, conscientiousness, emotional sta-
2. Results bility, and openness to experience). Results of the analyses are
presented in Table 2.
Data screening for missing data, univariate and multivariate The overall model was significant [F(12, 306) = 4.491,
outliers, and violations of normality (Tabachnick & Fidell, p = .001] in the prediction of first-quarter GPA. The first Step,
2001), resulted in the exclusion of 13 cases who were outliers including high school GPA and parental education level,
and were found to have used a non-valid response patterns. accounted for 10% of the variance in the school performance
Once removed, the final data set (n = 315) met the assumptions outcome [F (3,315) = 11.458, p = .001]. The addition of intrinsic
and extrinsic motivation and the big five personality factors in
Step 2 explained an additional 6% of the variance in first-
Table 2 quarter GPA. This change was also significant [Fchange (9,306) =
Regression results for predicting first-quarter GPA 2.054, p = .03]. In the final model, four variables produced
Variable B SE B Β significant coefficients. These were high school GPA (β = .24),
intrinsic motivation orientation (β = .15), extrinsic motivation
Step 1
HSGPA .34 .08 .25 orientation (β = −.16), and conscientiousness (β = .12). Complete
Mother education .03 .03 .07 results of the regression analysis are presented in Table 2.
Father education .04 .03 .11
3. Discussion
Step 2
HSGPA .34 .08 .24
Mother education .03 .03 .08 The results provide support for non-cognitive characteristics
Father education .03 .03 .09 in the prediction of early academic success in a sample drawn
KAMS-I .08 .04 .15 from a largely non-traditional college student population at a
KAMS-E − .08 .04 − .16 Hispanic-serving institution. Specifically, results provide sup-
Extraversion − .02 .05 − .01
port for the personality trait conscientiousness, and for indivi-
Agreeableness − .08 .08 − .06
Conscientiousness .12 .06 .12 dual motivation orientation, considering both intrinsic and
Emotional stability .04 .05 .05 extrinsic orientations independently. Each and all of these var-
Openness .07 .07 .06 iables offered significant prediction of first-quarter academic
Note: N = 315. Step 1 r 2 = .10 (adjusted r 2 = .09); Overall r 2 = .16 (adjusted performance, even taking into account the effects of high school
r 2 = .14). Significant Beta's are in italics ( p b .05). GPA and parental education.
J.C. Kaufman et al. / Learning and Individual Differences 18 (2008) 492–496 495

Conscientiousness, intrinsic motivation, and extrinsic high school GPA may be a less valid indicator of current abili-
motivation were all significant predictors of first-quarter ties. Some students transferred in from community college;
academic success. These results are consistent with past others graduated high school many years ago. The same factors
studies, including the finding of Robbins et al. (2006) on the that limited the relationship between high school and first-quar-
importance of academic discipline (a key component of ter college GPA, however, may have also had an effect on per-
conscientiousness) and commitment to college (a key sonality and motivation. Consequently, these results may not
component of intrinsic motivation). In addition, our work generalize to all college populations, and continued considera-
adds to the literature on the previously discussed relation- tion of non-traditional populations is warranted. Another caveat
ship between motivation and college success by providing to the study is that survey response itself may be related to
support for the importance of intrinsic motivation and for college success. Woosley (2003) found that the simple variable
the possible harm of extrinsic motivation (e.g., Cooper, of whether college first year students responded to a survey
Clasen, Silva-Jalonen, & Butler, 1999). predicted first-quarter success. It is possible that participating
The primary finding from this study is that non-cognitive students were already primed for higher levels of success. Fi-
constructs proved to be significant predictors of performance nally, the positive correlation between intrinsic and extrinsic
at a Hispanic-serving institution with a population com- motivation orientation suggests that the dimensions may not be
prised substantially of non-traditional students. The general as independent as proposed (Kaufman & Agars, in press). Des-
consistency of the current findings with past research on pite this, the regression weights for each are consistent in di-
traditional populations suggests that key variables that have rection with their respective zero-order correlation to GPA,
previously surfaced may be equally important across which provides support for their meaningful and independent
different ethnicities and populations. Indeed, one practical relationship with college performance.
implication of this study is that diverse college communities Our study adds to the growing body of research that identifies
may equally benefit from research findings conducted on individual-level variables that predict first-quarter academic
traditional populations. Although the effects of personality success. Specifically, this study demonstrates the importance of
and motivation are not large (i.e., explaining 6% of the high levels of conscientiousness, intrinsic motivation, and low
variance in GPA after high school performance and parental levels of extrinsic motivation in a non-traditional population. We
education are accounted for), they represent a meaningful see future work continuing to explore these differences at the
contribution. Even modest predictors of academic perfor- individual level – and moving beyond motivation and per-
mance are rare, and these findings encourage the continued sonality to include such variables as thinking styles, emotional
consideration of factors that may enhance the early academic intelligence, self-efficacy, creativity, and other non-cognitive
experiences of college students. variables (e.g., Kyllonen et al., 2005).
Building on these and previous results, interventions can be
developed to maximize student success. Such interventions are References
currently used with success, typically focusing on academic-
related topics such as study skills (e.g., Sayer, De Saintonge,
Baker, S. R. (2004). Intrinsic, extrinsic, and amotivational orientations: Their
Evans, & Wood, 2002). Additional programs could be tailored role in university adjustment, stress, well-being, and subsequent academic
to more individual characteristics that are associated with aca- performance. Current Psychology, 23, 189−202.
demic success. Conscientiousness, for example, was shown to Britton, B. K., & Tesser, A. (1991). Effects of time management practices on
be an important predictor in this and other studies. This factor is college grades. Journal of Educational Psychology, 83, 405−410.
Chao, R., & Good, G. E. (2004). Nontraditional students' perspectives on college
comprised of such traits as being organized and following a
education: A qualitative study. Journal of College Counseling, 7, 5−12.
schedule (Kyllonen, Walters, & Kaufman, 2005). One skill that Church, M. A., Elliot, A. J., & Gable, S. L. (2001). Perceptions of classroom
requires the same traits is time management – which has also environment, achievement goals, and achievement outcomes. Journal of
been found to be correlated with higher grades (Britton & Educational Psychology, 93, 43−54.
Tesser, 1991). A possible intervention could be to help students Cooper, B. L., Clasen, P., Silva-Jalonen, D. E., & Butler, M. C. (1999). Creative
performance on an in-basket exercise: Effects of inoculation against extrin-
develop time management skills by holding seminars on how to
sic reward. Journal of Managerial Psychology, 14, 39−57.
use the calendar function on Microsoft Outlook or by giving out Daugherty, T. K., & Lane, E. J. (1999). A longitudinal study of academic and social
free software that helps students plan out their semester. Such predictors of college attrition. Social Behavior and Personality, 27, 355−362.
interventions, which would dovetail well with pre-existing in- Gloria, A. M., Castellanos, J., & Lopez, A. G. (2005). An examination of
terventions, could lead to better time management skills and, academic nonpersistence decisions of Latino undergraduates. Hispanic
Journal of Behavioral Sciences, 27, 202−223.
perhaps, encourage conscientious behavior.
Goldberg, L. R. (1999). A broad-bandwidth, public domain, personality inventory
It is worth noting, however, that high school GPA explained measuring the lower-level facets of several five-factor models. In I. Mervielde,
less variance in college GPA in our sample than past studies have I. Deary, F. De Fruyt, & F. Ostendorf (Eds.), Personality psychology in Europe,
shown (e.g., Daugherty & Lane, 1999). One likely reason for this Vol. 7. (pp. 7−28) Tilburg, The Netherlands: Tilburg University Press.
discrepancy is that first-quarter performance represents only one Harackiewicz, J. M., Barron, K. E., Tauer, J. M., & Elliot, A. J. (2002).
Predicting success in college: A longitudinal study of achievement goals and
small sample of overall college performance, which is more
ability measures as predictors of interest and performance from freshman
susceptible to the influence of external factors than cumulative year through graduation. Journal of Educational Psychology, 94, 562−575.
college GPA after one or more years of education. Another Higgins, D. M., Peterson, J. B., & Pihl, R. O. (2007). Prefrontal cognitive ability,
possibility is that in a sample of non-traditional college students, intelligence, Big Five personality, and the prediction of advanced academic
496 J.C. Kaufman et al. / Learning and Individual Differences 18 (2008) 492–496

and workplace performance. Journal of Personality and Social Psychology, Robbins, S. B., Allen, J., Casillas, A., Peterson, C. H., & Le, H. (2006).
93, 298−319. Unraveling the different effects of motivational and skills, social, and self-
International Personality Item Pool (2001). A Scientific collaboratory for the management measures from traditional predictors of college outcomes.
development of advanced measures of personality traits and other individual Journal of Educational Psychology, 98, 598−616.
differences (http://ipip.ori.org/). Internet Web Site. Robbins, S. B., Lauver, K., Le, H., Davis, D., Langley, R., & Carlstrom, A.
Kaufman, J.C., & Agars, M.D. in press. The Kaufman–Agars motivation (2004). Do psychosocial and study skill factors predict college outcomes? A
orientation scale. unpublished manuscript. meta-analysis. Psychological Bulletin, 130, 261−288.
Kyllonen, P. C., Walters, A. M., & Kaufman, J. C. (2005). Noncognitive Sayer, M., De Saintonge, M. C., Evans, D., & Wood, D. (2002). Support for
constructs and their assessment in graduate education: A review. Educa- students with academic difficulties. Medical Education, 36, 643−650.
tional Assessment, 10, 153−184. Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics, 4th ed.
Lounsbury, J. W., Welsh, D. P., & Gibson, L. W. (2005). Broad and narrow Needham, MA: Allyn & Bacon.
personality traits in relation to cognitive ability in adolescents. Personality Woosley, S. A. (2003). How important are the first few weeks of college? The
and Individual Differences, 38, 1009−1019. long term effects of initial college experiences. College Student Journal, 37,
McKenzie, K., Gow, K., & Schweitzer, R. (2004). Exploring first-year academic 201−207.
achievement through structural equation modeling. Higher Education
Research & Development, 23, 95−112.
Mouw, J. T., & Khanna, R. K. (1993). Prediction of academic success: A review
of the literature and some recommendations. College Student Journal, 27,
Noftle, E. E., & Robins, R. W. (2007). Personality predictors of academic
outcomes: Big five correlates of GPA and SAT scores. Journal of Perso-
nality and Social Psychology, 93, 116−130.
Phillips, P., Abraham, C., & Bond, R. (2003). Personality, cognition, and
university students' examination performance. European Journal of Perso-
nality, 17, 435−448.
Pintrich, P. R., & Schunk, D. H. (1996). Motivation in education. Englewood
Cliffs, NJ: Prentice Hall.