Retention implications of a relationship between age and GPA

Retention implications of a relationship between age and GPA

T. Ross Owen

The purpose of this study was to identify the extent to which age and GPA are related among 158 students attending community college. The resulting positive correlation was .33 and statistically significant at the .0001 alpha level. It was concluded that a significant positive relationship exists between age and GPA. Retention implications for a convenience sample are discussed according to a Coordinated Program Studies Model.

**********

Although some students enter college around age 18, others choose to enter or return in their 30s and beyond (Haggen, 2000). Since 1970, the number of students over 25 years of age entering American higher education institutions has increased from 28% to 44% (U.S. Department of Education, 1996). The postponement of one’s collegiate experience warrants an awareness of the age-related characteristics of students. Middle-aged females represent the fastest growing segment of the American post-secondary student population (National Center for Educational Statistics, 1995). Johnson, Schwartz, and Bower (2000) found age-related health concerns to be important stress factors among community college women. Moreover, “adult women would benefit from interventions by college staff and faculty members to help them manage these stresses and to increase their chances for successful completion of their programs” (p. 294).

Community college students are becoming increasingly diverse. According to Bishop-Clark & Lynch, “age hetero-geneity creates a unique atmosphere for learning” (1998, p. 21). Their study concluded that community college students and faculty prefer mixed-age classroom experiences. Gustentine & Keim (1996) found significant age differences among community college students on learning style. Traditionally aged students processed information through reflective observation, while nontraditionally aged students processed information through active experimentation. Age was found by Gonzenback (1993) to be a factor affecting community college students’ decisions to continue their education beyond the associates’ degree. Chi-square test of independence showed no dependence on GPA.

“Grade point average (GPA) is one widely accepted means of determining academic success and the degree to which students have learned what they are expected to learn” (McAloon, 1994, p. 13). In a survey of university students who began their academic career on a community college campus, Carlin (2001) found that transfer students did not differ significantly in GPA from students who began freshman studies at a major university. Carlin concluded that “community colleges are a good investment for most adult students who wish to pursue a baccalaureate degree” (2001, p. 169). Alcohol-related academic problems tend to be less apparent at non-residential community colleges than on residential university campuses (McAloon, 1994). Data further revealed that the more frequently a community college student reported drinking, the more likely they were to also report a lower GPA.

The characteristics of age and GPA exist to some extent among all students. The purpose of this study was to identify the extent to which age and GPA are related among community college students. Data concerning the relationship between age and GPA are sparse. This may be the first study to investigate the relationship between age and GPA among community college students.

Design

This study investigates the extent to which the variables of age and GPA are related among 158 students attending community college. According to Ary, Jacobs, and Razavieh, “[t]he possibility of the existence of relationships between variables is a reasonable question to investigate in educational research” (1985, p. 329). Due to its focus on relationship, this study is correlational in nature. “Correlational research methods help to clarify relationships and patterns of relationships among variables” (Ary, Jacobs & Razavieh, 1996, p.390).

Sample

The convenience sample for this study consisted of 158 students attending Haywood Community College (n=51) and Southwestern Community College (n= 107) during Winter 1996 (n=34), Spring 1997 (n=51), Fall 1997 (n=24), Spring 1998 (n= 18), Fall 1998 (n=23), and Spring 1999 (n=8). Included in the sample were day (n=83) and evening (n=75) students enrolled in ACA 111 College Student Success (n=18), ACA 118 College Study Skills (n=5), ENG 0085 Sentence Skills (n=3), ENG 1100 Vocational Communication Skills (n=7), MAT 060 Essential Mathematics (n=50), MAT 1125 Vocational Mathematics (n=18), and PSY 109 Human Relations (n=57).

Student-disclosed demographic data were confirmed by the registrar’s office. Demographic data included age, race, gender, and GPA. Table 1 illustrates age frequencies and percentages. Forty-eight (30.38%) students were teenagers and 26 (16.46%) were 41 years of age or older. Table 2 illustrates race frequencies and percentages. The majority of students (89.24%) were Caucasian at a frequency of 141. Table 3 illustrates gender frequencies and percentages. Sixty (37.97%) were male and 98 (62.03%) were female. Table 4 illustrates GPA frequencies and percentages. Twenty-eight (17.72%) students had GPAs of 1.96 or below. Twenty-three (14.56%) students had 4.0 GPAs.

Results

The most commonly used descriptive statistic of correlation is the Pearson product moment correlation coefficient. Because it is most appropriately used with interval data, the Pearson product moment correlation coefficient was the statistical test used to address the research question: What is the relationship between age and GPA?

The resulting correlation was +.33 and statistically significant at the .0001 alpha level. The probability of the relationship between age and GPA being due to chance is one in 10,000 or less. It can therefore be concluded that among this sample of community college students, a significant positive relationship exists between age and GPA.

Discussion

Pearson product moment correlation coefficient is a statistical test that illustrates both the direction and strength of a relationship. A correlation coefficient of -1.00 is evidence that a perfect negative relationship exists. A correlation coefficient of +1.00 is evidence of a perfect positive relationship. And a correlation coefficient of zero would suggest that age and GPA are completely independent of one another, or totally unrelated. Such exact relationships between variables are rarely or never calculated. In this case, a positive correlation coefficient of .33 means that in general, age increases as GPA increases, or similarly, age decreases as GPA decreases.

A correlation coefficient’s strength may be referred to as either significant or insignificant depending upon the level at which a researcher chooses to make probable explanations. An alpha level of .0001 allows for predictions of 99.999% accuracy. Thus, a probable or significant correlation coefficient of +.33 enables an extremely accurate prediction about age on the basis of GPA, or GPA on the basis of age. Significance should not be confused with numeric value. A low correlation of .33 represents a low positive relationship on a scale from -1.00 to +1.00. However, a correlation of +.33 does suggest that a weak relationship between age and GPA can be predicted with 99.999% accuracy.

Implications

A correlation coefficient is not a proportion. It is the square root of a proportion expressed in the form of a fraction. Thus .33 does not indicate the percentage of a perfect relationship between age and GPA. One way to determine the percentage or proportion of a relationship between age and GPA is to calculate a coefficient of determination. A coefficient of determination is calculated by squaring a correlation coefficient. That is, it is the percentage of variance in age that is determined by the variance in GPA, or vice versa. A coefficient of determination is always positive and varies from 0 to 1.

In this case a coefficient of determination may be more informative than a correlation coefficient. For instance, a correlation coefficient of .33 yields a coefficient of determination of 11%. Thus, 11% of the variance in GPA is predictable by the variance, or dispersion in age. Dispersed scores are characteristic of a heterogeneous group. Only 11% of the total variation between GPA and age is accounted for. The remaining proportion (89%) must be attributed to something besides age. In other words, there are additional variables responsible for determining the characteristics that accompany age and GPA that prevent segregating students only according to age for the purpose of predicting academic success. Additional variables include financial assistance (Makuakane-Drechsel & Hagedom, 2000); dispositional variables (Armstrong, 2000); goals and intentions (Borglum & Kubala, 2000; Mason, 1998); assertiveness (Haggan, 2000); special attention and class size (Waycaster, 2001); instructor characteristics (Armstrong, 2000); and crisis intervention (Gilliland & James, 1997). Family background, individual attributes, and pre-college schooling are also accurate predictors of academic success (Tinto, 1975, 1987, 1993).

Indeed, a positive yet weak relationship between age and GPA suggests the need for a comprehensive student retention model designed to enhance the academic experiences of young, poorly performing community college students. Tinto, Russo, & Kadel (1994) have developed a Coordinated Studies Program (CSP) model that increases student learning and retention by structuring college programs in a way that stresses the importance of community. Their CSP model demonstrates that “despite the many obstacles, community colleges can successfully involve students in education, thus enhancing their learning and increasing their persistence” (Tinto et al., 1994, p. 27).

CSPs are theme-based and team-taught by two to four instructors from contrasting yet complementary disciplines. Students enroll in separate courses (i.e., art, political science, English, and sociology), but they attend CSPs as one course and as members of a collaborative learning community. Multiple instructors facilitate cross-disciplined learning activities using lectures, guest speakers, films, small and large group discussions, seminars, and field trips. CSP assignments include readings, papers, group projects and presentations, quizzes, and exams. Attendance and participation provides students with a distinctively different learning experience characterized by in-depth exploration, analysis, and synthesis of multidisciplinary concepts.

An in-depth look at the Coordinated Studies Program at Seattle Central Community College revealed a number of positive effects among young, full-time, day students (Tinto et al., 1994). Participating CSP students reported being significantly more involved than non-CSP students in course-related activities, activities with other students, library resources, with faculty, and in arts activities on campus. CSP students were more socially and academically involved in college life and more positive in their views of the institution and their involvement in college. Over the course of a year, CSP students made greater academic gains than did their non-CSP peers. “And perhaps most importantly, independent of individual attributes, students were more likely to stay in school” (Tinto et al., 1994, p. 29).

Tinto et al. (1994) concluded that community colleges can promote student learning and retention among young, poorly performing students using CSPs as collaborative learning environments. “Equally important, [collaborative learning] appears to work as well for community college students with substantial remedial needs” (Tinto et al., 1994, p. 29). It is not surprising that the predictive utility of a relationship between age and GPA is especially important when predictions are used to make important decisions about individuals, such as placement in a remedial course. “What is surprising is that it has taken so long to rediscover the importance of community in college and its impact upon student learning” (Tinto et al., 1996, p. 29).

Recommendation

A limitation of this study is that due to the absence of random assignment, sample results are not generalizable to a larger population. Thus it is recommended that the relationship between age and GPA be studied using an experimental design with the potential to identify a generalizable causal relationship. A causal relationship should not be confused with a predictive relationship. A causal relationship would imply that GPA varies according to the influence of age.

Table 1

Age Frequencies and Percentages

Cumulative Cumulative

Age Frequency Percent Frequency Percent

16 1 0.63 1 0.63

17 3 1.90 4 2.53

18 20 12.66 24 15.19

19 24 15.19 48 30.38

20 8 5.06 56 35.44

21 9 5.70 65 41.14

22 9 5.70 74 46.84

23 5 3.16 79 50.00

24 4 2.53 83 52.53

25 7 4.43 90 56.96

26 3 1.90 93 58.86

27 6 3.80 99 62.66

28 4 2.53 103 65.19

29 7 4.43 110 69.62

30 3 1.90 113 71.52

31 2 1.27 115 72.78

32 5 3.16 120 75.95

33 1 0.63 121 76.58

34 2 1.27 123 77.85

35 2 1.27 125 79.11

36 1 0.63 126 79.75

37 3 1.90 129 81.65

38 1 0.63 130 82.28

39 2 1.27 132 83.54

41 3 1.90 135 85.44

42 6 3.80 141 89.24

44 1 0.63 142 89.87

45 2 1.27 144 91.14

46 3 1.90 147 93.04

49 1 0.63 148 93.67

50 2 1.27 150 94.94

51 2 1.27 152 96.20

54 2 1.27 154 97.47

57 1 0.63 155 98.10

60 1 0.63 156 98.73

62 1 0.63 157 99.37

71 1 0.63 158 100.00

Table 2

Race Frequencies and Percentages

Cumulative Cumulative

Race Frequency Percent Frequency Percent

Caucasian 141 89.24 141 89.24

African

American 1 0.63 142 89.87

Native

American 15 9.49 157 99.37

Hispanic 1 0.63 158 100.00

Table 3

Gender Frequencies and Percentages

Cumulative Cumulative

Gender Frequency Percent Frequency Percent

Male 60 37.97 60 37.97

Female 98 62.03 158 100.00

Table 4

GPA Frequencies and Percentages

Cumulative Cumulative

GPA Frequency Percent Frequency Percent

0.00 4 2.53 4 2.53

0.50 1 0.63 5 3.16

0.75 1 0.63 6 3.80

0.82 1 0.63 7 4.43

0.93 1 0.63 8 5.06

1.10 1 0.63 9 5.70

1.20 2 1.27 11 6.96

1.21 1 0.63 12 7.59

1.25 1 0.63 13 8.23

1.29 1 0.63 14 8.86

1.30 1 0.63 15 9.49

1.47 1 0.63 16 10.13

1.50 2 1.27 18 11.39

1.52 1 0.63 19 12.03

1.54 1 0.63 20 12.66

1.67 1 0.63 21 13.29

1.70 1 0.63 22 13.92

1.78 1 0.63 23 14.56

1.86 1 0.63 24 15.19

1.92 2 1.27 26 16.46

1.94 1 0.63 27 17.09

1.96 1 0.63 28 17.72

2.O0 3 1.90 31 19.62

2.03 1 0.63 32 20.25

2.07 1 0.63 33 20.89

2.11 1 0.63 34 21.52

2.13 1 0.63 35 22.15

2.25 1 0.63 36 22.78

2.32 1 0.63 37 23.42

2.33 1 0.63 38 24.05

2.36 1 0.63 39 24.68

2.39 1 0.63 40 25.32

2.41 1 0.63 41 25.95

2.47 1 0.63 42 26.58

2.49 1 0.63 43 27.22

2.50 4 2.53 47 29.75

2.52 1 0.63 48 30.38

2.54 1 0.63 49 31.01

2.55 1 0.63 50 31.65

2.56 1 0.63 51 32.28

2.59 1 0.63 52 32.91

2.60 1 0.63 53 33.54

2.63 2 1.27 55 34.81

2.66 1 0.63 56 35.44

2.67 2 1.27 58 36.71

2.72 1 0.63 59 37.34

2.73 1 0.63 60 37.97

2.83 2 1.27 62 39.24

2.85 1 0.63 63 39.87

2.86 1 0.63 64 40.51

2.90 2 1.27 66 41.77

2.91 1 0.63 67 42.41

2.92 2 1.27 69 43.67

2.94 2 1.27 71 44.94

2.96 1 0.63 72 45.57

2.97 1 0.63 73 46.20

2.98 1 0.63 74 46.84

3.00 8 5.06 82 51.90

3.04 1 0.63 83 52.53

3.10 1 0.63 84 53.16

3.13 1 0.63 85 53.80

3.16 1 0.63 86 54.43

3.18 1 0.63 87 55.06

3.20 1 0.63 88 55.70

3.21 2 1.27 90 56.96

3.22 1 0.63 91 57.59

3.24 2 1.27 93 58.86

3.27 1 0.63 94 59.49

3.28 1 0.63 95 60.13

3.33 4 2.53 99 62.66

3.38 1 0.63 100 63.29

3.40 2 1.27 102 64.56

3.42 2 1 27 104 65.82

3.43 1 0.63 105 66.46

3.46 1 0.63 106 67.09

3.47 1 0.63 107 67.72

3.50 1 0.63 108 68.35

3.56 2 1.27 110 69.62

3.57 3 1.90 113 71.52

3.60 1 0.63 114 72.15

3.67 1 0.63 115 72.78

3.68 1 0.63 116 73.42

3.69 1 0.63 117 74.05

3.7 1 0.63 118 74.68

3.71 1 0.63 119 75.32

3.73 2 1.27 121 76.58

3.75 1 0.63 122 77.22

3.76 1 0.63 123 77.85

3.78 3 1.90 126 79.75

3.81 1 0.63 127 80.38

3.83 1 0.63 128 81.01

3.85 1 0.63 129 81.65

3.86 1 0.63 130 82.28

3.90 2 1.27 132 83.54

3.91 1 0.63 133 84.18

3.95 1 0.63 134 84.81

3.98 1 0.63 135 85.44

4.00 23 14.56 158 100.00

References

Ary, D., Jacobs, L. C., & Razavieh, A. (1985). Introduction to research in education (3rd ed.). Fort Worth, TX: Holt, Rinehart, and Winston.

Ary, D., Jacobs, L. C., & Razavieh, A. (1996). Introduction to research in education (5th ed.). Fort Worth, TX: Holt, Rinehart, and Winston.

Armstrong, W. B. (2000). The association among student success in courses, placement test scores, student background data, and instructor grading practices. Community College Journal of Research & Practice, 24, 8, 681-695.

Bishop-Clark, C., & Lynch, J. (1998). Comparing teacher and student responses to the mixed-age college classroom. Community College Review, 25, 4, 21-34.

Borglum, K., & Kubala, T. (2000). Academic and social integration of community college students: A case study. Community College Journal of Research & Practice 24, 7, 567-577.

Carlan, P. E. (2001). Adult students and community college beginnings: Examining the efficacy of performance stereotypes on a university campus. College Student Journal, 35, 2, 169-181.

Gilliland, B. E., & James, R. K. (1997). Crisis intervention strategies (2nd ed.). Pacific Grove, CA: Brooks/Cole.

Gonzenbach, N. M. (1993). Factors affecting decisions by office occupation students to continue education beyond the community college. Community College Review, 21, 2, 22-32.

Gusentine, S. D., & Keim, M. C. (1996). The learning styles of community college students. Community College Review, 24, 3, 17-27.

Haggan, P. S. (2000). Transition counseling in the community college. Community College Journal of Research & Practice, 24, 6, 427-442.

Johnson, L. G., Schwartz, R. A., & Bower, B. L. (2000). Managing stress among adult women students on community colleges. Community College Journal of Research & Practice, 24, 4, 289-300.

Makuakane-Drechsel, T., & Hagedorn, L. S. (2000). Correlated of retention among Asian Pacific Americans in community colleges: The case for Hawaiian students. Community College of Research & Practice, 24, 8, 639-655.

Mason, H. P. (1998). A persistence model for African American male urban community college students. Community College Journal of Research & Practice 22, 8, 751-761.

McAloon, D. T. (1994). The effect of alcohol abuse on academic achievement on two-year campuses. Community College Review, 22, 1, 12-18.

National Center for Educational Statistics (1995). Profile of older undergraduates: 1989-1990. Washington, D.C.: U.S. Department of Education.

Tinto, V. (1975). Dropouts from higher education; A theoretical synthesis of recent research. Review of Educational Research, 45, 89-125.

Tinto, V. (1987). Leaving college: Rethinking the causes and cures of student attrition. Chicago, IL: University of Chicago Press.

Tinto, V. (1993). Leaving college: Rethinking the causes and cures of student attrition (2nd ed.). Chicago, IL: University of Chicago Press.

Tinto, V., Russo, P., & Kadel, S. (1994, February/March). Constructing educational communities: Increasing retention in challenging circumstances. Community College Journal, 64, 4, 26 29

U.S. Department of Education (1996). Digest of educational statistics. Washington, D.C.: U.S. Government Printing Office.

Waycaster, P. (2001). Factors impacting success in community college developmental mathematics courses and subsequent courses. Community College Journal of Research & Practice, 25, 5/6, 403-416.

T. Ross OWEN, ED.D.

Assistant Professor and Coordinator

Adult & Higher Education Graduate Program

Morehead State University

COPYRIGHT 2003 Project Innovation (Alabama)

COPYRIGHT 2003 Gale Group