Self-Concept and the College Student

This topic submitted by Stephanie Lee, Pulkit Datta, Allison Croft, Emily Hackett, and Hilary Stevens ( stevenhk@muohio.edu ) on 12/12/04 .

Natural Systems 1 Syllabus---Western Program---Miami University

Allison Croft, Pulkit Datta, Emily Hackett,
Stephanie Lee, & Hilary Stevens
Chris Myers
Natural Systems 1
October 7, 2004
Self-Concept and the College Student
Introduction
This lab will examine self-concept in Miami University students. The purpose of this study is to explore the relationship between the year a student is in college and the level of their self-concept. We have chosen a statistical study method in order to show whether the year one is at Miami University leads to a greater self-concept. Do upperclassmen have a greater self-concept than lower classmen? We hypothesize that there is a relationship between the year of a Miami University Oxford undergraduate student and their self-concept, such that an upper classman has a greater self-concept. The null hypothesis in this study is that there is no relationship between the year an undergraduate is at Miami University Oxford and self-concept.
This research is interesting to us because we are first year students at Miami University Oxford campus. We are part of a main group we will be using in this study, thus the information we find will be directly applicable to us. As first year students we are in a brand new place, perhaps having left home for the first time. This puts us in a very vulnerable position. It will be extremely interesting to find how much being in this liminal place affects our self-concept now and even if it stretches into later years of college. This research will help us to better understand ourselves at this critical point in our lives. It will also help us to appreciate our peers in a more empathetic light.

Background
When we refer to Miami University students, we mean undergraduate students between the ages of 17 and 25 enrolled in and attending the Miami University Oxford Campus. When we mention self-concept, we mean Òthe totality of a complex, organized, and dynamic system of learned beliefs, attitudes, and opinions that each person holds to be true about his or her personal existenceÓ(Purkey, 1988). OneÕs self concept has to do with the confidence they have in themselves and the self-esteem they possess. Our research further supports our investigation.
In their article ÒSelf Esteem Is Up, But Society Has Little To Show For It,Ó Michael Smith and Gary Vagin state that college students do seem to be happier and healthier than in times past. Society seems to have benefited little from the improvement. Overall, college studentsÕ self-esteem has continually risen in the past twenty-four years, but at the same time, test scores have decreased and there has been a great increase in anxiety. In a related article, the concept of cross-temporal meta-analysis to determine the above issues is discussed. This method provides for age and group effects to be studied simultaneously, therefore increasing the understanding of generational differences in self-esteem (Osborne 2001). In this analysis, it was discovered that self-esteem for children is at an all-time low, while college studentsÕ self-esteem is rising.
Jennifer Crocker, PhD, has found that the reasons for this rising self-esteem in college students are partially due to internalization of self-esteem. If a student is driven in life by religion, morals, or values, he or she is more likely to have higher grades and fewer stress-related issues. If a student bases his or her self-esteem on external components Ð appearance, receiving compliments or criticisms from others, or academic performance Ð he or she is more likely to suffer from low self-esteem, as well as stress, lower grades, relationship issues and eating disorders. Unfortunately, an overwhelming amount of students base their self-esteem on the external components. 80% of students reported that their self-esteem comes from their academic performances, while another 66% depend on having more achievements than others and 65% (70% of which are women) depend on appearance (Crocker 2002).
Clinical practitioners and researchers have also long had difficulty in enhancing low self-esteem. Brockner (1979) has found that low-self-esteem people often do worse in achievement settings Ð settings where oneÕs accomplishments will directly affect their self-esteem Ð and as a result their self-esteem remains low. He also discovered that the low self-esteem people could be more easily manipulated by using distractions such as audiences and video cameras. Brockner (1979) stated that stimuli focusing on the self caused more anxiety in low self-esteem people, but the anxiety was lessened when the stimuli focused on the task itself. It was also revealed that if the low self-esteem people were surrounded by both stimuli that focused on the self and stimuli that focused on the task, their performances improved greatly. Duval and Wicklund (1972/1975) have also suggested that positive self-focused stimuli alone will enhance positive affect.
In a study done by Brockner, subjects of both high an low elf-esteem arbitrarily received success and failure feedback on a fake test of their social insight, and then completed a concept formation task in the presence and absence of self-focusing stimuli (Brockner 1979). Low self-esteem people with positive feedback and self-awareness were predicted to have a more positive outlook, and the same people with negative feedback and a lack of self-awareness would then have a more negative outlook. In summary of the study, the major hypotheses were supported. The fact that low self-esteem people performed better when they were told they were successful is a very positive step in breaking the cycle of low self-esteem. From the results of the study, Brockner (1979) also argued that sheer distraction rather than self-anxiety caused the low self-esteem people to be negatively affected by the self-focused stimuli. In a follow-up to BrocknerÕs study, Carver (1979) suggests that self-awareness is the key to success in improving self-esteem. He summarizes that the nature of the focus was crucial, rather than the degree to which the low self-esteem people could not focus.
Another study on self esteem was done by Madonna and Philpot (1996), who studied the use of positive to negative self-statements, loss of control, and self-esteem ion distinguishing between scores on the Beck Depression Inventory, this test was administered to undergraduates. A stepwise discriminate analysis indicated that five variables (ratio of positive to negative self statements, self esteem, age, gender, ethnic background) combined to yield a sizable discrimination among low, middle and high scores on the Beck Depression Inventory, the classification analysis indicated that an overwhelming amount of students were correctly classified as low or high scorers.
Michael, Denny, Lee and Michael (1984) were concerned with academic self-concept measures for college students; the major purpose of their study was to present information regarding the development and determination of the construct validity. They addressed four main areas of concern in this study; the underlying theory or rationale of affectivity in school learning, a description of the outline of the test, an exposition in the steps followed in the study, and a quick review of the samples making up the database, their rational was that Òan unrealistic level of aspiration Ð either too low or too high Ð was hypothesized to be related to the probable subsequent occurrence of anxiety. (Michael 1984).
Philpot et al. assert that positive self-talk influences oneÕs sense of a locus of control, helping to bolster self-esteem and Òover-all well-beingÓ (p.1007). The study involved an examination of a sample of 145 undergraduate college students, ranging from 19 to 53 years in age. The ratio of black to white was 1:4. Philpot et al. used the Coopersmith Self-esteem Inventory- Adult Form (Coopersmith, 1981), along with other tests of self-concept. They examined the frequency of positive self-statements as an indicator of positive self-image, noting that this correlated directly to positive scores on the Self-esteem inventory.
On the same note, Chang hypothesizes that perfectionism and social problem solving have Òaddictive and interactive effects in predicting psychological maladjustmentÓ (p. 581). Testing involved a sample of 371 college students, with conclusions towards the connection between social problem solving over perfectionism as being a prime predictor of suicide ideation and depressive symptoms. Chang thereby proposed the ÒPerfectionism x Social Problem Solving interactionÓ (581), whereby problems with self-esteem and social problems result in correlation to the presence of maladjustment.
The study conducted by Gibb et al attempted to Òexamine the relation between attributional styles assessed during the freshman year and academic achievement over the entire college careerÓ (309). The study found that those students with an internal locus of control and strong self-concept were more negatively affected by lower academic performance and were less likely to improve their academic records. Thus, students who believed they were responsible for their academic performance could not assume the ability to change, and thus psychologically convinced themselves of failure (thus fulfilling a self-fulfilling prophecy).
One study on explanatory style, by Seligman (1996), implies that self-esteem and achievement are connected by one intervening variable, which is the studentÕs level of optimism. This is defined as whether the student sees his/her world as one where positive things will occur, or one where negative events will occur.
Also concerning academic self esteem, Gage and Berliner (1992) state that "the evidence is accumulating, however, to indicate that level of school success, particularly over many years, predicts level of regard of self and one's own ability (Bridgeman & Shipman, 1978; Kifer, 1975); whereas level of self-esteem does not predict level of school achievement. The implication is that teachers need to concentrate on the academic successes and failures of their students. It is the student's history of success and failure that gives them the information with which to assess themselves" (p. 159).
In contrast, Franken (1994) states that "there is a great deal of research which shows that the self-concept is, perhaps, the basis for all motivated behavior. It is the self-concept that gives rise to possible selves, and it is possible selves that create the motivation for behavior" (p. 443). Franken also suggests that self-concept is related to self-esteem in that "people who have good self-esteem have a clearly differentiated self-concept.... When people know themselves they can maximize outcomes because they know what they can and cannot do" (p. 439).
Along the same lines, whilst considering the aspect of ethnicity/race in relation to the studentsÕ self-esteem, we can refer to the studies conducted by Mooney et al (1991), where eighty-eight female college freshmen answered a questionnaire designed to assess academic locus of control, self-esteem, and geographical distance from home, as indicators of adjustment in college. The results showed that positive responses to all three areas were related to the personal, social, academic, and attachment dimensions of college adjustment. However, no associations were found between the studentsÕ actual distance from home and the dimensions of adjusting to college.
Methods
Experimental Study Design
In order to conduct our experiment, we will be using the Rosenberg Self-esteem Scale. This scale is perhaps the most widely used self-esteem measure in social science research. Dr. Rosenberg is the author or editor of numerous books and articles, and his
work on the self-concept, particularly the dimension of self-esteem, is world-renowned.
The original sample for which the scale was developed in the 1960s consisted of 5,024 high school juniors and seniors from ten randomly selected schools in New York State and was scored as a Guttman scale. The scale generally has high reliability: test-retest correlations are typically in the range of .82 to .88. (Rosenberg, 1989.)
We will distribute the survey to 200 students. If our respondents are not varied enough, we will distribute more. The survey will consist of the following:

The student will rate their level of agreement with each statement on a scale of zero to four. Zero means least characteristic of the student, and four means the most characteristic. Questions two, five, six, eight, and nine will be scored oppositely, as in if the student answered with a number one, the amount added into the score would actually be three.
A total score of twenty or below is considered to be low self-esteem, resulting in a low self-concept (Rosenberg, 1989). With a score of 30 or above the student is considered to have a very high self-esteem level. The highest score attainable on this scale is a 40. Through analyzing and comparing these scores we will be able to determine each individualÕs personal self-concept. At the same time, we are asking each studentÕs year in college so we can later define whether or not there is a relationship between it and their self-concept. We also are collecting data regarding the individualÕs ethnicity, number of credit hours, and gender to see if we can find other patterns.
The data we collect will then be counted and recorded into data charts. From these tables, we will create graphs to act as visual aids showing whether or not there is a sharp increase or decrease in levels of self concept in relation to college year. Other graphs will show if variables affected the studentsÕ scores, and if so, which ones affected them the most.
Sampling
For the sampling portion of this project, we will be going around to some of the most populated places on the Miami University campusÑboth main campus and the Western campusÑand distributing surveys to some of the people there to obtain our information. We have determined the places on campus that will be the most helpful to us in completing this project; we will be more likely there to find a great difference in majors, genders and races, as well as personality types. The places we chose on main campus are as follows: within the various academic and residential quads, inside the Shriver Center, and various different classes on campus. On Western campus, the best place to hand out the surveys is Alexander Dining Hall, as that is where large groups of students, both Western and from the other residence halls located on the Western campus, can be found at almost any time.

Materials
The materials needed for this project are people and surveys. The people include not only the representative sample of Miami UniversityÕs population that we will study (two hundred people out of the approximately twelve hundred that live on the campus), but also the members of the group that will be distributing the surveys. The surveys themselves are psychological surveys provided by a member of our group, and said surveys will measure self-esteem and the effects of different living situations a

Timeline of Research Execution
September 2, 2004: Create SGL Group, named ÒSpe?Ó

September 9, 2004: Brainstorm Session for SGL Idea
September 14, 2004: Research Idea posted (Squirrel Fishing)
September 24, 2004: Research Idea changed (Honeysuckle)
September 28, 2004: Final Research Idea developed (Self-concept)
September 28-30, 2004: Discuss proposal requirements, study references
October 1-6, 2004: Write research proposal
October 7, 2004: Make copies of surveys and begin distribution
October 11-28, 2004: Organize, distribute surveys, collect and analyze data
November 2004: Dates left open if repeat or further testing is necessary, data analysis, final report preparation
December 2004: Final report write-up
Results
As stated in the Methods, the range of the scores from the questionnaire goes from 0 to 40. Since the students were asked to rate themselves on a scale of 0 to 4 on each question asking about their self-esteem, we then added up the score for each student. This enabled us to create a chart of the results for comparison purposes, and then graph the relevant data. A score of 20 or below was considered to signify low-esteem, and a score of 30 or above was considered a very high self-esteem level.
Some of the questions on the questionnaire were scored opposite, due to the phrasing of the particular questions, and the attempt at making the subject think about the answer rather than simply choosing the answer that will guarantee them a higher score, thus making our results invalid. Therefore, questions two, five, six, eight, and nine were scored opposite, as in if the student answered with the number one, the amount added into the score would be three.
Once all the scores are calculated, we created one primary chart (see Appendix), which illustrated the main results that we sought in this investigation: that of the relationship between the Miami studentÕs year with his/her score of self-esteem. From there, we branched out the results from the questionnaire and chart and graph the self-esteem scores with variables such as gender, ethnicity, and number of credit hours taken by the students.
From the charts, we will then be able to derive our conclusion and assess whether it supports our initial hypothesis, and essentially, how it answers our research question. The following are relevant graphs and data from our study.

Bivariate Fit of score by credit hours

Linear Fit
Score = 25.695215 + 0.2539574 credit hours

The linear model suggests a positive correlation between credit hours and self-esteem measures, though not a strong, positive correlation. R=0.07528 is too small to be conclusive of a linear relationship. The wide range of data points, a well as the large variance suggest a strong conclusive relationship, thus leading us to be doubtful of the dataÕs accuracy.

Oneway Analysis of score By Year

Oneway ANOVA
Summary of Fit

Rsquare 0.01957
Adj Rsquare -0.00054
Root Mean Square Error 6.415585
Mean of Response 29.865
Observations (or Sum Wgts) 200

Graph of 95% confidence levels show an upward trend in self-esteem scores over years. Indeed, when one examines the means within the years, there seems to be a dip in means during the second year, with a steady upward progression, the highest means being in the 4th and 5th years.

Oneway Analysis of score by ethnicity

Means and Std Deviations
Level Number Mean Std Dev Std Err Mean Lower 95% Upper 95%
A 12 26.2500 8.04674 2.3229 21.137 31.363
B 14 30.6429 6.82312 1.8236 26.703 34.582
BA 1 36.0000 . . . .
C 152 29.9803 6.23807 0.5060 28.981 30.980
H 4 32.7500 4.99166 2.4958 24.807 40.693
I 5 25.6000 8.90505 3.9825 14.543 36.657
IRAN 1 28.0000 . . . .
NA 1 34.0000 . . . .

CONCLUSIONS:
Questions that arose during our investigationÉ

The discussion of identity and self-concept has always played a prominent part in the transition from high school to college. Moreover, the college years have been touted as the years of change and progression to independence, the stage between childhood and responsible adulthood. Inherent to this process of change and transition is an underlying emphasis on the attitudes and personal conception of Òself.Ó An individualÕs self-esteem, personal attitude and feelings of control and direction in life change dramatically during the turbulence and confusion of the college years.
We examined the data carefully for a correlation between academic year (or ÒexperienceÓ) and the improvement or decline of self-concept (or self-esteem). Upon analyzing the statistical relationship between experience and self-esteem, we determined if self-concept is truly a function of life experience. Our deduction of the results is that a studentÕs stage in the college experience is a variable too simple to attribute his/her self-concept to. As Purkey stated, self-concept is Òthe totality of a complex, organized, and dynamic system.Ó
The greatest problem in our results, and a factor that consequently affects our conclusion of the study, was the uneven sample sizes. Out of a total of 200 students sampled, 102 (51%) were freshman/first years only. The sample sizes for sophomore, junior and senior years were 42 (21%), 26 (13%), and 27 (13.5%), respectively. These sample sizes (except for the first and perhaps the second year samples) were clearly not large enough to be accurately representative of the population. The statistical rule that a minimum sample size of 30 is required for the results to be significantly representative of a population should be applicable in this case. Therefore, our hypothesis that the self-concept of seniors (4th year students) at Miami University would be significantly higher than that of freshmen (1st years) has been proven wrong. There is no statistically significant difference, and this can be attributed to the fact that the sample size of seniors is almost a quarter that of the freshmen.
Another difficulty in the analysis of our data was the prominence and significance of human perspective in this aspect of experimentation. Because the human volunteers completed the surveys based wholly on their own feelings of ÒcompletenessÓ (a very vague and subjective term), a very specific and detailed scale was necessary to apply across different subject groups.
In analyzing the data, we were also careful to avoid confirmation bias, and to maintain objectivity. We also kept in mind that the answers and values placed on sentiments and feelings are entirely subjective, and thus not conclusive measurements of secure identity, but merely reflections of a larger idea and comfort with the self. When an individual evaluates him/herself as high on the evaluative scale, he/she is making a quantitative judgment on the quality of his/her life experience. But because the experiences, and thus the interpretations of those experiences, will differ largely for each person, due to the inevitability of subjectivity and perceptive complications inherent to human subjects and their judgments, our data was examined in terms of other controllable variables for the sake of comparison. Additionally, the individuals were included within the context of a larger population for general comparison. Thus, the data was examined for signs of a group norm or pattern.
Another consideration for the experimental design was to test other variables or characteristics, to determine their effects on self-concept. We took into consideration that there is more to take into account in the development of oneÕs identity, beyond simply the amount of experience and acclimation one has had to a given situation. Analysis of psychological literature revealed aspects of identity that form as a result of other factors that contribute or affect experience, such as gender, race, cultural background or societal influences, sexuality or academic focus or determination, and distance from home. Therefore, we asked the sampled students to tell us their gender, ethnicity, number of credit hours, and whether or not they were a transfer student, to make secondary analyses to see how these variables affected the scores of self-concept. The results for these variables were not significantly indicative of any trend. This, again, can be attributed to the greatly uneven sample sizes.
Going deeper into one of the variables, ethnicity, we did notice a slight trend. Despite the lack of a great difference between the mean scores of the different ethnicities, it could be seen that students of Asian backgrounds had slightly lower scores than others. It also seemed as if the Caucasian and African American students had almost equally high levels of self-concept. This could be explained by the general perception of Asian cultures as more humble, and less self-assertive, perhaps more so in a Caucasian-majority community such as Miami University.
In conclusion, there are several changes that can be made to this study. First and foremost, the sample sizes need to be more fairly representative of the population. Also, perhaps the survey could have been expanded to make the sampled students think more about their self-concept in terms of the other variables we hinted at. Self-concept in college students is definitely an area that needs to be studied further, and this is taking into consideration that Miami University is not a typical example of an American university. Perhaps a college with a greater international student population would have an overall higher self-concept amongst its students. Perhaps there are other factors to be taken into consideration, such as personal life, sexual orientation and the resulting societal pressures, and physical distance from home. The possibilities for this subject of study are, indeed, endless.

Bibliography

¥ Brockner, Joel. (1979) The Effects of Self-Esteem, Success-Failure, and Self-Consciousness on Task Performance. Journal of Personality and Social
Psychology, 1979, Vol.37, No. 10, 1732-1741.

¥ Chang, Edward C. (2002) Examining the Link Between Perfectionism and Psychological Maladjustment: Social Problem Solving as a Buffer. Cognitive Therapy and Research, Vol. 26, No. 5, October 2002, pp. 581-595.

¥ Crocker, J. (2002) Journal of Social Issues (Vol. 58, No. 3).

¥ Dittman, M. (2002) Self-esteem that's based on external sources has mental health consequences. Monitor on Psychology. Volume 33, No. 11 December 2002.
¥ Franken, R. (1994). Human motivation (3rd ed.). Pacific Grove, CA: Brooks/Cole Publishing Co.

¥ Gage, N., & Berliner, D. (1992). Educational psychology (5th ed.). Boston: Houghton Mifflin.

¥ Gibb, Brandon E., Zhu, Lin, Alloy, Lauren B., Abramson, Lyn Y (2002). Attributional Styles and Academic Achievement in University Students: a Longitudinal Investigation. Cognitive Therapy and Research, Vol. 26, No. 3, June 2002, pp. 309-315.

¥ Madonna Jr, Stephen and Vincent D. Philpot. (1996) Self-Statements, Self-Esteem, and Locus of Control In Discriminatory College Students' Scores on the Beck Depression Inventory. Psychological Reports, 1996.

¥ Michael, William B., et al. (1984) The Development and Validation of a Preliminary Research Form of an Academic Self-Concept Measure for College Students.
Educational and Psychological Measurements, 1984.

¥ Mooney, S.P. et al. (1991). Academic Locus of Control, Self-Esteem and Perceived Distance from Home as Predictors of College Adjustment. The Journal of Counseling and Development. Vol. 69, p.445-448.

¥ Osborne, Kim. (2001) Self-Esteem of College Students Increased Substantially Over 25-Year Period, But Benefits to Society Unclear. 2001.
http://www.eurekalert.org/pub-releases/2001-10/uog-soc10150.php

¥ Philpot, Vincent D., Holliman, W. Bruce, Madonna Jr., Stephen. (1995) Self-Statements, Locus of Control, and Depression in Predicting Self-Esteem. Psychological Reports, 1995, v. 76, p.1007-1010.

¥ Purkey, W. (1988). An overview of self-concept theory for counselors. ERIC Clearinghouse on Counseling and Personnel Services, Ann Arbor, Mich. (An ERIC/CAPS Digest: ED304630)

¥ Rosenberg, Morris. 1989. Society and the Adolescent Self-Image. Revised edition. Middletown, CT: Wesleyan University Press.

¥ Seligman, M. (1996) The optimistic child: How learned optimism protects children from depression. New York: Houghton Mifflin.

¥ Smith MD, Michael and Gary Vagin, MD. Self-Esteem is Up, But Society Has Little to Show For It. WebMD Medical News Archive. Oct 22, 2001
http://my.webmd.com/content/article/35/1728_91573

Next Article
Previous Article
Return to the Topic Menu