College Student Retention: Instrument Validation and Value for Partnering Between Academic and Counseling Services
By Coll, Kenneth M Stewart, Roger A
Designed to explore the utility of a retention assessment of students within a college of education, and using Pascarella and Terenzini’s (1983) academic and social integration scales, this study found several pertinent results: (a) Factor analyses of the scales support earlier validation studies with first-year college students, and (b) The scales differentiate academic integration, social integration, and career decidedness between students who have been identified as at-risk (i.e., on probation, previously on academic suspension) and those who have not been so identified. The utility of these scales for use in partnered programming between university counseling services and faculty is also considered. Universities and the undergraduate professional colleges within them (e.g., business, engineering, education) need to first attract and then retain high quality students (Archer & Cooper, 1999). Retention is a long-standing challenge that institutions continue to address (Braxton, Bray, & Berger, 2000). Counseling can play an important role in this process; but because of heavy case loads, college student counselors need efficient ways to identify students at-risk and ways to partner with professional schools and academic departments so they, too, can help with retention issues (Archer & Cooper, 1999).
Two variables important to retention efforts are academic and social integration (Tinto, 1993). Students are academically and socially integrated when they have positive regard for their academic performance and they value the social relationships they have established at the institution. Tinto (1975) articulated a model of college student persistence/withdrawal based on these variables, and Pascarella and Terenzini (1983) generated scales to measure them.
In this study, Pascarella and Terenzini’s scales were used to measure academic and social integration in a sample of students enrolled in an introductory education course. The course was part of the required course work leading to state teacher certification. The scales were part of a survey developed by the authors that was administered to the students at the beginning of the semester and then at the end of the semester. Data from the scales was employed in the following analyses. First, the scales were explored through principal component analysis to add to the extant literature on the internal structure, validity, and reliability of these scales. second, the scales were used to explore differences between those students in the sample who had been identified as at-risk and those who had not been so identified. This was done to ascertain whether the scales discriminated between two identifiable and quite different subpopulations in the sample. The discrimination function was seen as a preliminary exploratory step to building meaningful norms that could be used by college student services personnel as they work with students and faculty on student retention and institutional fit issues.
In the discussion section of this paper, the results of these analyses become the centerpiece in a discussion of the initiator- catalyst (I/C) model, whereby counselors can work with undergraduate professional schools to build collaborative relationships to enhance a wide range of institutional objectives (Archer & Cooper, 1998; 1999). We believe such collaborative relationships are important for two reasons. First, such relationships can help undergraduate professional schools better achieve their overarching objective of attracting and retaining the most able and highly committed students to their programs. Second, collaboration between counseling services and professional schools provides avenues for such services to be more deeply embedded in the academic culture of the institution thereby enhancing counseling’s important contribution to institutional goals. In the remainder of this introduction, we explore Tinto’s model of college student retention, including the academic and social integration constructs, which are a direct outgrowth of Tinto’s model. We then describe the I/C approach for college student counseling.
Tinto ‘s Model of College Student Retention
Tinto’s theory of college student departure has near paradigmatic stature (more than 400 citations to the theory) (Braxton, Bray, & Berger, 2000). However, the application to professional programs within universities remains largely unconsidered, since the bulk of the work in this area focuses on first year students who are not yet part of professional preparation programs that usually induct students after the freshman year. Tinto’s (1975; 1993) theory is based on the degree of fit between the individual student and the college environment. As Pascarella and Terenzini (1983) note in their review of Tinto’s theory, students come to a particular institution with a range of background traits (e.g., race, secondary school experiences, academic aptitude, family background). These traits lead to initial commitments, both to the institution attended and to the goal of graduation from college. Together with background traits, these initial commitments then influence not only how well students perform in college, but also how they interact with and subsequently become integrated into the institution’s social and academic systems.
Tinto (1975) postulated that “academic integration is characterized by structural and normative dimensions. Structural integration involves the meeting of explicit standards of the college or university, whereas normative integration pertains to an individual’s identification with the normative structure of the academic system” (p. 104). “Social integration pertains to the degree of congruency between the individual student and the social system of a college or university. Mechanisms of social integration include informal peer group associations, extracurricular activities, and interactions with faculty and administrators” (Tinto, 1975, p. 107). Both Tinto (1975; 1993) in his original theory and Braxton et al. (2002) underscore the important relationship between levels of academic and social integration and students’ commitments to the institution and college graduation. High levels of academic and social integration lead to greater institutional commitment and persistence that result in higher graduation rates (Braxton, et al., 2000).
The role of the classroom environment in the college student departure process has recently begun to receive consideration. This environment serves as a critical gateway for student involvement in the academic and social communities of a college (Braxton, et al., 2000). Thus the undergraduate professional school constitutes a source of influence on academic and social integration, and subsequent college departure.
Research evidence also suggests that the utilization of counseling services has a positive influence on college student retention (Bishop & Walker, 1990; Coll & Stewart, 2002). However, academic and social integration as applied to at-risk students enrolled in an undergraduate professional program of study remains an important yet unexplained phenomenon in the college student departure process.
The UC Approach
The initiator-catalyst (I/C) approach to college counseling combines academic and counseling influences by emphasizing the counseling staff’s role in energizing other campus groups (e.g., academic programs) to become involved in the operation and leadership related to student success (Archer & Cooper, 1998; 1999). For example, counseling staff may act as initiators and catalysts for system changes within undergraduate professional colleges at universities (e.g., architecture, business, education) leading to higher levels of student academic and social integration. Archer and Cooper (1999) specifically encourage counseling staff to: 1) initiate efforts to change and modify norms for greater student success and development, 2) participate in assessment related to college objectives, 3) secure a role in faculty development activities, and 4) delegate the development of outreach to faculty and other constituencies to the fullest extent possible.
It is within this partnered academic programs and student services context that this study was developed and executed. The first author is a counselor educator interested in college student counseling issues. The second author is a teacher educator who is interested in recruiting and retaining the highest quality candidates for teacher certification while also helping all students find their place in the institution. The impetus behind this study came about as a consequence of the second author noticing that the pre-service teachers, who subsequently became the participants in this study, manifested a wide range of commitment to and interest in the dynamic and challenging field of teaching. He realized that just because students were enrolled in a class in a professional school doesn’t mean that they are fully integrated into the university and completely sure of their career direction. This study was a response to these insights and the need for counseling services to be strategically applied in one professional school setting. Theoretical Constructs and How They Were Employed in this Study
Pascarella and Terenzini developed four scales (1983) based on Tinto’s theory. They developed a 14 item self-report questionnaire that measures the scales. Tinto’s work and that of Pascarella and Terenzini focused on first-year undergraduate students. Thus, understanding how the key constructs of academic and social integration factor out for a population of teacher education students all of whom were of sophomore standing or higher, when compared to a more general first-year college student population, can be useful when applying the scales and the I/C model to professional school populations. For example, recent research suggests (Coll & Stewart, 2002) that some students in teacher education populations may struggle with academic and social integration, but the structural factors of these constructs as related to this population have not been extensively studied. Teacher preparation programs are under increasing pressure to produce large numbers of high quality teachers who can immediately move into classrooms and enhance student achievement. This puts preparation programs under tremendous pressure, since simultaneously increasing quality and graduation numbers represents a formidable challenge (Darling-Hammond, 2000).
In addition to the Pascarella and Terenzini scales, the authors developed a four item scale for principle component analysis which were hypothesized to capture a missing link between institutional fit and career direction (Coll & Stewart, 2002). Teacher education students were asked a series of questions about how sure they were about their chosen career direction (e.g., “How confident are you in your ability to teach?” How would you rate you desire to become a teacher?”). This construct was titled “career decidedness”.
Career decidedness is considered an important variable within the realm of an undergraduate professional program of study (Coll & Stewart, 2002). Career Decidedness directly relates to the “goal commitment” component of Tinto’s model, which influences and is influenced by academic and social integration (Coll & Stewart, 2002). Goal commitment is defined as how sure, confident, and committed a student is to an academic and career path (Tinto, 1993). Undergraduate professional programs, including teacher preparation programs, need highly committed individuals who are sure of their career direction. Professional programs should strive for academic and social integration so that students persist, graduate, and have strong, positive dispositions toward the institution and profession. These professional programs need students who are sure of what they want to do, so that finite energy and resources within the program are focused on those who will benefit the most. It is within this complex of variables that we see the I/C model becoming important to professional schools. For example, after counseling staff teach professional school faculty about these important variables, they can work collaboratively to measure these variables (especially in at-risk students) and build programs and intervention strategies.
Counselors are also under pressure to show how their work furthers the institutional vision and mission. The intersection of all of these variables produces a powerful framework in which the I/ C model can operate. Additionally, Pascarella (2001) has noted the importance of using psychometrically reliable instruments for estimating college impact. He indicates that, unfortunately, this is all to rare in retention studies.
All undergraduate teacher education students enrolled in a multiple section introductory education course at a Rocky Mountain regional University (total undergraduate enrollment, approximately 10,000 students) were included in the sample (n=304). The sample included three semesters of students who enrolled in the course. The largest group of participants were second-year (41.7%), majoring in education (81.3%), taking at least 12 credit hours, were single (85%), and averaged 21 years of age, with a range of 18 to 47 years. The ethnic composition of the participants was as follows: 92% White, 2.1% African American, 1.9% Hispanic, 2.9% Native American, and 1.0% labeled as “other,” a category composed of groups with less than 1% representation in the sample. The majority of the participants were female (60% women, 40% men). Atrisk students were identified within this large group.
Pascarella and Terenzini’s (1983) factor analyzed their scales using a sample of 1906 persons drawn by computer from the first- year class of an independent residential university in the Mid- Atlantic region of the United States. The analysis yielded results supporting the four-factor solution initially suggested by them in 1979. This structure strongly parallels the two theoretical constructs of academic and social integration.
Pascarella & Terenzini (1983) reported that twenty-four items were factor analyzed and four factors were isolated. Sample questions included “I am satisfied with the extent of my intellectual development since enrolling in this university” and “The student friendships I have developed at this university have been personally satisfying”. The two highest loading items on each factor were retained for the final scale (8 items) along with six (6) items from a pool of non-factorially analyzed but theoretically driven items (total of 14 items). Pascarella and Terenzini (1983) did not explain the rationale for not factor analyzing all of the items (see Table 1).
Pascarella and Terenzini (1983) reported internal consistency reliabilities (alpha) in their four-factor solution as follows; factor one (academic integration academic/intellectual development, .72), factor two (academic integration – faculty concern for student academic development, .77), factor three (social integration – peer group relations, .84), and factor four (social integration – informal interactions with faculty, .83). The overall constructs of academic and social integration had internal consistency reliabilities (alpha) of .64 and .46 respectively, and were correlated (r=.36; p
A recent study of these constructs has revealed similar alpha reliabilities for the subscales along with their predictive validity. Braxton, Bray, and Berger (2000) report two social integration subscale reliabilities (social integration – peer group relations= .76 and social integration informal interactions with faculty= .82). They constructed a composite of these two subscales and employed it in a path analysis exploring teacher effects on social integration and other variables. Results revealed that the more students perceived faculty as organized and prepared, the greater the level of student social integration.
As previously mentioned, the instrumentation used in this study included the 8 items factorally derived from PascarelIa and Terrenzini’s work. The instrumentation also included the six (6) non- class contact questions not factor-analyzed by Pascarella and Terenzini. Three of these six non-factor analyzed items related to academic integration and explored how often students access basic information about academic programs, discuss intellectual or course – related matters, and discuss matters related to future career. The other three of the non-factor analyzed items relate to social integration, and ask how often students and faculty socialize informally, discuss a campus issue or problem, and receive help in resolving a personal problem. Each of these questions was rated on a six point Likert-type scale, with Ox being no contact of at least 10 minutes per month with faculty and/or other academic professionals; 1x = once; 2x = twice; 3x = three times; 4x = four times; and over 5x = over five contacts). One additional social integration item was added (recommended by Pascarrela and Terenzini) which asked, “What is your involvement in university extracurricular activities” (1-4 scale: 1 = four or more hours per week, 2 = two to three hours per week, 3 = less than two hours per week, 4 = no involvement).
As previously mentioned and in keeping with Tinto’s hypothesis that academic integration and social integration influence goal commitment, 4 items termed “career decidedness” (i.e., How sure are you about your career direction?; How sure are you that you want to be a teacher?; How would you rate your overall desire to become a teacher?; and How would you rate your overall confidence in your ability to teach and perform the duties of a classroom teacher?) were also (developed by Coll and Stewart, 2002)included in the principle component analysis (1-5 scale: e.g., 1 = highly confident, 2 = very confident; 3 = confident; 4 = somewhat confident; 5 = not confident). These items were developed and piloted using Dillman’s (2000) procedures for development of survey questions.
In the planning stages of the project, the authors agreed to target the introductory teacher education course because it is often the entry point for teacher education undergraduate professional school training. The data collection period spanned three semesters. At the beginning of each semester, a brief presentation was made concerning the purpose of the study and confidentiality of participating. Questions from participants were also addressed. After the presentation and questions, asurvey was distributed that took approximately 15 minutes to complete. As a posttest measure, students completed the same survey at the end of the semester followed by a brief discussion about the nature of the variables explored in the study and their importance to students. At the conclusion, an offer was extended for individual participants to meet and discuss the study and any questions they might have.
Data Analysis Data analysis was a three-step process. The first step was an exploratory principal components analysis with varimax rotation. The second step was the computation of factor scores using summated scales (Hau-, Anderson, Tatham & Black, 1995). The questionnaire items are listed in Table 2 that were used in the summated scales for each factor. Step 3 was t-test analysis of the summated scale scores using a grouping variable for at-risk vs. not- at-risk students. The group of students, termed “at-risk” were defined as those students who did not persist, who had been placed on academic probation, and/or who had been suspended from school. In the remainder of this section, we will first describe results of the principle component analysis. This will be followed by the results of the t-test analyses of the summated scale scores of at-risk and not-atrisk students.
Principal component analysis
A total of nineteen items were factor analyzed: eight items factor analyzed from Pascarella and Terenzini’s original factor analysis, the six items from Pascarella and Terenzini that they did not factor analyze, one item on extracurricular activities recommended by Pascarella and Terenzini but not included in their original questionnaire, and four career decidedness items added by the authors (see Table 2). The Kaiser criterion (Kaiser, 1960) and interpretability were used to determine the number of factors to be rotated. The Kaiser criterion recommends that all factors with eigenvalues greater than one be considered. Seven factors met this criterion (see Table 2).
Varimax rotation was used to rotate the factor matrix. An orthogonal rotation was chosen for two reasons. First, it forces the factors to be independent from one another thus making interpretation less complex (Cohen, 1988). second, the researchers wanted to use a methodology similar to the original study, which employed a varimax rotation. According to Pascarella and Terenzini (1983), the scales reflect two distinctly different constructs- academic and social integration.
Factor loadings were quite strong (see Table 2). Only items loading at .5 or higher were retained for a factor. Given the sample size for this principle component analysis, these loadings are considered moderate to high (Field, 1999; Hair, et al., 1995; Stevens, 1986). Additionally, all items loaded at .5 or higher on only one factor, thus all 19 items were retained and interpretation of the factor structure was relatively clear. We now turn to a discussion of the seven factors. The seven-factor structure was interpreted within the theoretical constructs of academic and social integration and career decidedness.
Factor 1, called Career Decidedness, had high positive loadings for three items. These three items explore how sure respondents were about their choice of teaching as a career. The fourth item was also retained with a moderate factor loading of .55.
Factor 2 (three of the 6 items not factor analyzed originally) is dominated by high positive loadings for three items that explore the frequency of nonclass contacts with faculty of 10 minutes duration or more for academic purposes (Pascarella & Terenzini, 1983). These questions correspond to the theoretical definition of academic integration, but within a specific domain of this construct, namely faculty contact. This factor will be called Academic Integration – Faculty Contact.
Factor 3 (also Pascarella and Terenzini’s Factor 3) contains high positive loadings for two items. The items represent Social Integration – Peer Group Relations. These same questions correspond to the operationalized and factorially analyzed definition used in Pascarella and Terenzini’s study (see Table 1). A third item was added (extra curricular activities) as recommended by Pascarella and Terenzini), and is retained with a moderately high factor loading of .6.
Factor 4 (the other three of the 6 items not factor analyzed originally) is dominated by high, positive loadings for three items. These questions relate to the frequency of non-class contacts with faculty of 10 minutes duration or more for social purposes (Pascarella & Terenzini, 1983). This factor will be called Social Integration – Faculty Contact
Factors 5, 6, and 7 contain high positive loadings for two items each (see Table 2). The items represent Social Integration – Informal Interactions with Faculty (Pascarella and Terrinzini’s Factor 4); Academic Integration – Academic/intellectual Development (Pascarella and Terrinzini’s Factor 1); and Academic Integration – Faculty Concern for Student Academic Development (Pascarella and Terrinzini’s Factor 2), respectively.
Pascarella and Terenzini arrived at a four-factor structure, although they did not factor analyze 6 items that were included in this study. Our analysis resulted in a 7 factor structure accounting for 73.7% of the variance and revealing additional detail about the constructs of academic and social integration. This 7 factor structure is both similar to and different from that of Pascarella and Terenzini in the following ways. First, the original 4 factors emerged in this study also. Second, an additional factor, called Career Decideness, emerged as significant for this professional school population. Finally, the 6 items not analyzed by Pascarella and Terenzini resulted in two factors: Academic integration-Faculty Contact and Social Integration-Faculty Contact.
At-Risk and Not-at-Risk Comparative Summated Scale Analysis
For further validation, summated scales were constructed based on the factor structure (Hair, et al., 1995). Means from the summated scales were used in independent samples t test analyses exploring differences between students who were identified as at-risk and those who had not been so identified. Alpha level for these tests was set at p
T test analysis of Summated Scale 1: Career Decideness revealed that at-risk students were significantly lower in their career decideness than those not-at-risk (t=2.68, df=288; p=.008; d=.46). Analysis of Summated Scale 5: Social Integration-Informal Interactions with Faculty revealed significant differences between groups (t=2.24, df=283, p=.03; d=.41). Analysis of the means (Table 3) revealed that students not at-risk reported to a greater degree that their nonclassroom interactions with faculty had had positive influences on them. Similarly, analysis of Summated Scale 6: Academic Integration-Academic/Intellectual Development revealed that students not-at-risk were more satisfied with their intellectual development (t=1.71, df=287, p=.09; d=.30). Finally, analysis of Summated Scale 7: Academic Integration-Faculty Concern for Student Academic Development revealed a similar trend (t=2.53; df=284; p=.01 ; d=.45). Students not-at-risk reported higher faculty interest in students. They also evaluated faculty teaching ability more highly.
No significant differences between the at-risk and not at-risk groups were found for Summated Scale 2: Academic Integration- Faculty Contact (new factor), Summated Scale 3: Social Integration- Peer Group Relations (Pascarella and Terrinzini’s Factor 3), and Summated Scale 4: Social Integration-Faculty Contact (new factor). Scales 2 and 4 explored how often students had contact with faculty during a month. Amounts of faculty contact were equal between the groups, and were quite low overall. Scale 3 explored the quality of student friendships that had developed since coming to campus, and overall the quality was quite good. Again no differences were found between the two groups.
Discussion, Implications, Recommendations
Related to the purpose of the study principal components analysis and summated scales analysis revealed meaningful constructs with which to explore at-risk students enrolled in a professional school. Additionally, these analyses further validated Pascarella and Terenzini’s academic and social integration constructs.
This study, however, has some limitations. This study utilized a single-institution sample, which may not be representative of all teacher education professional schools. A limitation inherent in survey research is self-administration, which may lead to inconsistency in the way participants responded to the questions. Inconsistency in administration influences the quality of the data. In addition, the over-representation of females and the under- representation of diverse ethnic groups may have influenced the results. Our decision to interpret the orthogonally rotated factor solution, although in itself not a limitation, suggests that different results might have been achieved if other factor methodologies had been used.
The authors recommend that this procedure be repeated using a different sample of undergraduate professional school students. The stability of the solution over time is unknown until such studies are conducted. Utilizing a more stringent p-value (e.g, p.
Archer and Cooper (1999) indicated that collaboration among faculty and counseling services staff is crucial for the success of the Initiator/Catalyst approach. Indeed, a major recommendation from the professional literature is for strong and formalized collaboration between student and academic services to use the institutional environment (e.g., classroom and student services) for affecting academic and social integration (Pascarella & Terenzini (1991). The I/C approach encourages student services involvement with academic units, initiating joint student success projects (e.g., academic clubs), and relying on sound assessment data (Archer & Cooper, 1999). The assessment tool that has emerged from this study can help college counseling services staff operationalize the I/C model to successfully collaborate with professional school faculty and staff. For example, the assessment data derived from the tool can be used as a foundation upon which meaningful, data-driven conversations and interventions can be constructed. First, counseling staff could make professional school faculty aware of the importance of these variables. This would be followed by an ongoing collaborative relationship in which the scales would be used to assess and monitor program performance on the variables. Products of this collaborative relationship might include adjusting professional program course work, extra-curricular activities, or custom tailored counseling services and/or faculty interactions designed to impact the variables. Importantly, this collaborative process enlists professional school faculty as change agents. Such activity can bring about auricular development, show counseling’s value to student success, and perhaps most importantly, maximize student development and success.
Utility for Academic Units Working with At-Risk Students
Our results show construct integrity for several variables important to both the larger institution and the individual professional program. Granted, only prospective teachers are represented in this sample; so additional research is needed exploring these variables in other professional schools, but our results show strong promise for providing the I/C model with a framework with which to function effectively within professional programs. Counselors and their support staff now have three salient constructs (i.e., academic integration, social integration, career decidedness) with which to approach professional schools to begin a dialogue about collaboration between programs that will be mutually beneficial and helpful to students. Once professional preparation program staff and faculty understand the three constructs and how they are measured, they can then begin to build interventions with the help and support of the counseling staff. Interventions that will help students better understand salient variables in their success and thus help them make better decisions about their academic, social, and career paths include a strong referral network, retention trainings for all campus personnel, early warning alerts for struggling students, and on-going career planning and development opportunities for students.
Another important intervention to promote greater integration was described in a large-scale study by Braxton, et al. (2000). They found that when students perceived faculty to be well organized and wellprepared for their teaching and when faculty taught with instructional skill and clarity, social integration was enhanced. They outlined teacher behaviors that led to increased student integration. As counseling staff collaborate with academic units to recruit and retain students, this growing body of “best practices” can become an empirical foundation upon which to build student integration and professional collaboration. We are not advocating that counseling staff become instructional coaches; however, counseling staff could be excellent messengers for this sometimes sensitive information about instructor quality and wonderful sources of information about where to turn for help with instruction.
The significant and nonsignificant differences found between the at-risk and not-at-risk groups provide additional insights for counselors and professional school staff as they begin their work together. Non-significant findings will be focused on first. Interestingly, both groups were equally satisfied with their peer group relations. It appears that even at-risk students find friends and become socially integrated into college life at the level of having satisfying peer interactions. We are not asserting that exploring peer group relations with students is not an important component of interventions. What we are recommending here is that counselors and faculty be cognizant that in aggregate this variable did not discriminate between the two groups so it should be looked at the individual student level. If a student is exhibiting problems in school, peer relations may be a contributing factor and thus should be explored on an individual basis. Both groups were also similar in their interactions with faculty. Regrettably, however, all participants, regardless of group membership, reported relatively low amounts of interaction with faculty. This is problematic given the powerful role faculty can play in fostering academic and social integration in the institution. Counseling staff would be well advised to discuss this issue with professional school faculty and help monitor these variables as students move through the program. Respondents in this study were taking an introductory education course in the college and thus were just beginning their teacher preparation course work, so their low level of faculty interaction may be understandable given that most of the respondents were sophomores just coming out of their general course work that involved a number of large lecture classes. But if this low level of interaction with faculty persists throughout the college experience, strong evidence would accrue for significant curricular and advising modifications.
Significant findings also inform practice for faculty and counseling services personnel. Even though significant differences were found on 4 of the 7 factors and all effect sizes were in the moderate range, means analysis reveals that the institution’s response needs to dramatically improve. For example, the mean for the not-at-risk group for “Academic Integration – Faculty concern for student academic development” was only 4.05 on a 7 point scale. In other words, when students were asked to respond to the following two items, “Few of the faculty members I have had contact with are generally interested in students” and “Few of the faculty members I have had contact with are genuinely outstanding or superior teachers,” the mean responses hardly exceeded the midpoint of the scale. One would hope that faculty would be more highly rated on these critically important variables. In short, the factors represent omnibus evaluations of the institution, and as such awareness of them becomes critically important if faculty and staff are to work together to improve institutional outcomes. Opportunities for collaboration with other student affairs offices (e.g., Career Planning Services, programs providing academic support for at-risk students) are also relevant to the study’s results.
The I/C model coupled with relevant data from valid and reliable instruments can become a centerpiece for deep conversations about institutional strengths and weaknesses and how the former are to be enhanced and the latter remediated. The coupling of the model and the data can bring faculty and student services personnel together in ways that have not often occurred on college campuses. We find this potential exciting because of the mutual benefits that accrue for all stakeholders.
Archer, J. Jr., & Cooper, (1999). An initiator-catalyst approach to college counseling outreach. Journal of College Counseling. 2. 76- 88.
Archer, J. Jr., & Cooper, S. (1998). Counseling and mental health services on campus: A handbook of contemporary practices and challenges. San Francisco: Jossey-Bass.
Bieber, S. L. & Smith, D. V. (1986). Multivariate analysis of sensory data: A comparison of methods. Chemical Senses. 11. 19-47.
Bishop, J. B. & Walker, S. K. (1990). What role does counseling play in decisions relating to retention. Journal of College Student Development. 31. 88-89.
Braxton, J. M., Bray, N. J., & Berger, J. B., (2000). Faculty teaching skills and their influence on the college student departure process. Journal of College Student Development. 4L 215-227.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd Edition). Hillsdale, NJ: Lawrence Earlbaum Associates.
Coll, K. M., & Stewart, R. (2002). Collaboration between counseling services and an academic program: An exploratory study of student outcome. Journal of College Student Counseling. 5, 135-141.
Darling-Hammond, L. (2000). Solving the dilemmas of teacher supply, demand, and standards: How we can ensure a competent, caring, and qualified teacher for every child. New York, New York: National Commission on Teaching and America’s Future.
Dillman, D.A., (2000). Mail and telephone surveys (2nd Ed.). New York: John Wiley & Sons.
Fields, A. (2000). Discovering Statistics. Thousand Oaks, CA: Sage Publishing.
Gall, M., Borg, W., & Gall, J. (1996). Educational research: An introduction (6th edition). White Plains, NY: Longman.
Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1995). Multivariate data analysis (4th ed.). Upper Saddle River, New Jersey: Prentice Hall.
Kaiser, H. F. (1960). The application of electronic computers to factor analysis. Educational and Psychological Measurement. 20. 141- 151.
Pascarella, E. (2001). Using student self-reported gains to estimate college impact: A cautionary tale. Journal of College Student Development. 42. 488-492.
Pascarella, E. T., & Terenzini, P. T. (1991). How college affects students. Jossey-Bass: San Francisco. Pascarella, E. T., & Terenzini, P. T. (1983). Predicting voluntary freshman year persistence/ withdrawal behavior in a residential university: A path analytic validation of Tinto’s model. Journal of Educational Psychology. 75(2). 215226.
Pascarella, E. T., & Terenzini, P. T. (1979). Interaction effects in Spapy’s and Tinto’s conceptual models of college dropout. Sociology of Education. 61. 197-210.
Stevens, J. (1986). Applied multivariate statistics for the social sciences. Hilldale, NJ: Erlbaum.
Tinto, V. (1975). Dropout from higher education: A theoretical synthesis of recent research. Review of Educational Research. 45. 89- 125.
Tinto. V. (1993) Leaving College. Chicago: The University of Chicago Press.
KENNETH M. COLL
Department of Counselor Education
ROGER A. STEWART
Department of Literacy
Boise State University
Dr. Coll is Professor of Counselor Education; Dr. Stewart is Professor of Literacy at Boise State University. Please address correspondence to Dr. Ken Coll at 1910 University Drive, Boise ID 83725, 208-426-1821, email@example.com.
Copyright Project Innovation, Inc. Mar 2008
(c) 2008 College Student Journal. Provided by ProQuest Information and Learning. All rights Reserved.