______________ Amy Vliek, LMSW, is the Director of Admissions and Student Services in the School of Social Work and a current PhD student in the Interdisciplinary Health Sciences PhD Program at Western Michigan University. Kieran Fogarty, PhD is a professor of Interdisciplinary Health Sciences PhD Program at Western Michigan University. Robert Wertkin, DSW is the director and professor of School of Interdisciplinary Health Programs at Western Michigan University. Copyright © 2015 Advances in Social Work Vol. 16 No. 2 (Fall 2015), 390-405, DOI: 10.18060/20822 Are Admissions Models Working? An Analysis of MSW Admissions Models as Predictors of Student Success Amy Vliek Kieran Fogarty Robert Wertkin Abstract: Admissions models facilitate the selection of MSW candidates who are both academically prepared for graduate study and professionally suited to social work, characteristics often referred to as fit-to-the-profession. This study attempted to identify the relationship between the criteria used in the admissions model of a medium-sized MSW program and student success in the program. Specific criteria and the associated measurement tools within the model were evaluated in relation to student success. The study found that undergraduate grade point average is significantly associated with MSW grade point average while neither academic criteria nor fit-to-the-profession criteria were predictive of graduation. The study also offered new information regarding the admission of students with a criminal history. Students with a criminal history were eight times less likely to graduate than their non-offending peers, even when controlling for age, gender, race, and program type. The study supports the development of new fit-to-the-profession criteria with the possible elimination of the written statement. Keywords: Admissions criteria, student success, fit-to-the-profession Ideally, the admissions models used by Master of Social Work (MSW) programs ensure the selection of the best candidates for advanced social work practice. Admissions models facilitate the selection of MSW candidates who are both academically prepared for graduate study and professionally suited to social work, characteristics often referred to as fit-to-the-profession. In this manner, admissions models enable MSW education programs to meet many administrative and professional standards including adherence to requirements of the Council on Social Work Education (CSWE), values and ethics of the profession, and the gatekeeping role for the profession. In general, CSWE’s Educational Policies and Academic Standards (EPAS) require admissions models of MSW programs to be made public and to be easy to understand. Historically, CSWE required MSW programs to develop admissions models allowing only for the admission of the most highly qualified applicants who show evidence of fit-to-the- profession (CSWE, 1994). However, this language was removed from the EPAS in 2002 and replaced with less selective language (CSWE, 2002). The most recent EPAS standards use even less selective language, but still require that admissions models promote “an educational culture that is congruent with the values of the profession” (CSWE, 2015, pp. 14). The National Association of Social Worker’s (NASW) Code of Ethics also outlines and details the professional values of social work in the United States and its outlying territories (NASW, 2008). The NASW Code of Ethics states that social work is grounded ADVANCES IN SOCIAL WORK, Fall 2015, 16(2) 391 in core values including service, social justice, integrity, and competence (NASW, 2008). The admissions process is a built-in form of gatekeeping with a loosely defined process which MSW programs use to select the best candidates for study with an expectation that graduates will master the basic competencies for beginning-level MSW practice (Grady & S., 2009; Reynolds, 2004; Ryan & Habbis, 1997). Admissions models attempt to fulfill the gatekeeping mandate by providing a framework which facilitates the selection of the most qualified applicants. Gatekeeping is also grounded in the professional belief that social workers must protect current and future clients from harm (Leedy & Smith, 2005), guiding social work educators, practitioners, and professional organizations in safeguarding those served by the profession. Literature Review: Admissions Models and Criteria Although the research on admissions models is outdated, it provides some insight into admissions models and their effectiveness. The limited research supports the findings that MSW programs use similar admissions models which assess two primary areas: an applicant’s readiness for graduate study and an applicant’s fit-to-the-profession (Dailey, 1974, 1979; Dunlap, 1979; Dunlap, Henley, & Fraser, 1998; Fortune, 2003; GlenMaye & Oakes, 2001; Johnson, 1980; Pelech, Stalker, Regehr, & Jacobs, 1999; Pfouts & Henley, 1977; Shubert, 1963; Thomas, McCleary, & Henry, 2004). Fit-to-the-profession was typically defined by how well an applicant understands social work and demonstrates an ability to apply the NASW Code of Ethics. Undergraduate grade point average (UGPA) was also frequently reported to measure an applicant’s academic readiness for graduate study (Bogo & Davin, 1989; Dunlap, 1979; Dunlap et al., 1998; GlenMaye & Oakes, 2001; Pelech et al., 1999; Thomas et al., 2004). An applicant’s written statement, human services experiences, reference letters, and criminal background have been used to measure an applicant’s fit-to-the-profession and field performance (Dailey, 1979; GlenMaye & Oakes, 2001; Pelech et al., 1999; Schubert, 1963; Thomas et al., 2004). Table 1 summarizes the statistically significant correlations found in the twelve studies (signified with X) between specific criteria and student success in MSW programs (type of success denoted by asterisk, if no asterisk the study did not specify type of student success). Faculty rating is defined as the faculty rating of the fit-to-the-profession criteria. Undergraduate GPA (UGPA) As can be seen in Table 1, UGPA is the only criterion that has consistently correlated with student success in graduate school. More often than not, it is correlated with academic success, showing that as UGPA increases so does an individual’s academic success in the program (Bogo & Davin, 1989; Dunlap, 1979; Dunlap et al., 1998; Thomas et al., 2004). However, two of the studies reviewed did not differentiate between academic success and fit-to-the-profession (GlenMaye & Oakes, 2001; Pelech et al., 1999). The research supporting UGPA is consistent and generalizable due to the large sample sizes, lack of selection bias in the studies, and the examination of multiple academic years. The studies are spread over 25 years and span four decades (Bogo & Davin, 1989; Dunlap, 1979; Dunlap et al., 1998; Thomas et al., 2004). Pelech et al. (1999) summarized it well saying “undergraduate GPA continues to be one of the most valid predictors of subsequent academic performance and success in the overall program” (p. 219). Vliek et al./ARE ADMISSIONS MODELS WORKING? 392 Table 1. Statistically Significant Correlations Between Admissions Criteria and Student Success in MSW Programs Author UGPA Written Statement Reference letters Work Experience Faculty Rating Bogo & Davin, 1989 X* Dailey, 1974, 1979 X*** Dunlap, 1979 X* X Dunlap, Henley, & Fraser, 1998 X* Fortune, 2003 GlenMaye & Oakes, 2001 X X** X** Johnson, 1980 Pelech, Stalker, Regehr, & Jacobs, 1999 X X Pfouts & Henley, 1977 Shubert, 1963 X**** Thomas, McCleary, & Henry, 2004 X* X X *=Significantly correlated with academic performance **=Significantly correlated with field performance ***=Faculty ratings unreliable and inconsistent ****= For marginal students only Graduate Record Examination (GRE) The large majority of MSW programs do not currently require the GRE for admission or they require it at the university level and not the program level (National Association of Deans and Director of Social Work Admissions membership, personal communication, October 15, 2014). This trend reflects the concerns of social work educators about the bias of the GRE, the content validity of the reading comprehension section of the GRE, and evidence that the GRE is not predictive of performance in social work programs (Donahue & Thyer, 1992; Milner, McNeil, & King, 1984; Thomas et al., 2004). Two popular studies in social work influenced the trend. The first of these studies made the bold statement that the “GRE is largely irrelevant in social work education” (Milner et al., 1984, p. 949). The study compared two groups of students in one MSW program-- one group that was required to take the GRE before admission and one group that was not. The study found no significant difference in the students’ field performance, retention rates, or graduation rates. Milner et al. (1984) also pointed out that the admission rate of minority students increased from 9.8% in schools which required the GRE to 17.5% in schools that did not require the GRE. The second study questioned the validity of the reading comprehension portion of the GRE. This study showed that factors other than reading comprehension, such as guessing ability or general knowledge, greatly impact an individual’s score (Donahue & Thyer, 1992). Donahue and Thyer (1992) caution against using a tool that is shown to discriminate against minorities and does not correlate with MSW student’s performance in the program. ADVANCES IN SOCIAL WORK, Fall 2015, 16(2) 393 Written Statement The written statement is highly valued in social work admissions as it is the most commonly used tool for assessing fit-to-the-profession (Gibbs, 1994). However, there is a lack of consistency in the examination of written statements (Bogo & Davin, 1989; Dailey, 1979; Fortune, 2003; GlenMaye & Oaks, 2001; Pelech et al., 1999; Pfouts & Henley, 1977; Thomas et al., 2004). Many studies lump the analysis of written statements with the analysis of other criteria used to measure fit-to-the-profession making it difficult to discern the individual relationship between a student’s written statement and fit-to-the-profession performance (Bogo & Davin, 1989; Fortune, 2003; Pfouts & Henley, 1977; Pelech et al., 1999; Thomas et al., 2004). The rubrics for fit-to-the-profession criteria are also varied, and many studies showed inter-rater reliability was weak (Dailey, 1979; Fortune, 2003; GlenMaye & Oakes, 2001). This makes it difficult to examine the effectiveness of written statements in predicting an applicant’s fit-to-the-profession. Regardless of how the studies assessed the written statement as an admissions indicator, the results do not support its use as an effective criterion for fit-to-the-profession. Although eight studies examined written statements, only one found a correlation between it and fit-to-the-profession, which the study equated with field performance (Bogo & Davin, 1989; Dunlap, 1979; Dunlap et al., 1998; Fortune, 2003; GlenMaye & Oakes, 2001; Pelech et al., 1999; Pfouts & Henley, 1977; Thomas et al., 2004). One study found that faculty ratings of written statements did not increase the predictive value of the application, stating that the remaining criteria were as effective at predicting student success as the application as a whole (Fortune, 2003). The remaining studies either examined written statements as they related to academic performance or lumped written statements with other fit-to-the-profession criteria. Therefore, the results do not provide helpful information in discerning the effectiveness of written statements as a fit-to-the-profession criterion. Work and Life Experience Work and life experience, specifically in human services, is another criterion often used in MSW admissions models (Gibbs, 1994; Miller & Koerin, 1998; Vliek, Kothari, Huizen, & Curtis 2011) although research supports that it is inconsistent at predicting student success (Dailey, 1974, 1979; Dunlap, 1979; Dunlap et al., 1998; GlenMaye & Oakes, 2001; Johnson, 1980; Pelech et al., 1999; Shubert, 1963; Thomas et al., 2004). Previous experience has not been as widely studied as UGPA or written statements; however, it is commonly viewed by admissions professionals as a positive attribute of applicants (Johnson, 1980). Two studies found that work experience in the human services field significantly correlated with field performance; as an applicants’ experience in human services increases, their performance in field improves (GlenMaye & Oakes, 2001; Thomas et al., 2004). In contrast, a third study found that students who experienced problems in their MSW programs were more likely to have human services related work experience (Pelech et al., 1999). This was not true of students who had non-human-services work experience, even when the work time equaled the human-services-related students’ work time (Pelech et al., 1999). Vliek et al./ARE ADMISSIONS MODELS WORKING? 394 Faculty Rating/Interview Margaret Schubert (1963) was the first to publish an article about MSW admissions models. Her article was published in the Social Services Review and focused primarily on individual faculty ratings of admissions criteria, specifically the impact of interviews on changing faculty ratings. Schubert’s study is consistent with the other five published studies that examine faculty ratings and interviews. Four of the five studies found that faculty ratings and interviews were related to student academic success, showing that as faculty ratings increased, a student’s MSW GPA also increased (Dailey, 1974, 1979; Dunlap, 1979; Pelech et al., 1999; Schubert, 1963). However, the studies were not consistent in how or what the faculty rated and thus should be viewed with caution. The studies were all consistent in showing that faculty ratings and interviews did not correlate with fit-to-the- profession performance (Dailey, 1974, 1979; Dunlap, 1979; Fortune, 2003; Pelech et al., 1999; Schubert, 1963). This is important because faculty ratings align with fit-to-the- profession criteria and are meant to predict student success in professional performance. There is some evidence in the literature that faculty ratings are inconsistent and unreliable. Dailey’s 1979 study was a repeat of his 1974 study. The studies were done at different universities; however, they used the same study design which was developed using a published validation paradigm. Both studied admissions data over multiple academic years (Dailey, 1974, 1979). Both studies found that faculty ratings were unreliable and inconsistent with inter-rater reliability .47 in the earlier study and .22 in the second study (Dailey, 1974, 1979). Both scores are far too low to indicate any reliability in faculty’s ability to rate in a consistent manner. Dailey (1979) suggested the examination of faculty as a variable in admissions models stating that: …it seems reasonable to conceive of the decisional process as including faculty with high variability making judgments about students with equally high variability. Couple this with the apparent weakness and contradiction in admissions criteria and the suggestion that faculty make differential inferences from the same admissions material, and the decisional process is likely to be subject to a good deal of slippage. (p. 21) Although these studies are somewhat dated, these findings should be seriously considered given that admissions models have not changed substantially since 1979 and faculty still rank applications in the large majority of programs (Fortune, 2003). Criminal History While MSW programs consistently review criminal history, it is not used in a unified manner (Vliek et al., 2011). Criminal history is reviewed for two main reasons: as a gatekeeping tool and because of state and federal regulations regarding an individual’s ability to work as a professional social worker if said individual has a criminal history. In some states, individuals with a criminal background are limited by state law as to where and with what populations they can practice. For example, in Michigan individuals with a criminal background are held to the standards in Public Health Code 333.16221 which outlines the criminal convictions, such as criminal sexual conduct, assault, or obtaining, possessing or selling controlled substances, that limit employment for health care workers (Vliek, 2013). The variances in state laws cause MSW programs to review criminal history ADVANCES IN SOCIAL WORK, Fall 2015, 16(2) 395 in different ways. The program examined in this study does not deny admissions due to criminal history, but instead reviews an applicant’s history and informs the applicant of the impact said history may have on field placement, licensing, and future employment as a professional social worker (Vliek & Way, 2014). Literature Summary As can be seen in the above cited literature, there has been little consistency in the results of past studies regarding admission models in MSW programs. Only one criterion, UGPA, has consistently correlated with student academic performance (Dunlap et al., 1998; GlenMaye & Oakes, 2001 Pelech et al., 1999; Thomas et al., 2004). The majority of MSW programs have consistently ranked UGPA as the most important factor in admissions academic criteria (Bogo & Devin, 1989; GlenMaye & Oakes, 2001 McNeece, 1978). However, the majority of programs also include criteria that examine fit-to-the-profession, emphasizing the importance of assessing an applicant’s fit to social work (Dailey, 1974, 1979; Dunlap, 1979; Fortune, 2003; GlenMaye & Oakes, 2001; Johnson, 1980; Pelech et al., 1999; Pfouts & Henley, 1977; Shubert, 1963; Thomas et al., 2004). The research clearly indicates that the ability of admissions models to predict fit-to-the-profession needs improvement (Miller & Koerin, 1998). Further evidence-based study is imperative in order for programs to determine which measures to include in admissions models, enabling them to increase the effectiveness of gatekeeping, meet educational standards, and safeguard future clients. Research Questions In order to offer new information that will enhance the body of research evaluating the standard MSW admissions model, this study examined the following research questions: 1) Are academic-readiness admissions criteria associated with MSW academic performance?, 2) Are fit-to-the-profession admissions criteria associated with professional competence?, and 3) Are fit-to-the-profession and academic-readiness admissions criteria associated with graduation? Methods The purpose of this study was to identify the relationship between the admissions model used at a medium-sized MSW program and student success in the program. Specific criteria within the model were evaluated in relation to student success. Student success was measured by graduate grade point average, professional competence, and whether or not students successfully graduated from the program. Sample The current sample consists of all applicants to one MSW program from the academic year 2004/2005 through the academic year 2008/2009. The program had 657 applications during these years. One hundred thirty-five of these applications were left out of the final sample as they were denied admission, leaving a sample size of 522 admitted applicants. Vliek et al./ARE ADMISSIONS MODELS WORKING? 396 Measures We received Human Subjects Institutional Review Board (HSIRB) approval to examine data for students who enrolled in the academic years 2004/2005 through 2008/2009. Two sets of secondary data were analyzed. The school’s student information system provided data related to grades, grade point average, and graduation dates. The Social Work Admissions Office provided information on admissions criteria scores, professional review committee meetings, leaves of absence from the MSW program, and criminal history. The study examined five independent measures for each application, three of which were reviewed by faculty members, resulting in 4,557 separate ranking scores. Written statement scores ranged from 0 to 35, work and life scores ranged from 0 to 30, and reference scores ranged from 0 to 5, with higher scores indicating better performance. The admissions office evaluates the final two remaining independent measures. Undergraduate grade point average scores are calculated based on the last 60 credit hours and given a point value ranging from 0 to 20 with 0 equaling a GPA below 2.5 and 20 equaling a GPA between 3.89 and 4.0. Criminal history is a dichotomous variable (yes/no) and is asked on the application and evaluated by the admissions director. A section of the rubric including the rating of the first two sections of the written statement can be seen in Figure 1. Figure 1. Excerpt of Rubric for Admissions Decisions Written Statement: Maximum score is 30 (5 points per category) The reader rates the applicant’s coverage of required areas. Each part is rated 0–5. Readers are advised to begin with examining the personal statement with a rating of “average” and then examine whether this statement section is stronger or weaker than “average.” Category Points A) Understanding of problems and concerns relative to social work. Discussion absent; does not refer to social problems; or expresses misunderstanding of social problems � 0 Briefly identifies a social problem but does not provide detail � 1 Briefly identifies a social problem, and provides little/less than satisfactory depth of understanding � 2 Identifies details about a social problem and provides a basic/satisfactory understanding � 3 Identifies details about a social problem and provides an advanced/more than satisfactory understanding � 4 Articulates an exceptionally thorough and detailed understanding of 1 or more social problems and implications for social workers � 5 B) Understanding of diversity. Discussion absent; does not refer to diversity issues � 0 Does not identify a personal philosophy, or does not provide discussion or understanding of diversity issues � 1 Does identify a personal philosophy, but does not provide discussion or understanding of diversity issues � 2 Identifies a personal philosophy, and provides a satisfactory discussion and understanding of diversity issues � 3 Identifies a personal philosophy, and provides a more than satisfactory discussion and understanding of diversity issues � 4 Identifies a personal philosophy, and provides an exceptional discussion and understanding of diversity issues � 5 ADVANCES IN SOCIAL WORK, Fall 2015, 16(2) 397 The dependent measures reflect student success as defined by the school of social work. Problems in professional competence were measured by participation in a professional review committee (PRC). A PRC is the procedure employed by the school when a student shows evidence of problems with professional competence in either the classroom or field education. It is a dichotomous variable and a student is considered successful if they do not have a PRC. Masters of Social Work academic performance is a continuous measure ranging from 2.00-4.00. A higher GPA equates to student success. Graduation is a dichotomous variable, and students are considered successful if they graduate from the MSW program. The dataset contained demographic information on age, race, gender, and program type (Table 2). The age of accepted applicants ranged from 20 to 64 with a mean age of 30. Race had six categories, with 86% of enrollees identifying themselves as Caucasian and 14% identified in the remaining five categories. Eighty-five percent of the population was female, and 15% was male. Four percent of the population had a criminal history. Fifty- eight percent of the students in the population were part-time; 42% were full-time. Table 2. Demographic Data of Study Sample (n=522) Demographic Variable Mean(SD) Age 30 (9) Race n (%) Caucasian 433 (86%) Other 70 (14%) Female 445 (85%) Criminal History 14 (4%) Program Type Advanced standing 82 (16%) Part-time 282 (54%) Full-time 134 (26%) Part-time advanced 24 (4%) Table 3 presents the mean admissions scores for the 522 applicants. Table 3. Mean Admissions Scores of Study Sample (n= 522) Admissions Criteria Possible Score Range Minimum Score given Maximum Score given Mean (SD) Undergraduate GPA score 0-20 0 20 13 (5) Work & Life score 0-30 14.5 30 24 (3) Written statement score 0-35 15.5 35 28 (4) Reference score* 0-5 2 5 4.28 (.61) * 75% of references rated potential students as the highest possible score (5) Statistical Analyses Frequencies and descriptive statistics were obtained for all data. If potential confounders, such as age, gender, and race, were shown to have statistically significant Vliek et al./ARE ADMISSIONS MODELS WORKING? 398 correlations with the dependent measures, they were included in the final models. Research question one had a continuous dependent variable; therefore, a linear regression model was used for analysis. For the remaining questions both dependent measures were categorical variables; therefore, logistic regression was used. Results RQ1: Are academic-readiness admissions criteria associated with MSW academic performance? The sample size of 522 exceeds the minimal sample size of 400 using Green’s formula for the overall model, 50+8(k) where k represents the number of predictors (Green, 1991). The model met the assumptions of multicollinearity (tolerance=0.953 and VIF=1.05), independent errors (Durbin-Watson=1.590), and cross- validation (F Change=73.21, p<.001). The final model did show evidence of heteroscedasticity (Breusch-Pagan test=173.68, p<.001). Therefore, the regression analysis was run using heteroscedasticity-consistent estimators of the regression coefficient standard errors (HC3) as outlined by Hayes and Cai (2007). Table 4 shows the change in standard errors. Table 4. Changes in Standard Errors in Linear Regression Adjusted Beta OLSE SE OLSE P value HC3 SE HC3 P value Constant 3.542 .052 <.001 .059 <.001 Gender .036 .028 .260 .031 .240 Age -.001 .001 .058 .001 .313 Race -.22 .01 <.001 .042 <.001 GPA Score .018 .002 <.001 .002 <.001 The linear regression showed that undergraduate grade point average is significantly associated with MSW grade point average (F=12.96, p <.001). As the undergraduate grade point average increased by 1 unit (2.5 GPA points) the MSW grade point average increased by .02 (Table 5). RQ2: Are fit-to-the-profession admissions criteria associated with professional competence? Sample sizes for each criterion varied due to missing data. However, all sample sizes exceeded the minimal requirements per Green’s formulas for sample size for the overall models and individual predictors (Green, 1991). All models met the assumptions of linearity of the logit and multicollinearity. As seen in Table 5, there was no significant relationship between problems in professional competence and fit-to-the- profession criteria. RQ3: Are fit-to-the-profession and academic-readiness admissions criteria associated with graduation? Sample sizes varied due to missing data. However, all samples exceeded the minimal requirements per Green’s formulas for sample size for the overall models and individual predictors (Green, 1991). All models met the assumptions of linearity of the logit and multicollinearity. As reflected in Table 5, there was a significant relationship between criminal history and graduation (F=8.36, p=.018). Applicants without a criminal history were 8.4 times more likely to graduate than those with a criminal history. Also, in the crude model GPA score was significantly associated with graduation. However, the ADVANCES IN SOCIAL WORK, Fall 2015, 16(2) 399 association became non-significant in the adjusted model which accounted for race, gender, age, and program type. Table 5. Results of Linear and Logistic Regressions (n= 522) na Crude ß (95% CI)b p Adjusted ß (95% CI)c p RQ1: MSW academic performance UGPA 522 .2(.018-.022) <.001 .02(.013-.022) <.001 RQ2: Problems in professional competence Work & life 247 .79(.82-1.16) 0.824 .94(.78-1.14) 0.515 Statement 234 .96(.83-1.11) 0.583 .95(.82-1.11) 0.512 References 249 .87(.38-1.99) 0.746 .93(.36-2.38) 0.875 Criminal hx 322 1.49(.18-12.07) 0.71 1.79(.20-15.74) 0.601 RQ3: Graduation GPA 522 1.076(1.007-1.150) 0.03 1.055(.986-1.129)c 0.122 Work & life 247 .920(.805-1.052) 0.224 .961(.831-1.112)d 0.595 Statement 234 .911(.809-1.026) 0.124 .921(8.14-1.041)d 0.188 References 249 1.380(.385-4.949) 0.621 1.105(.289-4.221)d 0.884 Criminal hx 322 5.790(1.674-20.030) 0.006 8.363(1.434-48.764)d 0.018 a. Population size for regression. Regression population demographics are statistically similar to full population. b. CI=confidence interval c. Adjusted model accounts for age, race, and gender d. Adjusted model included race, gender, age, and program type Discussion The current study differs significantly from previous studies in four important ways. It is the first study since the 1970s to use a reasonably large sample size when analyzing admission criteria (see also Dunlap, 1979). Secondly, it looks at admissions over time by examining four academic years. Third, the admissions committee used the same rigorous scoring rubric for reviews during all four academic years, increasing the consistency of the admissions ratings. Lastly, the fit-to-the-profession criteria were evaluated by examining professional competence problems in multiple domains whereas current published studies only associated fit–to-the-profession criteria with field performance. The study offers new and important information regarding graduation rates for students with a criminal history. The study findings show that students with a criminal history are less likely to graduate than students without a criminal history, even when controlling for age, gender, race, and program type. Social work is a profession that values self- determination and rehabilitation. Future research investigating why students with a criminal history are less likely to graduate is important to ensure that social work education remains true to these values. The study findings are consistent with past literature examining the academic criteria used in admissions models. The study supported UGPA as an academic criterion with results showing that as an applicant’s UGPA increased, the applicant’s academic performance in the MSW program improved. This is consistent with seven of the twelve published studies which support UPGA as a predictor of MSW grade point average (Bogo & Davin, 1989; Dunlap, 1979; Dunlap et al,, 1998; GlenMaye & Oakes, 2001; Pelech et al., 1999; Thomas et al., 2004). Given the consistent support of UGPA as an effective predictor, MSW programs may want to consider relying more heavily on undergraduate Vliek et al./ARE ADMISSIONS MODELS WORKING? 400 grade point average. However, this should be done with great fortitude and care and with strict adherence to the social work values of social justice and CSWE’s educational competencies. The study findings do not support current fit-to-the-profession criteria as effective predictors of professional competence. In fact, findings strengthen the argument that fit-to- the-profession criteria are poor indicators of professional competence as consistent with nine of the twelve published studies (Bogo & Davin, 1989, Dailey, 1974, 1979; W. Dunlap, 1979; Fortune, 2003; Johnson, 1980; Pelech et al., 1999; Pfouts & Henley, 1977; Shubert, 1963). These findings, which are consistent over a fifty year period, challenge the widely held belief among social work educators that certain fit-to-the-profession criteria, including written statements and human services work experience, should be used to assess fit-to- the-profession (Johnson, 1980). Given the large amount of resources needed to evaluate these criteria-- faculty/reader time, development of scoring systems, time devoted by applicants to write statements, recommenders' time, it makes sense that we further examine how these measurements tools can be useful. Further, many studies showed that faculty ratings of criteria were, in and of themselves, inconsistent and ineffective at predicting student success (Dailey, 1974, 1979; Dunlap, 1979; Fortune, 2003; Pelech et al., 1999; Schubert, 1963). If programs are looking for inter-rater reliability, it may be wise for researchers to follow Dailey’s (1979) suggestion and examine faculty as a variable in the admissions process. If faculty members are to continue rating applications, then tools that allow for consistency and accuracy in faculty ratings must be designed. Strengths and Limitations There is limited research on admissions criteria for MSW programs, with only 12 published studies conducted since 1963. In an increasingly important area, this is the first study of its kind to have a sample size over 200, review applicants from four or more academic years, and to have the consistency of a rubric for admissions ratings. This study is also the first to examine fit-to-the-profession in both the classroom and field placements, allowing for a more comprehensive examination of fit-to-the-profession criteria. Because demographic and graduation data is not available regarding all MSW applicants, it is difficult to generalize the results of this study. It could be that the admissions process is working effectively and applicants who are not a good fit for social work are being denied admission. However, it was not the goal of this study to examine if admissions models were denying applicants appropriately but rather to examine if models were admitting students appropriately. Generalizability is also a limitation but was not a goal for this study. Instead, the study adds to the body of evidence that will offer evidence of the appropriateness and effectiveness of current admissions models. Recommendations In sum, the research indicates that fit-to-the-profession criteria consistently fail to predict student professional performance. MSW programs need to seek out effective criteria and their associated measurement tools for predicting professional performance and competence. Perhaps programs should consider building admissions models based on applicant competencies, specifically social work competencies. In 1977, Duehn and ADVANCES IN SOCIAL WORK, Fall 2015, 16(2) 401 Mayadas published an exploratory study examining a competency-based direct practice curriculum. They recommended five knowledge skills for use in the development of competency-based direct practice curriculum (Duehn & Mayadas, 1977). Although this study is related only to direct practice, it has implications for MSW admissions models as it could serve as an outline in how to develop competency-based admissions criteria. This change would be consistent with CSWE’s EPAS standards which are based on measurement of student competencies. It may be appropriate for MSW programs to examine competency-based admissions criteria and re-examine Duehn and Mayadas’s work. Another possibility is to model fit-to-the-profession criteria after Stein, Linn, and Furdon’s work (1975) which studied non-academic factors and intelligence as predictive factors of student professional performance. The study examined scores on three attitude scales and one intelligence scale. The study was small, with a sample size of 58; however, it yielded some interesting and relevant information. Overall, the study found that student's professional performance could be predicted by attitude ratings. Further, it showed that attitude ratings that are desirable to the social work profession (i.e., less dogmatic, less authoritarian, and less alienated) are associated with higher intelligence. Interestingly, the study also found that students who are closed-minded become less interested in social work values as their careers progress. The study provides some information regarding effective fit-to-the-profession criteria, which warrants study replication. Johnson (1980) supports retaining written statements as an admissions criteria despite the fact that her research did not find any significant correlations between written statements and student performance. In fact, none of the publications reviewed for this study support written statements as valid criteria (Bogo & Davin, 1989; Dailey, 1974, 1979; Dunlap, 1979; Dunlap et al., 1998; Fortune, 2003; GlenMaye & Oakes, 2001; Johnson, 1980; Pelech et al., 1999; Pfouts & Henley, 1977; Shubert, 1963; Thomas et al., 2004). Johnson’s suggestion is based on the fact that admissions committee members felt better about their recommendation when the written statement was included in the application. The author can only surmise that this is the main reason written statements continue to be included in admission applications. However, as we move toward evidence-based admissions models, we must acknowledge that models should not include criteria that have continuously been shown as ineffective. Perhaps written statements can serve a different purpose in admissions models. It is possible that written statements could contain ‘red flags’ for students who are more likely to have problems related to professional competence. More studies replicating Pelech et al.’s (1999) regression model may discover evidence that would give justification for continuing to use written statements in admissions models. However, it may be time to consider more drastic changes to the admissions models currently used by most MSW programs. It may be time to acknowledge that our fit-to-the- profession criteria are not effective and we need to start over. Other healthcare professions use well-studied admissions models, some of which contain fit-to-the-profession criteria. A meta-analysis of other health care admission models, focused specifically on fit-to-the- profession criteria, may help social work educators build an effective model. Many health programs, including nursing, physician’s assistant, and speech pathology, use centralized Vliek et al./ARE ADMISSIONS MODELS WORKING? 402 application systems (CAS). The CAS require all U.S. applicants to use the same base application. These systems are not designed to improve the effectiveness of admissions models. However, having a national application may increase social work educators’ interest in improving the admissions model. Research into the impacts of a CAS is certainly warranted. Perhaps fit-to-the-profession cannot be adequately measured before admission. Social work programs can focus more diligently on fit-to-the-profession criteria while students are matriculating. This would allow programs to examine, over time and in a comprehensive manner, a student’s fit-to-the-profession. Through licensing requirements, the social work profession acknowledges that social work education alone does not fully prepare someone for social work practice. Fit–to-the-profession requirements can also be added to licensure requirements. For example, as part of the application procedure, an applicant could be required to provide evidence of the application of the NASW Code of Ethics in their practice. In an increasingly competitive applicant pool, it is important to have evidence-based admission models to guide decision-making. Cole (1991) addresses arbitrary admissions decisions as one of the largest legal issues faced by social work programs. Cole argues “there should be some evidence that there are relationships between admission standards, likely program success, and alumnae ethical and responsible conduct as professionals” (p. 23). Born and Carroll (1988) emphasizes the need to be ethical in the process of admitting students to MSW programs, which includes using admissions models that are evidence- based. The literature indicates that current admissions models do not meet these standards. This leaves educational institutions at risk for legal action, and more importantly, it leaves social work clientele vulnerable to practitioners who are not competent and do not adhere to professional social work values. Further research examining admissions models will ensure that schools of social work are using evidence-based admissions models that effectively meet CSWE standards, ensure gatekeeping, and reflect core principles of the profession. References Bogo, M., & Davin, C. (1989). The use of admissions criteria and a practice skills measure in predicting academic and practicum performance of MSW students. Canadian Social Work Review, 6(1), 95-109. Born, C. E., & Carroll, D. J. (1988). Ethics in admissions. Journal of Social Work Education, 24(1), 79-85. Cole, B. (1991). Legal issues related to social work program admissions. Journal of Social Work Education, 27(1), 18-24. Council on Social Work Education [CSWE]. (1994, 2002, 2015). Educational Policies and Academic Standards. Alexandria, VA: CSWE Press. Dailey, D. (1974). The validity of admissions predictions: Implications for social work education. Journal of Education for Social Work, 10(2), 12-19. doi: http://dx.doi.org/10.1080/00220612.1974.10672304 http://dx.doi.org/10.1080/00220612.1974.10672304 ADVANCES IN SOCIAL WORK, Fall 2015, 16(2) 403 Dailey, D. (1979). The validity of admissions predictions: A replication study and implications for the future. Journal of Education for Social Work, 15(2), 14-22. doi: http://dx.doi.org/10.1080/00220612.1979.10671561 Donahue, B., & Thyer, B. A. (1992). Should the GRE be used as an admissions requirement by schools of social work? Journal of Teaching in Social Work, 6(2), 33- 40. doi: http://dx.doi.org/10.1300/J067v06n02_04 Duehn, W. D., & Mayadas, N. S. (1977). Entrance and exit requirements of professional social work education. Journal of Education for Social Work, 13(2), 22-29. doi: http://dx.doi.org/10.1080/00220612.1977.10671431 Dunlap, W. (1979). How effective are graduate social work admission criteria? Journal of Education for Social Work, 15(3), 96-102. Dunlap, K. M., Henley, H. C. J., & Fraser, M. W. (1998). The relationship between admissions criteria and academic performance in an MSW program. Journal of Social Work Education, 34(3), 455-462. Fortune, A. E. (2003). Comparison of faculty ratings of applicants and background characteristics as predictors of performance in a MSW program. Journal of Teaching in Social Work, 23(1), 35-54. doi: http://dx.doi.org/10.1300/J067v23n01_04 Gibbs, P. (1994). Gatekeeping issues in BSW programs. Arete, 19(2), 15-27. GlenMaye, L., & Oakes, M. (2001). Assessing suitability of MSW applicants through objective scoring of personal statements. Journal of Social Work Education, 38(1), 67-82. Grady, M., & Mr. S. (2009). Gatekeeping: Perspectives from both sides of the fence. Smith College Studies in Social Work, 79(1), 51-64. doi: http://dx.doi.org/10.1080/00377310802634616 Green, S. (1991). How many subjects does it take to do a regression analysis? Multivariate Behavior Research, 26(3), 499-510. doi: http://dx.doi.org/10.1207/s15327906mbr2603_7 Hayes, A., & Cai, L. (2007). Using heteroscedasticity-consistent standard error estimators in OLS regression: An introduction and software implementation. Behavior Research Methods, 39(4), 709-722. doi: http://dx.doi.org/10.3758/BF03192961 Johnson, J. (1980). Admission decisions and student selection in social work education: A study of values and writing ability as predictors of educational performance in a graduate school of social work. Social Work Abstracts, 17(3), 189-204. Leedy, G. M., & Smith, J. E. (2005). Felony convictions and program admissions: Theoretical perspectives to guide decision-making. Journal of Social Work Values and Ethics, 16(34), 1-17. McNeece, C. A. (1978). Admissions policies and procedures in graduate schools of social work. Arete, 5(3), 143-155. http://dx.doi.org/10.1080/00220612.1979.10671561 http://dx.doi.org/10.1300/J067v06n02_04 http://dx.doi.org/10.1080/00220612.1977.10671431 http://dx.doi.org/10.1300/J067v23n01_04 http://dx.doi.org/10.1080/00377310802634616 http://dx.doi.org/10.1207/s15327906mbr2603_7 http://dx.doi.org/10.3758/BF03192961 Vliek et al./ARE ADMISSIONS MODELS WORKING? 404 Miller, J., & Koerin, B. (1998). Can we assess suitability at admission? A review of MSW application procedures. Journal of Social Work Education, 34(3), 437-453. Milner, N., McNeil, J.S., & King, S.W. (1984). The GRE: A question of validity in predicting performance in professional schools of social work. Educational and Psychological Measurement, 44, 945-950. doi: http://dx.doi.org/10.1177/0013164484444015 National Association of Social Workers. (2008). Code of ethics of the National Association of Social Workers. Washington, DC: NASW Press. Pelech, W., Stalker, C. A., Regehr, C., & Jacobs, M. (1999). Making the grade: The quest for validity in admissions decisions. Journal of Social Work Education, 35(2), 215- 226. Pfouts, J. H., & Henley, H. C. (1977). Admissions roulette: Predictive factors for success in practice. Journal of Education for Social Work, 13(3), 56-63. doi: http://dx.doi.org/10.1080/00220612.1977.10671457 Reynolds, L. (2004). Gatekeeping prior to point of entry. Advances in Social Work, 5(1), 18-31. Ryan. M., & Habbis, D. (1997). Guarding the gates of the profession: Findings of a survey of gatekeeping mechanisms in Australian bachelor of social work programs. Australian Social Work: The Journal of Australian Association of Social Workers, 50(3), 5-12. doi: http://dx.doi.org/10.1080/03124079708414092 Schubert, M. (1963). Admissions decisions. Social Service Review, 37(2), 154-165. doi: http://dx.doi.org/10.1086/641418 Stein, S., Linn, M. W., & Furdon, J. (1975). Predicting social work student performance. Journal of Education for Social Work, 10(3), 85-92. doi: http://dx.doi.org/10.1080/00220612.1974.10671307 Thomas, M., McCleary, R., & Henry, P. (2004). Effectiveness of admission criteria on student performance in classroom and field instruction. Advances in Social Work, 5(1), 33-46. Vliek, A. (2013). Critical and ethical analysis of the Michigan Public Health Code Act 368 Of 1978 Section 333.16221: Should criminal convictions be considered in social work licensing? Manuscript in preparation. Vliek, A., Kothari, K., Huizen, D., & Curtis, A. (2011, November). Criminal history and social work admissions decisions: recidivism-related factors or social work-related factors. Poster presented at the 57th Annual Program Meeting for CSWE, Atlanta, GA. Vliek, A., & Way, I. (2014). Western Michigan university school of social work admissions committee manual of policies and procedures: 2014-2015. Unpublished document, School of Social Work, Western Michigan University, Kalamazoo, MI. http://dx.doi.org/10.1177/0013164484444015 http://dx.doi.org/10.1080/00220612.1977.10671457 http://dx.doi.org/10.1080/03124079708414092 http://dx.doi.org/10.1086/641418 http://dx.doi.org/10.1080/00220612.1974.10671307 ADVANCES IN SOCIAL WORK, Fall 2015, 16(2) 405 Author note Address correspondence to: Amy Vliek, MSW, BSW, School of Social Work, Western Michigan University, 4430 Health and Human Services Building, Mail Stop 5354, 1903 W Michigan Ave, Kalamazoo MI, 49008-5354. Email: amy.vliek@wmich.edu mailto:amy.vliek@wmich.edu Keywords: Admissions criteria, student success, fit-to-the-profession Literature Review: Admissions Models and Criteria Methods Sample The current sample consists of all applicants to one MSW program from the academic year 2004/2005 through the academic year 2008/2009. The program had 657 applications during these years. One hundred thirty-five of these applications were left out of ... Statistical Analyses Results Discussion Strengths and Limitations