1 EXAMINING THE INFLUENCE OF RESEAR CH MENTORING AND TRAINING MODEL ON CLINICAL AND COUNSELING GRADUATE STUDENTS SCHOLARLY ACTIVITY. By GEOFFREY ANTON LEE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009
2 2009 Geoffrey Anton Lee
3 To my family and friends, who supported me throughout this process
4 ACKNOWLEDGMENTS I extend special thanks to m y wife and all of my friends and family for their constant love and support. I am also grateful to my committee for their continual guidance. I could not have completed this dissertation project without the guidance of my extraord inary supervisor, Dr. Greg Neimeyer.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4LIST OF TABLES ...........................................................................................................................7ABSTRACT ...................................................................................................................... ...............8 CHAP TER 1 INTRODUCTION .................................................................................................................. ..92 LITERATURE REVIEW .......................................................................................................11Research Training Environment ............................................................................................. 11Research Mentoring ................................................................................................................14Academic Training Model ......................................................................................................193 METHODS ....................................................................................................................... ......24Participants .................................................................................................................. ...........24Measures ...................................................................................................................... ...........24Demographic information ....................................................................................................... 28Procedure ..................................................................................................................... ...........304 RESULTS ....................................................................................................................... ........32Descriptive and Preliminary Analyses .................................................................................... 32Measurement Reliability ....................................................................................................... ..32Regression .................................................................................................................... ...........33Hypothesis 1 and 2 .......................................................................................................... 33MANOVA ....................................................................................................................... 355 DISCUSSION .................................................................................................................... .....42APPENDIX A RESEARCH TRAINING ENVIRONMENT SCALE-REVISED SHORT FORM (RTES-R) ...................................................................................................................... ..........52B SELF-EFFICACY IN RESE ARCH MEASURE (SERM)..................................................... 54C THE RESEARCH OUTCOME EXPECTATIONS QUESTIONNAIRE (ROEQ) ............... 55D SCIENCTIST-PRACTITIONER INVENTORY ...................................................................56E RESEARCH MENTORING EXPE RIENCES SCALE (RMES) ........................................... 57
6 F PAST ATTITUDES TO WARDS RESEARCH ..................................................................... 60G SCHOLARLY ACTIVITY SCALE (SAS) ............................................................................ 61LIST OF REFERENCES ...............................................................................................................62BIOGRAPHICAL SKETCH .........................................................................................................67
7 LIST OF TABLES Table page 4-1 Intercorrelations, Means, and St andard Deviations of Variables ......................................394-2 Summary of Multiple Regression Analys es for Variables Predicting Scholarly Activity (N =205) ...............................................................................................................404-3 Differences in Satisfacti on between Training Models ....................................................... 414-4 Differences in Perceptions of RTE and Scholarly Activity between Training Model .......41
8 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EXAMINING THE INFLUENCE OF RESEAR CH MENTORING AND TRAINING MODEL ON CLINICAL AND COUNSELING GRADUATE STUDENTS SCHOLARLY ACTIVITY. By Geoffrey Anton Lee August 2009 Chair: Greg Neimeyer Major: Counseling Psychology This study focused on graduate student resear ch training through expl oration of research mentoring experiences and training model as pr edictors of clinical and counseling psychology graduate students scholarly activity. Findings re vealed research mentoring and training model served as significant predictors of students scholarly activity. Training model served to differentiate students based on their training experiences and sc holarly activity. While, students from practitioner-scholar programs reported greater satisfaction with the clinical emphasis of their training, they reported le ss positive perceptions of their research training environment, and reported lower scholarly activity than students from science-practitioner or clinical-scientist programs. This study adds to the developing lit erature concerning gradua te student research mentoring and training model with in professional psychology.
9 CHAPTER 1 INTRODUCTION Research training of professi onal psychology doctoral students has received a fair am ount of attention in the last 2 decades (Hollingsworth & Fassinger, 2002). This attention developed, in part, from recognition that despite counseling and clinical psychologys commitment to the development of competent practitioners and productive scientists (Szymanski, Jovanovic Ozegovic, Phillips, & Briggs-Phillips, 2007), most applied psychologists pursue practiceoriented careers (Fitzgerald & Osipow, 1988). For example, 71% of counseling psychology graduates report initial employment in human serv ice settings (APA, 1996). With such clinically oriented placements, it is not surprising that few clinical or counse ling psychologists conduct research after completing their degrees (B rems, Johnson, & Gallucci, 1996). Reports among counseling psychologists suggest research constitutes less than 10% of their time and does not generally lead to publication (Fitzgerald & Osipow, 1988; Watk ins et al. 1986). Arguably, these findings reflect lower scientific output than is expected by endorsement of a scientist-practitioner model of training, espousing equal emphasis on re search and practice (Kahn & Scott, 1997) and raises the question of what needs to be done to change this trend. Despite the current nature of the field, increasing student and graduate involvement in scholarly activity is a desira ble goal (Kahn, 2001) for several re asons. Research represents a primary tool in informing practice (Belar, 2000) and serves to advance professional psychology (Belar, 2000; Gelso, 1979, 1993; Gelso & Lent, 2000). Because of the importance of quality research to the applied fields of psychology, increased attention to factors of graduate training that contribute to students scholarly activity is needed (Gelso, & Lent 2000). Thus, the present study was designed to augment recent efforts in th e research training literature by 1) examining the influence of research ment oring and training model on scholar ly activity, and 2) broadening
10 the examination of research training environments to include scholarly activity among clinical and counseling psychology gradua te students. The subsequent section will review existing literature in these areas to establish an em pirical warrant and hypothe size relationships among the variables of interest in this investiga tion. First, findings from the research training environment literature will be reviewed. We will th en examine the available research on the role of mentoring relationships in contributing to scholarly activity. Finally, we will conclude the review with recent findings examining training models and their contributions to scholarly activity.
11 CHAPTER 2 LITERATURE REVIEW The factors that influence studen t involvemen t in scholarly activity have been gaining increasing attention. Generally, research has found the interaction of individual factors such as personality and interests and c ontextual factors such as the Research Training Environment (RTE) exert a significant influen ce on attitudes toward research and subsequent scholarly activity (Bishop & Bieschke, 1998; Gelso, 1997; Kahn, 2001; Kahn & Scott, 1997; Krebs, Smither & Hurley, 1991; Mallinckrodt, Gelso, & Royalty, 1990; 2002). Howe ver, several authors have suggested broadening the examination of factor s influencing scholarly activity within the literature to include other releva nt factors such as faculty me ntoring (Betz, 1997; Gelso, 1997; Hill, 1997; Hollingsworth & Fassinger, 2002) an d training model science-practice emphasis (Szymanksi et al. 2007). Research Training Environment Gelso (1979, 1993,1997) proposed exam ining the influence of Research Training Environments (RTE) in contribu ting to graduate student trai ning outcomes. Gelso (1993, 1997) proposed RTE theory to explain how contextual f actors impact students research interest and scholarly activity. This perspe ctive suggests that academic training environments often lack critical research training necessary to facilitate positive student attitudes toward research (i.e., interests) and scholarly activity (Kahn & Scott, 1997). Gelso (1997) describes effective RTEs as represented by both instructional and interpersona l components critical to scholarly activity (e.g., publication & professional presentations). While instructional elements encourage idea generation and research implementation, interpersonal elements encourage involvement in minimally threatening research early in training, provide support and encouragement of students efforts, focus on the collaborative nature of research, and utilize f aculty role-modeling of
12 appropriate scientific beha viors (Gelso, 1993, 1997). Accord ing to RTE theory, these interpersonal and instructional components work together to in crease scholarly productivity both directly and indirectly through enhancing research self-efficacy and research interest. More specifically, Gelso (1993) proposed that effective RTEs influe nce scholarly activity in two primary ways: they enhance student interest in research by demonstrati ng the exciting elements of research and increase students research self -efficacy. Thus, research interest and self-efficacy mediate the relationship between RTE and scholar ly activity (Gelso, 1993). This conception has been consistently supported by research findings suggesting students perceptions of their RTE predict interest in a variety of research ac tivities (Bishop & Bieschke 1998; Kahn & Miller, 2000; Kahn & Scott, 1997; Mallinckrodt, Gelso & Royalty, 1990; Royalty, Gelso, Mallinckrodt, & Garrett, 1986). Surveys of graduate students also provide support for RTE enhancing students research self-efficacy (Bishop & Bieschke, 1998; Gelso, Mallinckrodt, & Judge, 1996; Kahn & Miller, 2000; Kahn & Scott, 1997; Phillips & Russell, 1994). Thus, effective RTEs facilitate research interest and self-efficacy among students, which encourages greater involvement in research (Kahn & Scott, 1997; Kahn, 2001). Research interest and self-efficacy have also been shown to predict scholarly activ ity for current students (Kahn, 1997; Krebs et al. 1991; Phillips & Russell, 1994) and gradua tes (Royalty & Magoon, 1985; Royalty & Reising, 1986). While the influence of research interest and research self-efficacy on scholarly activity has become increasingly well documented, authors have also begun to consider the role of research outcome expectations (Bishop & Bieschke, 1998; Kahn, 2001) and prior levels of research interest (Gelso, 1997; Hollingsworth & Fassinger, 2002) in predicting scholarly activity. For example, Kahn and Scott (1997) proposed th at individual differences (e.g., personality components) and contextual factors (e.g., environmental supports and ba rriers) interact to
13 influence research self-efficacy (e.g., confidence in being able to effectively complete research task) and research outcome expectations (e.g., ex pected consequences of doing research: getting published). Thus, the more confiden t a student feels regarding doing research, the more they will perceive positive outcomes regarding research involvements (Kahn, 2001). Kahn and Scott (1997) suggested research self-efficacy and rese arch outcomes expectations lead to research interests, with research efficacy and interest ultimately predicting scholarly activity. Numerous studies have provided support for these predic tions (Bieschke, Bishop, & Herbert, 1995; Bieschke, Herbet, & Bard, 1998; Kahn & Miller, 2000; Phillips & Russell, 1994). Although it has gone largely unaccounted for in recent res earch training literature, Gelso (1997) and Hollingsworth and Fassinger (2002) ha ve suggested individual diffe rences in prior levels of research interest may also influence students later research productiv ity. The role of prior research interest in predicting scholarly activity is yet unknown. La stly, Taylor and Neimeyer (in press) have highlighted the importance of mentor ing in graduate student training outcomes such as scholarly activity and training program satis faction. The authors not ed several findings relevant to the current investig ation. They found that the amount of time spent with ones mentor was positively associated with greater scholarly activity (e.g. publications and presentations) and students overall level of satisfaction was positively associated with particular types of mentoring (e.g. socioemotional and instructio nal). The later finding lead the authors to suggest that the nature of ones mentoring experience may promote overall satisfaction with their graduate training program. While noteworthy and informativ e, these findings have limited generalizability to students from programs with a strong research emphasis. A broader examination of students from professional psychology graduate programs w ith varying levels of commitment to practice and science might serve to inform the literature regarding how these differences affect scholarly
14 activity. Despite this limitation, these findings offer encouraging support for mentoring as a contributor to scholarly activity. Research Mentoring While the relationship of RTEs to research training outcomes has received substantial attention in the literature, several authors (e.g., Betz, 1997; Gelso & Lent 2000; Hill, 1997; Hollingsworth & Fassinger, 2002; Kahn, 2001; Taylor & Neimeyer, in press) have suggested broadening the examination to consider the influence of faculty mentoring on scholarly activity. Hollingsworth and Fassinger (2002) have sugge sted faculty mentoring emerges as a consistently important undercurrent in the rese arch training environment (p. 324). This is consistent with reports regarding the influence of mentoring on graduate training more generally. Graduate students suggest havi ng a mentor is a crucial component of graduate training (Atkinson, Neville, & Casa, 1991; Lark & Crot eu, 1998; Luna & Cullen, 1998). Authors have also noted the profound effect mentoring can have on a variety of key student outcomes including: the mentees professional identity a nd skills development (Elman, Illfelde-Kaye, & Robiner, 2005; Vespia, 2006), docto ral training satisfaction (Clark et al., 2000), and salary and career satisfaction (Hume & Johns on, 2003). With regards to rese arch mentoring, Royalty and Reisling (1986) reported advisor-ad visee interaction regarding resear ch activities was one of the strongest positive influences on students research interest. Others have noted student reports highlighting their relationship with faculty memb ers as a critical elem ent in their research training (Gelso, 1997; OBrien, 1995 ). Among minority students su rveyed, reports suggest the mentors encouragement of student in research involvement is crucia l (Atkinson et al. 1991). Faculty research mentoring has also been shown to influence greater research involvement and scholarly activity among current psychology stud ents and recent graduates (Cronan-Hillix, Gensheimer, Cronan-Hillix, & Davi dson, 1986; Galassi, Brooks, Stol tz & Trexler, 1986; Krebs
15 et al. 1991). Cronan-Hillix et al. (1986) found mentoring was significantly relate d to several measures of research productivity. Kahn and Sco tt (1997) proposed that faculty mentoring may serve as a mechanism to enhance research intere st and research self-efficacy. Hill (1997) has compared the research mentori ng relationship to the counselor-c lient working alliance, leading her to suggest the mentoring relationship may be a crucial element in the research training environment, deserving a more explicit focus in RTE literature research. Gelso (1993), acknowledging the impact of faculty research ment oring, identified a series of interpersonal and instructional faculty behaviors to help establish a positive research mentoring relationship. Gelso (1997) and Hollingsworth (2000) have noted the influence of the mentoring relationship on students self-efficacy and research attitudes. Despite authors suggesting the positive benefits of mentoring on stude nt research training, only two studies examining mentor ing as a predictor of scholarly activity have been conducted. In one study, Kahn (2001) extended the work of Kahn and Scott (1997) by measuring the mentoring relationship as part of the research training environment. To assess the mentoring relationship, Kahn (2001) utilized the Mentori ng Functions Scale (Noe, 1988), which measures students perceptions of the adequacy of their relationship with their ment or. Results in this study did not support student perceptions of the mentor ing relationship as a significant predictor of scholarly involvement. However, these findings s hould be considered with caution, given several important considerations. Surveys of graduate students suggest many students do not have an identifiable mentoring relati onship (Cronan-Hillix et al.1986 ; Smith & Davidson, 1992). This may be due to lack of agreement in the mentor ing literature regarding what constitutes a mentor (Schlosser & Gelso, 2001), which furt her attenuates the st rength of this relationship with no clear definition of the role of mentoring in research. Consistent with this co ncern, Kahns (2001) study
16 lacked specificity regarding this role. He notes possible confusion on the part of students on who to choose as a mentor, students were free to iden tify any faculty member as a mentor, and it is possible that many identified a person who has been a strong mentor in their growth as a practitioner, but has not stimulated interest or efficacy as a prac titioner with respect to research activities (p. 353). Consistent w ith this point, Kahn (2001) did not report the nature of sciencepractice emphasis of the sample. It is possible that for many students, even if they had an established mentoring relationship, their pr ograms do not emphasize research training or scholarly activity. A likely outco me, given 40% of the participants in Kahn and Scott s (1997) study reported not having to write a thesis dur ing their doctoral traini ng. Without measurement of the programs science-practice emphasis, it becomes difficult to accurately assess their comparative level of commitment to scholarly activity. Further complicating assessment, even within the same program, advisors may differ in their science-practice commitments (Mallinckrodt, 1997). In other c onsiderations, a gender effect may have impacted the studys results. Kahn (2001)s sample was predominantly female and previous research has suggested females perceive research training environments less favorably and experience less research selfefficacy than men (Kahn & Scott, 1997; Landino & Owen, 1988). These perceptions may interact with faculty reactions to further impact the mentoring relationship through faculty members willingness to invest energy in research training (Schlosser & Gelso, 2005). Lastly, Kahns (2001) predictive model had an insuffici ently large sample given the large number of parameters estimated. Failing to meet the mi nimum number of observations (e.g., 5-10 per estimated parameters, Bentler & Chou, 1987), thus reducing the ability to adequately measure these relationships. Given these findings and th e inherent limitations regarding Kahns (2001) study, further work is needed to ex amine the mentoring relationship.
17 In the only other study of research ment oring among counselin g psychology doctoral students, Hollingsworth and Fassing er (2002) attempted to provide a conceptual bridge between research training environment and mentoring relationships, ex amining research mentoring experiences as predictors of upper division counseling psychology doctoral students. A link believed to be critically impor tant to understanding the resear ch training process (Betz, 1997; Gelso, 1997; Hill, 1997; Hollingsworth & Fassinge r, 2002). The authors also included prior research interest and training model as additional variables be lieved to influence scholarly activity. The inclusion of these variables has been supported by findings suggesting higher levels of research interest differentiate between mo re and less productive c ounseling psychologists (Royalty & Magoon, 1985) and training model has been found to lead to differential scholarly productivity for clinical and counseling student s (Cherry et al. 2000; Ne imeyer, Saferstein, & Rice, 2005; Szymanski et al. 2007). Hollingsworth and Fassinger (2002) developed the Research Mentoring Experiences Scale (RMES) that broa dly measured students me ntoring experiences, and more specifically measured two aspects of the mentoring relationship: psychosocial (e.g., affective aspects and interpersona l elements of the faculty-student research training relationship) and career mentoring (e.g.,. mentors efforts to facilitate acquisition of necessary skills to complete research tasks successfully). These re lationally based concepts were analogous to Gelsos broader, programmatically based instrumental and interp ersonal elements of effective RTEs. Hollingsworth and Fassinger (2002) proposed that the research mentoring relationship would mediate the relationship between research training environment and scholarly activity, while controlling for students prior interests might attenuate the relationship between RTEs and self-efficacy and research producti vity. Finally, they investigated the potential moderating effect of gender and training program science-practice emphasis on scholarly productivity. Consistent
18 with previous research, findings supported resear ch training environment, research self-efficacy, and past research attitudes as direct predictors of scholar ly productivity. While gender and training model were not significant predictors of productivity in their study, mentoring experiences were found to be a significant pr edictor of scholarly activity. The mediating relationship of students ment oring experiences between R TE and productivity was also supported. These findings support proposals regarding the importance of res earch mentoring as a critical element within res earch training environments (e .g., Betz, 1997; Gelso, 1997; Gelso & Lent, 2000; Hill, 1997) and provide tentative empirical evidence rega rding the facilitative nature of research mentoring in stude nts research training (Holling sworth & Fassinger, 2002). The authors note two additional cons iderations regarding implications of their findings: 1. Consistent with previous proposals (e.g., Gelso, 1997; Mallinck rodt, 1997), the strong correlation between research training environments and research mentoring may suggest that strong research training programs can facilitate strong research mentor ing relationships. 2. However, the meditational nature of research mentoring suggests it serv es as a vehicle through which RTE exerts the greatest influence on students scholarly product ivity. Thus, while a training programs sciencepractice commitments may serve to define the tr aining program culture present during training, the relative influence may be enhanced or hindered by the nature of the faculty-student relationship. These interpersonal elements (e.g., role modeling, support, and guidance) can exert a powerful influence, above and beyond RTE, on student attitudes toward research and subsequently, their scholarly activity (Mallinckrodt, 1997). While more conclusive than Kahn (2001) re garding the importance of mentoring as a critical ingredient of research training in contributing to sc holarly productivity, Hollingsworth and Fassingers (2002) findings sugg est the need for further inves tigation. Several limitations of
19 this study warrant acknowledgement. Most notably, Hollingsworth and Fassinger (2002) reported scale modifications that may qualify their findings. The authors reported concerns regarding multicollinearity be tween the RTES-R and RME and utilized a modified 16-item RTES. The authors also reported us e of shortened version of the scholarly activity scale (SAS). Thus, the association reported by Hollingsworth and Fassinger (2002) may represent different constructs than previous research. Other limitations include failure to include research outcomes expectations, which have been supported as mediating the relationship between RTE and scholarly productivity (Bishop & Bieschke, 1998; Kahn, 2001).Similarly, additional examination of training model is needed to determine its role in research productivity. Lastly, our confidence in the generalizability of these findings may be enhanced by broadening the investigation to include both clinical and counseling graduate students. But first, we will briefly review research supporting the importance of academic training m odels in scholarly productivity for these students. Academic Training Model While RTEs have been of interest to resear chers attem pting to understand the nature of scholarly activity and other res earch training outcomes, investigations of training models and their associated science-practice emphasis within psychology graduate pr ograms (Belar & Perry, 1992; Hoshmand & Polking-Horner, 1992; Murdock, Alcorn, Heesacker, & Stotlenberg, 1998; Stoltenberg, et al. 2000) may offer additional insight into graduate student scholarly activity. For example, Galassi, Brooks, Stoltz, & Trexler ( 1986) surveyed training directors from APA accredited programs and found more productive programs (based on actual research published or presented) tended to involve students in research earlier in training, were more likely to require research participation, and placed greater emphasis on philosophy of science. These findings
20 suggest the importance of a programs chosen training model, but before reviewing more recent work, a brief review of the conceptual founda tions of these models will be discussed. Generally, training models offer a conceptual framework that serves to guide a particular programs professional training focus, expecta tions, and outcomes (Neimeyer, Saferstein, & Rice, 2005; Neimeyer, Rice, & Keilin, 2007). Traini ng models are primarily distinguished by the nature and degree of their commitment to scient ific and professional ende avors (Neimeyer et al. 2005). Notably, there is considerab le variability in programs id entification with a particular training model within the clinic al and counseling psychology speci alties. On one hand, there are programs primarily or exclusively dedicated to a professional training model. These training programs, based on the practitioner-scholar mode l; focus on developing professional skills, and serving as consumers of scientific knowledge in the service of professi onal practice. Thus, the primary emphasis in practitioner-scholar programs is providing clinic ally oriented services that are effective and responsive to the varied needs of those served (e.g., individual, community, society) (McHolland, 1992). On th e other hand, the clinical-scien ce training model represents a primary or exclusive emphasis on scientific trai ning (Neimyer et al. 2005). These programs are characterized by a commitment to empirical appro aches to evaluating the validity and utility of testable hypotheses and to a dvancing knowledge by this met hod (Academy of Psychological Clinical Science, 2004). Despite these appare nt differences, nearly 98% of professional psychology doctoral programs endorse to a scie ntist-practitioner training model based on the Boulder Model (OSullivan & Qu evillon, 1992). The science-practi tioner model encourages the integration of science and clin ical practice knowledge to effectively inform the work of professional psychologists (Jones & Mehr, 2007). It also attempts to foster a career-long process of psychological investigation, assessment, and intervention (Belar & Perry, 1992). Notably,
21 even within science-practitioner programs, the degree of commitment to implementing tenets of the science-practitioner model varies widely, with programs emphasizing one or the other of their tandem commitments. Given these differences, several investigator s have begun examining the relation between training model and scholarly activity for clini cal and counseling psyc hology training programs (Cherry et al., 2000; Gaddy, Charlot-Swilley, Ne lson, & Reich, 1995; Neimeyer et al., 2005; Neimeyer et al., 2007). For example, Cherry et al. (2000) examined training outcomes for clinical psychology programs. The authors found that clinical-science and science-practice programs reported higher levels of scientific prod uctivity (e.g., greate r publications and presentations) among faculty and students. In contrast, practitioner-scholar programs demonstrated greater levels of professional se rvice delivery. Findings al so revealed employment setting and activities were in line with progr am emphasis. These findings represent outcomes generally consistent with program orientation (Cherry et al. 2000). Similarly, Neimeyer et al. (2005) found faculty and students in practice-oriented programs we re less likely to publish in professional or scientific journa ls or present at professional meetings, although programs did not differ in relation to faculty and students i nvolvement in professi onal service delivery. In considering results of these studies t ogether, Neimeyer et al. (2005) suggested noteworthy differences in regard s to outcome between the two sp ecialties within professional psychology. For example, counseling psychology pr ograms differences were generally with regards to scientific outcomes, while differences extended to professiona l outcomes in clinical psychology. Cherry et al. (2000) f ound practice oriented clinical programs reported significantly higher levels of service delivery than science-pr actice or clinical-science. Given these findings, Neimeyer et al. (2005) highlighted the need for further examination of the influence of training
22 model in scholarly outcomes with professional psychology. Th e training model literature collectively offers important implications for clinical and counseling scholarly productivity, however they are limited by primary focus on ei ther clinical or couns eling. Notably, only one study has examined the nature of training model and research outcome simultaneously for clinical and counseling psyc hology, although no study has yet to examine these outcomes for graduate students. In the only study to date to examine the in fluence of research tr aining environment and training model on scholarly productivity, Szymansk i et al. (2007) investig ated the impact of RTEs on scholarly activity of early career pr ofessionals. Szymanski et al. (2007) found the training model matters in the scholarly productiv ity during academic and internship training. Participants from programs emphasizing a science-practitioner training model reported significantly higher scholarly productivity than those from pr actitioner-scholar programs. Interestingly, perceptions of academic research training environment were not influenced by the nature of their programs training model. In interpreting these findings, the authors suggested that while it seems reasonable to assume di fferences would exist across academic RTEs, a possible explanation may be that participants perceptions of their RTE matched with their expectations. Consistent with pr evious work, the relationships among research training variables proposed in RTE theory was generally supporte d in Szymanski et al.s (2007) study. RTE, research self-efficacy, research outcome expectati ons, and research interest were all significantly related to scholarly productivity. Although, results of the path analysis specifying these relationships suggested that only research self-efficacy and research interest predicted scholarly activity. Consistent with meditational proposals from previ ous work, RTE did significantly account for variance in self-efficacy and RTE and self-efficacy significantly predicted outcome
23 expectations. Szymanksi et al. (2007) did not include training model in predicting scholarly activity. In concluding their wo rk, Szymanski et al. (2007) recommended the inclusion of additional variables such as rese arch mentoring in future models to explore their influence on scholarly activity. The current investigation attempts to adva nce the literature on research training by exploring the role that resear ch mentoring relationships and training model emphasis play in predicting scholarly productivity for clinical and counseling psychology students. The current investigation will also examine key differences in graduate student experiences and scholarly activity as a function of traini ng model. Consistent with the above-mentioned literature, the following hypotheses guide this work : 1. Research training environment, research ment oring relationships, research self-efficacy, level of research interest, level of practice interest, and research outcome expectations will predict clinical and counseling st udents scholarly productivity. 2. Training model will be positively related to scholarly activity and will serve as an additional significant predictor of scholarly act ivity for clinical and counseling students. 3. Students report of research, clinical, and ove rall programmatic satisfaction will differ as a function of training model. 4. Students from more practice-oriented pr ograms (e.g., practitioner-scholar) will report similar overall perceptions of their RTEs and lower levels of scholarly productivity than more research oriented programs (e.g., clinical -scientist) for both cl inical and counseling students. 5. The research mentoring relationship will me diate the relationship between research training environment and scholarly productivity for clinical and counseling students.
24 CHAPTER 3 METHODS Participants Letters requ esting participation in a survey were e-mailed to APAaccredited clinical and counseling programs. The solicitatio n letter stated that the survey was available to all graduate students with the purpose of exploring the natu re of scholarly activity among professional psychology graduate students. Tr aining directors were asked to disseminate the survey to graduate students. Participati on was voluntary and all participants were re quired to read and confirm their agreement to an informed consent form prior to participa ting in this study. All inventories were completed online and submitted to an online database. It took participants approximately 15 minutes to complete the instruments, and the study was conducted in accordance with APA ethical guidelines. Measures Perceptions of Research Training Environment: were measured by the 18-item Research Training Environment ScaleRevi sed (RTESRS; Kahn & Miller, 2000). This measure is a short form of the 54-item RTES R (Gelso et al., 1996) which includes items representative of the nine ingredients of the research training environment described by Gelso (1993, 1997). Each ingredient is measured by two items, rated on a scale ranging from 1 (disagree) to 5 (agree). Sample items include I have felt encouraged during my training to find and follow my own scholarly interests and Our faculty seems interested in understanding and teaching how research can be related to counse ling practice. Total scores for RTESRS range from 18 to 90. Higher scores represented more positive perceptions of the research training environment. Internal consistency values of the RTESRS reported range from .85-.88 (Kahn & Miller, 2000; Kahn, 2001). The RTESRS correlates .96 with the 54-ite m RTESR (Kahn &
25 Miller, 2000). Convergent validity is supporte d by positive correlations between RTESRS scores and measures of research self-efficacy and interest in scientist ac tivities (Kahn & Miller, 2000). Research Self-efficacy: was measured by the 12-item version of the Self-Efficacy in Research Measure (SERM; Phil lips & Russell, 1994) develope d by Kahn and Scott (1997). The 12-item SERM is characterized by items measuri ng four domains of research self-efficacy. Each domain is measured by 3 items: research desi gn skills (e.g., formulating hypotheses), practical research skills (e.g., keeping r ecords during a research project ), quantitative and computer skills (e.g., understanding computer printouts), and writing sk ills (e.g., writing the introduction and literature review for a dissert ation). Students completing the SERM indicated their confidence, from 0 (no confidence) to 9 (total confidence), in successfully performing each task or their belief that they possess the skill to complete each task. Scores range from 0 to 108, with higher scores suggesti ng greater research self-efficacy. The 12-item SERM has good internal consistency ranges .89-.90 (Kahn & Scott, 1997; Kahn, 2001). Kahn and Scott (1997) reported positive correlations between the 12-item SERM and measures of scholarly activity and perceptions of the resear ch training environment. Research Outcome Expectations: was measured using The Research Outcome Expectations Questionnaire (ROEQ; Bishop & Bi eschke, 1998). On the ROEQ, students rate 17 potential outcomes associated with doing re search on a 5-point Likert scale ranging from strongly disagree (1) to strongly agree (5). Sample items include Involvement in research will enhance my job/career opportunities, Involvement in research will allow me to contribute to practitioners' knowledge base, and Involvement in research will take time away from leisure activities Scores range from 17-85, with hi gher scores suggesting more positive research
26 outcome expectations. Internal consistenc y ranges from .88-.90 (Kahn, 2001) (Bishop & Bieschke, 1998) (Bieschke et al., 1995). Bish op and Biseschke (1998) report that ROEQ correlates positively with interest in research, pe rceptions of the research training environment, research self-efficacy, and investigative interests. Science versus Practice interest: was measured with the Science-Practice Inventory. SPI 20 is a 20-item scale that measures interest in 10 scientific and 10 pr actitioner activities in psychology using a 5-point scale ranging from 1 (very low interest ) to 5 ( very high interest ). In previous work, the internal consistency for the SPI scores was .92 (scientist) and .69-.85 (practitioner). Validity evidence has been demonstrated previously through a positive correlation with a measure of career interest in a sample of psychology graduate students ( Leong & Zachar, 1993). Research Mentoring: was m easured with the Research Mentoring Experiences Scale (RMES), a measure created by Hollingswort h and Fassinger (2002) based on comparable instruments developed for business settings (e.g., Noe, 1988b; Ragins & McFarlin, 1990). Hollingsworth and Fassinger (2002) report that the RME includes two subscales. The first subscale, Psychosocial Mentoring, measures the affective aspects of research training, and focuses on the interpersonal elements of the faculty-student relationship. Participants rate 13 items investigating the extent to which their f aculty mentor provides em otional support, respect and personal regard, and models positive research attitudes. The second subscale, Career Mentoring, examines the mentorss efforts to help students acquire necessary skills to complete research tasks successfully. This subscale includes 16 items representing faculty members' teaching of research skills, advice giving, a nd the provision of research opportunities. Instructions in the current st udy will ask students to rate thei r relationship with the faculty
27 member whom they considered most important in their current doctoral research training. Possible responses ranged from 1 (faculty member pa ys very little attention to ) to 5 (faculty member pays a great deal of attention to ). Student responses are summed and divided by the number of items to generate a total score. Po ssible scores range from 1 to 5. Hollingsworth and Fassinger (2002) reported a Cronbach's alpha of .74. Past Attitudes Toward Research: was measured by four items constructed by Royalty et al. (1986). Royalty et al. (1986) measured couns eling psychology students' recalled interest in conducting research prior to thei r enrollment in the doctoral pr ogram. The items included the following: 1. I would have preferred to have the option of completing my doctoral training without being required to complete research projects (Preference) 2. I had a strong interest in doing research (Interest), 3. I placed a high value on the place of research in my future career (Value), and 4. Participating in research activi ties after graduation was not a major priority for me (Priority). Students rate th eir agreement on a 5-point Likert scale, which ranged from 1 (strongly disagree) to 5 (strongly agree), the first and last items were reverse-scored. Responses are summed and divided by the number of items to produce a final score, having a potential range from 1 to 5. Research supports the intern al consistency, with alpha coefficients ranging from .87 to .90 (Gelso et al., 1996; Hollingsworth & Fassinger, 2002; Kahn & Scott, 1997; Royalty et al., 1986). Test-retest has been reported at .93 (Roya lty et al., 1986). Dependent variables. The dependent variable scholarly activity was measured utilizing Kahn and Scotts (1997) 9-item measure, the Scho larly Activity Scale (SAS ). According to Kahn (2001) these items measure both past accomplishm ents (e.g., number of manuscripts published) and the current research activities (e.g. whether st udent is currently collecting data). Thus, this measure provides a broad examination of student involvement in resear ch-related activities
28 including data collection and analysis, manus cript development, presentations, research convention attendance (Kahn & Scott, 1997). Students respond to each item by reporting a number of activities they are engaging for each ar ea of research involvement. Responses to all items summed to obtain a total scholarly activity index with higher scor es reflecting greater scholarly activity. Intern al consistency coefficients (K-R-20) for this scale for Kahn and Scott (1997) was .68, Kahn (2001) .70; Hollingsworth & Fassinger (2002) .75; Szymanski et al. (2007) was .79. Kahn and Scott (1997) report this measure pos itively correlates with interest in research and science relatedness of students career goals. The dependent variable Program Satisfaction was measured through a series of 3 likerttype questions developed for the cu rrent investigation due to a lack of suitable existing measures. These items measured clinical and counseling psychology students' satisfaction with their training programs clinical, research emphasis, as well as, their overall satisfaction with their training program experience. The items were repres ented by the following example: Please rate your SATISFACTION with your programs emphasi s on RESEARCH on a 5-poi nt Likert scale, ranging from 1 (completely unsatisfied ) to 5 (completely satisfied). Demographic information The sam ple consisted of 215 graduate student s from clinical and counseling psychology graduate programs (170 female, 45 male) with a mean age of 28.61 ( SD = 6.59). Of the 32 programs represented, the mean number of st udents per program for counseling was 6.72 ( SD = 3.10), and 7.93 ( SD = 4.92). The overall mean number of students was 7.25 (SD = 3.25). The sample was primarily Caucasian, 79.1% ( N = 170), followed by Asian American, 7.0% ( N = 15), Hispanic, 5.6% ( N = 12), African American, 4.2% ( N = 9), Multiracial, 1.9% ( N = 4), Other, 1.9% ( N = 4), and Native American, 0.5% ( N = 1).
29 There was a fairly even split of Clinical 47% ( N = 101) versus Counseling 53%( N = 114) graduate students, with pr imarily PhD programs, 83.7% ( N = 180) and some PsyD programs 16.3% ( N = 35). Graduate students repor ted their year in their doctora l program, with first year students representing 20.9% (N = 45) of the sample, second year 19.5% ( N = 42), third year 12.6% ( N = 27), fourth year 20% ( N = 43), fifth year 15.3% ( N = 33), and sixth year students representing 11.6% ( N = 25) of the sample. Results of the Pearsons Chi Square examining year in program and training model revealed year in program did not vary as a function of training model designation, X (10, N = 215) = 17.72, p = .06. Students were asked to pick one of the fo llowing that BEST DESCRIBES the degree of science-practice emphasis of your training program. The majority of students ra ted the degree of science-practice emphasis in their graduate programs to be science-practitioner, 71.6% ( N = 154), followed by practitioner-scholar, 20.5% (N = 44), and clinical scientist, 7.9% ( N =17). Examination of actual science-practice emphasis of participants gradua te programs based of training director ratings from The Insiders Guide to Clinical and Counseling Psychology (Norcross, Sayette, & Mayne, 2008) revealed a sli ghtly different picture of graduate programs science-practice emphasis: sc ience-practitioner, 42.3% ( N = 91), followed by practitionerscholar, 29.8% ( N = 64), and clinical scientist, 27.9% ( N =60). Students were additionally asked to rate their own personal rese arch versus clinical emphasis on a 7-point Likert type scale ranging from 1 (clinically oriented ) to 7 (research oriented). Stude nts reported a mean score of 3.35 ( SD 1.46). For purposes of the study, participan ts were classified by science-practice emphasis based on training directors ratings in an attempt to address individual differences in participants views of their programs.
30 Procedure In an attem pt to recruit a participant sample th at was representative of the participants of interest, training programs were se nt the online survey containing an informed consent, a brief demographics sheet, and the aforementioned m easures. A total of 50 professional psychology programs within clinical and counseling were initially identified for inclusion in the study. Programs were selected on the basis of the follo wing criteria: identificat ion as a clinical or counseling psychology doctoral program, espousing either a scholar-practitioner, sciencepractitioner, or clinical-scien tist, training model, having basi c recognition with its respective specialty area, and geographical location (e.g. Nort heast, Midwest, Southeast etc). The latter criteria attempted to provide geographical divers ity so as attempt avoid regional bias. Of the programs solicited, the specialties were roughly equally repres ented with 27 clinical and 23 counseling programs. Attempts were also made to obtain a stratified sa mple consisting of equivalent number of programs (e.g. 8 scholar-pract itioner, 8 scienc e-practitioner, and 8 clinicalscientist training programs) within each specia lty. Participants from 32 programs of the 50 programs solicited completed the survey. Of thes e 32 programs, 14 identified as clinical and 18 identified as counseling. Program matic training model was derived from training director ratings included in The Insiders Guide to Clinical and Counseling Psychology (Norcross et al. 2008) and stratification followed procedures utilized by Neim eyer et al. (2005) and Neimeyer et al. (2007). Training directors were asked to review the study to assess its usefulness and appropriateness for their program. Upon their review, they were asked to forward our study to their students. Thus, participants were graduate st udents recruited via email so licitation according to program membership. Participants were asked to read and electronically sign the informed consent form. Once the participant agreed to pa rticipate, they were asked to click on link to the survey. Once
31 participants completed the surveys and submitted their responses, they were thanked for their participation.
32 CHAPTER 4 RESULTS Results fro m the current study are described below. First, I will discuss measurement issues, followed by sample descriptives, gene ral correlations, regres sion analyses, MANOVA analyses, and finally results of the mediation analysis. Descriptive and Preliminary Analyses Power analyses were conducted on the hypothese s in the current study to ensure inclusion of sufficient number of participants to estimate parameters. For a multiple regression analysis with six predictors, a medium effect size ( 0.15), and Alpha level of 0.05, and strong power (0.95), a total sample size of 143 participants was needed. This was the most participants that were needed for each of the 5 hypotheses. Multivariate analyses of vari ance were conducted on the 5 meas ures used in the current study (RTES, SERM, ROEQ, SPI, RMES, PATR, SA S) and suggested that there were no differences in the mean scores of the variable s of interest among th e questionnaire forms (all ps > .004). Thus for all analyses, the data from different forms were combined. Measurement Reliability Measurem ent reliabilities for the RTES, SERM, ROEQ, SPI, RMES, PATR, and SAS scores appear in Table 4-1. Notably, Chronbachs coefficient alpha for the RTES of .88, was much higher than the .74 reported by Hollingsworth and Fassinger (2002). The reliability finding for the SERM was an alpha of .91. This is comparable to previous reports ranging from .89-.90 (Kahn and Scott, 1997; Kahn, 2001). Reliability findings for the ROEQ revealed a Chronbachs coefficient alpha of .93, which is also comparable to the internal consis tency ranges found from .88-.90 (Kahn, 2001; Bishop & Bieschke, 1998; Bieschke et al., 1995).
33 The measurement reliability findings for the SPI were an alpha of .87 for the scientist subscale and an alpha of .83 for the practice subscal e. This is similar to findings in previous work, where the internal consistency for th e SPI scores was .92 (scientist) and .69-.85 (practitioner) ( Leong & Zachar, 1993). Measurem ent reliability for the RMES reveal ed a Chronbachs coefficient alpha of .95, which was much higher than the .74 repor ted by Hollingsworth and Fassinger (2002). Reliability findings for the PATR revealed a Chronbachs coefficient alpha of .91 for the overall score. Research supports the internal co nsistency, with alpha coef ficients ranging from .87 to .90 (Gelso et al., 1996; Hollingsworth & Fa ssinger, 2002; Kahn & Scott, 1997; Royalty et al., 1986). Finally, reliability for the SAS revealed a Ch ronbachs coefficient alpha of .69. This was comparable to previous research indicated intern al consistency coefficients (K-R-20) for this scale for Kahn and Scott (1997) .68, Kahn (2001) .70; Hollingsworth & Fassinger (2002) .75; and Szymanski et al. (2007) .79. Kahn and Scott (1997) report this measur e positively correlates with interest in research and scien ce relatedness of st udents career goals. Regression Hypothesis 1 and 2 For the firs t two hypotheses a multiple regre ssion analysis was conducted. First, Person Product Moment correlations, using a criterion level of .05 (1-tailed), were computed between the criterion variable (SAS) and each of the predictor variables in an attempt to confirm that the relationships were in the predicted directions. Pearson Pr oduct Moment correlations were conducted between the SAS and the RMES, r = .34, p < .001, the SERM, r = .35, p < .001, the PATR, r = .32, p < .001, the ROEQ, r = .15, p < .016, science practice orientation r = .35, p < .001, and SPI science, r = 0.25, p < .001, revealing positive significant correlations in the
34 predicted directions. In contrast, a negative significant correlation o ccurred between SAS and SPI practice r = -0.31, p < .001. The correlation between SAS and the RTES was conducted to verify that participants with more positive perceptions of their research training environments reported greater scholarly activity. Results were not in the predicted di rections, revealing an insignificant correlation betw een the SAS and the RTES, r = 0.10, p = ns Thus, RTES was not included as a predictor in the re gression analysis for hypothesis 1. Review of the correlation matrix also revealed multicollinearity concerns due to significant correlations between PATR and ROEQ, r = .75, p < .001, and PATR and SPI science r = .80, p < .001 suggesting these variables shared 56% and 64 % of their variance re spectively, with PATR. Therefore, PATR was removed from the analysis as it did not serve as primary hypothesis within the study (See Table 4-1). In order to assess the capacity of the data to be in line with the normality assumptions of multiple regression, the data was subjected to te sts of skewness and kurtosis. Results of these analyses indicated that the assumptions for mu ltivariate normalcy were met. All skewness and kurtosis estimates for the variables fell between 2 and except for the Scholarly Activity Scale, which had a kurtosis value of 2.551. Examination of th e data revealed 8 outliers. As noted in the literature, outliers can have dele terious effects on statistical anal yses including: increasing error variance, reducing power, altering the odds of Type I and Type II errors, and influencing estimates of interest (Osbourne & Overbay, 2004). Thus, for the dependent variable (SAS) 4 outliers were removed based on the generally accep ted convention of 3 standard deviations from the mean (Osbourne & Overbay, 2004). As a further safe guard, alpha levels were also protected by conducting Bonferroni corrections (dividing the conventional alpha of .05 by the number of criterion variables), resulting in a more conservative test of the hypotheses.
35 Hypothesis 1 examined RTE, research ment oring relationships, research-self-efficacy, research interest, practice interest, and research outcome expectations as predictors of clinical and counseling students scholarly productiv ity. Hypothesis 2 included training model orientation as an additional pred ictor of scholarly activity. Thus for the first two hypotheses, a simultaneous multiple linear regression analysis was conducted to determine if these variables were significant predictors of the criterion variab le (scholarly activity). The predictor variables accounted for significant variation in scholarly productivity scores, F (6, 205) = 11.62, p < .001 ( R = .254). The standardized beta coefficients for the RMES ( = .137), t (205) = 1.96, p = .05, the SERM ( = .190), t (205) = 2.65, p = .009, and training model ( = .210), t (205) = 3.14 p = .002, were significant and in the pos itive direction. The direction of these effects indicated that the more the students endorsed research self-e fficacy, quality mentoring relationships, and the more scientifically oriented their program, the greater their report of scholarly activity. In contrast, the standardized beta coefficient for the SPI practice ( = -0.393) was significant and in the negative direction for the SAS, t (207) = -2.82, p = .005. The direction of the effect for SPI practice was in expected directi on, and was conceptually consistent indicating that the more the students endorsed practice-oriented interests, the lower their report of scholarly activity. The standardized beta coefficients SPI science and ROEQ did not reach significance (See Table 4-2). MANOVA For hypothesis 3, (Students report of resear ch, clinical, and overall programm atic satisfaction will differ as a function of trai ning model) a MANOVA was conducted with the independent variable being trai ning model as measured by the Insiders Guide to Clinical and Counseling Psychology (2008) Following Neimeyer et al. (2 005) and Neimeyer et al. (2007), Programs with a practice orient ation (values 1-3) were placed in the practice-oriented group, those with an equal emphasis (value of 4) were placed in the science-pr actice oriented group and
36 participants with a higher scien ce orientation (values 5-7) were placed in the science-oriented group. The dependent variables were participants ratings of their sa tisfaction with their programs emphasis on research, practice, and th eir overall satisfaction with their graduate program. Results of the MANOVA were significant F (3, 210) = 6.74, p < .001. Follow-up ANOVAs revealed significant differences in progra m science-practice orientation for satisfaction with programs emphasis on clinical practice F (3, 209) = 13.58, p < .001. Bonferroni Post-Hoc analyses revealed that participants with a pr actice orientation reported significantly more satisfaction with their programs emphasis on clinical practice ( M = 4.39) as compared to those with science-practice ( M = 3.99) and science orientation ( M = 3.47) (See Table 43). Participants from science-practice programs also reported signif icantly more satisfaction with their programs emphasis on clinical practice than those with a science orientat ion. Significant differences were not found for participants ratings of their satisfaction with pr ograms research emphasis and overall satisfaction with training program. For hypothesis 4, (Students from more practi ce-oriented programs will report similar overall perceptions of their Research Training Environments (RTEs) and lower levels of scholarly productivity (SAS) than more resear ch-oriented programs for both clinical and counseling students), the indepe ndent variable in the MANOVA was program training model (e.g. Practice, Science-Practice, or Science Orie ntation), with the depe ndent variables being participants perceptions of their Research Training Environments and Scholarly Activity. Results of the MANOVA were significant F (2, 209) = 10.41, p < .001. Follow-up ANOVAs revealed significant differences for training mo del along both dependent variables, Research Training Environment, F (2, 209) = 7.01, p < .001 and Scholarly Activity F (2, 209) = 15.92, p < .001. Participants with a practi ce orientation had significantly more negative perceptions ( M =
37 60.87) of research training environments as compared to those with science ( M = 66.40) and science-practice orientations ( M = 68.14), Bonferroni Post-Hoc analyses. There was not a significant difference found between science and science-practice program orientations. For scholarly activity similar findings occurred, with participants in programs with a greater practice orientation having significantly lower re ported levels of scholarly activity ( M = 10.45), compared to science-practice ( M = 18.83) and science ( M = 22.60) oriented programs Bonferroni Post Hoc analyses. No significant differe nces were found between scien ce and science-practice program orientations. These results were in the predicte d directions. To ensure that there was not a confound between clinical and counseling psyc hology student in these analyses, a MANOVA was conducted to assess for signi ficant differences between cl inical and counseling psychology students scores on the dependent variable s. Results revealed that the MANOVA, F (1, 210) = 1.11 and ANOVAs for type of program (clinical vs. counseling psychology) along research training environment, F (1, 210) = 0.20 and scholarly activity, F (1, 210) = 1.89, were insignificant p = ns (See Table 4-4). Lastly, hypothesis 5 assessed whether the re search mentoring relationship mediates the relationship between research tr aining environment and scholarl y productivity fo r clinical and counseling students. Per Baron and Kenny (1986), the following criteria were necessary for mediation: (I) the predictor (research training en vironment) is significantly associated with the outcome (scholarly activity); ( II) the predictor is significantly associated with the mediator (research mentoring relationship); (III) the mediator is associat ed with the outcome variable (with the predictor accounted for); and (IV) the add ition of the mediator to the full model reduces the relation between the predictor and criterion variable. These gui delines for mediation were not met to test the influence of research training environment on scholarly activity via research
38 mentoring relationship due to the lack of a si gnificant relationship be tween research training environment and scholarly activity ( r = .103, ns ).
39 Table 4-1. Intercorrelations, Means, a nd Standard Deviations of Variables Variable 1 2 3 4 5 6 7 8 ROEQ RMES 0.37** RTES 0.48** 0.45** SERM 0.33** 0.34** 0.48** PATR 0.75** 0.32** 0.35**0.41** SAS 0.15* 0.26 0.10 0.35**0.32** SPI Science 0.68** 0.32** 0.42**0.47**0.80**0.25** SPI Practice -0.26** -0.05 -0.06 -0.21**-0.38**-0.30** 0.24** Training model 0.23** 0.35** 0.18**0.31**0.27**0.35**0.25 -.019 M 72.04 3.15 65.37 75.35 3.31 17.44 27.99 38.874.09 SD 13.29 1.12 12.33 17.60 1.21 13.12 7.67 6.211.38 Note. N = 213 p < .05.* p < .01.**
40 Table 4-2. Summary of Multiple Regression An alyses for Variables Predicting Scholarly Activity (N =205) Scholarly Activity Variable B SE B RMES 1.61 .82 .14* SERM .14 .05 .19** Training model 1.74 .55 .21** ROEQ -.15 .21 .16 SPI Science .19 .15 .11 SPI Practice -.50 .14 -.23* R .50 R2 .25 F 11.62 *p < .05. **p < .01.
41 Table 4-3. Differences in Satisf action between Training Models PractitionerScholar (n =64) SciencePractitioner (n =91) ClinicalScientist (n = 60) Satisfaction M SD M SD M SD F Effect Size ( 2) Observed power Clinical 4.39 .88 3.99 1.05 3.47 1.00 13.58* .114 .998 Research 3.59 1.06 3.82 1.05 3.73 1.09 .998 .008 .200 Overall 4.09 .87 3.91 .80 3.79 .79 2.74 .025 .537 F (3, 209) = 13.58, p < .001.* Table 4-4. Differences in Perceptions of RTE and Scholarly Activity be tween Training Model PractitionerScholar (n =62) SciencePractitioner (n =90) ClinicalScientist (n = 60) Variable M SD M SD M SD F Effect Size ( 2) Observed power RTE 60.87 12.27 68.14 11.86 66.40 11.82 7.01* .063 .925 Scholarly Activity 10.45 7.40 18.83 12.84 22.60 15.13 15.92** .133 .999 F (2, 209) = 7.01, p < .001.* ; F (2, 209) = 15.92, p < .001.**
42 CHAPTER 5 DISCUSSION The curren t investigation attempted to advan ce the research training literature through exploration of research ment oring experiences and training model as potentially important factors in training outcomes am ong professional psychology graduate students. Previous work has supported research training environments, re search self-efficacy, research interest, and research outcome expectations as predictors of scholarly activity among counseling psychology students. The current investiga tion included these previously s upported variables, and broadened the investigation through examination of the influence of these variables on clinical and counseling students scholarly activity. In a ddition, the current study also examined key differences in graduate student experiences and scholarly activity as a function of programmatic training model. Results of the current investigation were ge nerally consistent with previous research, although several important departur es were noted. Consistent w ith previous work, research mentoring, research self-efficacy, research outcome expectations, and resear ch interests were all positively related to reports of scholarly activity. These data s uggested greater endorsement of these variables was associated with greater repo rts of scholarly activity for both clinical and counseling students. Interestingly, in the cu rrent study, the relationship between RTE and scholarly activity failed to reach significance. Th is may be explained, in part, by differences in RTE measurement, a decision in the current stud y guided by equivalent re liability and a decision to balance amount of data collected with potential risk of partic ipant fatigue influencing findings. That being said, the 54 item RTES-R utilized in earlier research (K ahn & Scott, 1997) may provide a more sensitive measure of the asso ciation between RTE a nd scholarly activity, particularly given the reported as sociation has generally been a small one. Future research may
43 benefit from measurement of R TE through use of this more extensive measure. Notably, of the variables measured, training model had the strong est association to scholarly activity. This was particularly interesting to the cu rrent investigation, given it has not been previously examined as a predictor of scholarly activity within the litera ture. Also relevant, par ticipants endorsement of practice interests was negatively associated with scholarly activity, suggesting greater endorsement of interest in practice activities wa s associated with lower levels of scholarly activity. These findings appear generally consiste nt with assertions th at the model matters (Neimeyer et al. 2005, 2007; Szymanski et al. 2007). When examining the predictive nature of th e above-mentioned variab les, departures from previous findings were observed. The data suggested only research self-efficacy, training model, practice interests, and research mentoring expe riences, served as signi ficant predictors of scholarly activity, the later fi nding supporting recent assertions regarding the influence of research mentoring in training outcomes (Gelso, 1993, 1997; Hill, 1997; Hollingsworth, 2000; Hollingsworth & Fassinger, 2002). Students level of research interest, and research outcome expectations did not reach significance as predictors of scholarly activity. It may be feasible to suggest that inclusion of research mentoring and training model as ad ditional predictors served to account for variance in scholarly acti vity previously attributed to other variables of interest. The lack of support for these previous ly supported variables may be also be explained by some of the following considerations. As noted by Gelso ( 1993,1997) and Kahn and Scott (1997) academic training programs often do not emphasize training experiences that facilitate positive attitudes toward research and scholarly activity. A possible consequence of this is that students feel less confident in their research abil ities and they are less likely to perceive positive outcomes of research involvement (Kahn, 2001). Another potentia l explanation comes from Neimeyer et als.
44 (2005, 2007) suggestion that traini ng models offer a conceptual framework, which serves to guide programmatic training focus, expectations, and outcomes. The inclusion of training model emphasis, with its overarching framework, may acc ount for the instructional and interpersonal elements associated with RTEs. The moderately strong association betw een training model and elements of RTE (e.g., interpersonal) found in the current study lends possible support for this conclusion. Alternatively, Hollingsworth and Fa ssinger (2002) have proposed that while the training model commitments define the cultur e of a program, its implementation can be enhanced or hindered by interpersonal elements, such as the nature of mentoring relationship. Thus, the inclusion of training model and research mentoring may have accounted for the variance in scholarly activity previously attri buted to the above-mentioned variables. Further work is clearly needed to clarify the nature of these relationships. In addition to attempts to enhance unde rstanding of the potential influence of key variables on scholarly activity, the current investigation also attempted to clarify whether differences in scholarly activity found in previous examinations of tr aining model literature extend to differences in satisfaction with key as pects of students trai ning experiences, and perceptions of their RTE. Cons istent with the current studys predictions, participants from practice-oriented clinical and counseling progr ams reported significantly less scholarly activity than their science-practice and sc ience-oriented peers. In the cu rrent investigati on, they reported 45% and 54% less scholarly activity on average, than participants from science-practice or science-oriented programs. Alt hough clinically oriented commitmen ts were not assessed in the current investigation, it may be in formative to examine whether less scholarly activity, translates into increases in delivery of clinical services. Given increases in interest in clinical activities was associated with decreases in scholarly activity, further attention to these variables as they
45 influence scholarly activity is warranted. Similarly, from a training standpoint, the examination of ones relative preference for service delivery compared to scientific investigation may serve to inform prospective students a bout their potent ial level of satisfacti on, given their chosen programs training model commitments. Potentially supportive of the proposition above, participan ts from practice-oriented programs were generally more satisfied with th eir programs emphasis on clinical practice than those from science-practice or science-oriented programs. Diff erences in satisfaction did not extend to ratings of research emphasis or overall satisfaction. These findings appear to imply that students from more science-or iented programs are unsatisfied with their programs relative commitment to professional service delivery, alth ough further review revealed, that on average, participants from these groups were moderately satisfied. Future research may benefit from greater specificity regarding st udents ratings of satisfaction through allowing for free response or multiple question assessment formats. With regards to participants perceptions of their RTEs the current study differed from previous research in importan t ways. Szymanski et al. (2007) found early career psychologists perceived of their academic RTE similarly regard less of their programs training model, while the current data suggested participants from practice oriented perc eived their RTEs less positively. Again, it is important to note that pa rticipants from all three groups generally perceived their programs RTE positively. Thus, the practical implications of these findings are not entirely clear. Szymanski et al.s (2007) conclusion that part icipants perceptions of their RTEs are generally inline with thei r expectations may provide a par tial explanation as to the lack of differences noted previously. Building upon this conclusion, the current studys failure to find differences in satisfaction with programmatic re search emphasis, may suggest students are less
46 satisfied with specific programmatic elements For example, students may appreciate the consumer of research philosophy held by their pr ogram, hence they are generally satisfied, yet may be frustrated with a research methods require ment or a mentors greater than average level of research interest, hence less positive percep tions of their RTE. The discrepancy between Szymanski et al.s (2007) finding and the current studys results may also be due, in part, to differences in sample demographics (e.g., earl y career psychologists vs. current students; retrospective evaluation vs. current experience). Additiona l research is needed to support these propositions and clarify the complex nature of these questions. The current investigation was una ble to provide support for the potential mediating role of research mentoring. While the association be tween RTE and scholarly activity has been demonstrated in prior work, the failure to find such an association, an assumption required for mediation, precluded the current investigations meditational analyses. Potential measurement issues (discussed in further detail below) may have played a role in the lack of association found. Interestingly, participants in the current study, had moderately positive perceptions of their RTE, yet these perceptions did not appear to influen ce scholarly activity in a perceivable way. Given the current sample examined participants from practice and science-oriented programs, the lack of association may suggest that for programs that do not emphasis research and scholarly activity, perceptions of RTE may be irrelevant. Future research may benefit from sampling from programs with a clear science commitment. The current study was also unable to evaluate recent proposals regarding the role of students prior research interest serves as predictor of their later scholarly activity (Gelso, 1997; Hollingsworth & Fassinger, 2002). Examination of th e association between pa st attitudes toward research (PATR) and measures of current res earch interests revealed high associations
47 suggestive of risk of multicollinearity. Subsequent ly, past and current research interests were deemed too conceptually similar to include as separate predictors in th e current investigation. Thus, the influence of previous research intere st on scholarly activity remains untested at the time of this study. In addition to discussing notable findings fr om the current investig ation, considering the strengths and limitations of the current study in the contex t of the previous literature and conclusions drawn from this work is also warranted. The current study served to advance research training literature thr ough inclusion of additional variab les such as research outcome expectations, training model emphasis, resear ch mentoring, and broadened examination to include both clinical and c ounseling psychology for potential ly greater generalizability. Generalizability may have been further enhanced by the current investigations sampling procedure, which included multiple programs w ithin each specialization within professional psychology. Although not hypothesized, an additional benefit of the current investigation was attention to possible differences in respondents as a function of disciple. Data suggested that participants did not differ in thei r responses as a function of their choice of clinical or counseling graduate training, thus this fi nding implies training model app ears to influence clinical and counseling students in similar ways. The current i nvestigation also attempte d to address previous confounds (e.g., unclear definitions of research mentor, inadequate sample sizes, reporting science-practice emphasis of participants) to cl arify the nature of th e relationships under investigation. Lastly, the current investigation attempted to maximize its comparative potential with previous or future work through use of ge nerally strong, and frequent ly utilized, measures to allow for greater comparis on of findings across studies.
48 The currents study strengths appear to outwe igh its limitations, yet several limitations serve to qualify the current studys findings a nd their acknowledgement will likely serve to improve future research. Notably, a primary hypot hesis remains untested and thus, the current study was unable to further our understanding of the potential meditational properties of research mentoring. As mentioned above, this may due to differences between the previous and current studies including: variations in the measurement of RTES, and lower reliability of dependent measure in the current study. Relatedly, it is possible the current st udy lacked measurement sensitivity (e.g., RTE-R) needed to support one of the primary assumptions of mediation (i.e. association). It also seems feasible to suggest there may have been variations in sample characteristics between the two studies. As alluded to in the previous limitation, many of the conclusions dr awn from the current study were based on a dependent variable with less than ideal reliability. While it is true that the current investigation revealed reliability esti mates well within those cited within previous literature, the overall low Chronb achs alpha reported for the SAS across the literature implies a need to revisit its psychometic properties a nd reassess of whether the SAS is adequately measuring the construct of interest. An examina tion of item reliabilities suggested the two items requesting students to endorse invo lvement in gathering data or conducting statistical analyses appeared to be driving the low reliability of the SAS. A possible revision to improve SAS reliability and the informative na ture of these items, might include a report format assessing the number of data gathering or da ta analysis projects rather than endorsement of involvement. Given the already brief length of the SAS, it would appear that the item modification would serve as a better alternative than item removal, although Ho llingsworth and Fassinger (2002) reported SAS reliability at .74 with a shorter version of the SAS. Alternatively, the varied nature
49 of scholarly activity suggests that future resear ch may benefit from exam ining the tenability of viewing scholarly activity as defined by multiple domains, to be assessed along separate dimensions. The current study attempted to enhance unders tanding of the influen ce of a variety of factors on the scholarly activity of professional psychology graduate students and differences in these factors as a function of the conceptual framework underlying the program, yet several questions require further investig ation. First, the predictors in the current study only accounted for 25% of the variance in scholarly activity. Th e question remains as to what accounts for the other 75 %? In support of the current investigatio n, it is noteworthy to men tion that the variables studied accounted for nearly 10% more variance in scholarly activity than previous work (e.g., Kahn, 2001), yet the fact remains, additional constructs of interest wait to be investigated to gain a broader understanding of factors influencing scholarly activity. Variables for possible inclusion in future work include: year in program, car eer goals (Kahn & Scott, 1997), and research competence (Kahn, 2001). Although at present, meas urement issues (e.g., lack of adequate measures for particular variable s) and feasibility (e.g., longitudina l analysis) have precluded such work (Kahn, 2001). While not previously hypothesized as a predictor, in the current st udy, data was collected on students graduate standing (i .e. year in graduate program) An exploratory ANOVA revealed a significant difference in scholarly activity by year in graduate program. The data suggested that sixth year students reported signif icantly greater levels of scholarly activity compared to all other years, except fifth year students. Forth and Fift h year graduate students also reported greater scholarly activity than first and second year st udents. This finding provi des support for graduate standing as a potentially relevant variable for consideration in the RTE literature. This variable
50 may warrant further investigation as a potential predictor of the sc holarly activity of clinical and counseling graduate students. Additionally, greater understanding of the factors influencing the research productivity of faculty advisors and mentors, charged with dissemination of research training is also warranted (Kahn & Scott, 1997). This particularly true in light of increasing support for the role of research mentoring in students scholarly activity, and the strong interpersonal nature of these relationships (S chlosser & Gelso, 2001, 2005). Lastly, it seems relevant to acknowledge that while the current i nvestigation primarily focused on the influence of training models on sc holarly activity, traini ng models and their associated science-practice commitments serve to contribute more to gradua te training than is implied by this singular focus on research produ ctivity. Future research may benefit from a broader examination of these infl uences such as how training m odel impacts other ways of being a scientist or a clinician. For example, training model emphasis might be reflected in the nature of an individuals view of the world as conceptu alized by their report of their scientific thinking or scientific skepticism broa dly. Or more specifically, in their understanding of empirical research, application of research to practice, or critical evalua tion of interventions and outcomes as it affects their clients, and th eir choices. Similarly, while scholar ly activity was conceptualized as influenced by the above mentioned variables, it may be the product of many things such as having a productive mentor, being attached to a federal research grant, etc. Thus, it may be important to qualitatively, as well as, quantit atively assess students and relevant others perceptions of these factors in fu ture research. Also, within any training environment, whether it be clinically or scientifically oriented, the be tter the student-program match and student-mentor relationship fit, the greater li kelihood of positively perceived outcomes. Future research may
51 benefit from an examination of the role of fit or match as it relates to relevant training outcomes (e.g., satisfaction, productivity, retention). In conclusion, research targeted at increas ing attention to factors of graduate training contributing to students scholarly activity is continually needed (Gelso, & Lent 2000). Applying research findings in the form of subsequent in terventions focused on cultivating greater scholarly activity among programs and students with interest in these pursuits may he lp to support the goal of advancement professional psychology (B elar, 2000; Gelso, 1979, 1993; Gelso & Lent, 2000). The current investigation attempted to provide a small contribution to understanding these factors and relevant differences amon g professional psychology students.
52 APPENDIX A RESEARCH TRAINING ENVIRONMENT SCALE-REVISED SHORT FORM (RTES-R) Below is a s eries of statements concerning research training. Please note that we define research broadly. "Res earch" when used in this survey includes the following types of activities: designing and executing research projects, preparing manuscripts of a theoretical nature or a cri tical review of literatu re, conducting program evaluations or needs assessments, making presentations at professional conferences, participating as a member of a research team engaged in any of the above activitie s, and advising the resear ch projects of others. Please respond to the following statements in terms of the doctoral program in which you are currently receiving your training. (Note: If you are currently on internship, please rate the graduate program in which you were previously tr ained.) It is important to answer each item, even if some of the items are difficult to answ er. Consider each statem ent using the following scale: 1 2 3 4 5 disagree somewhat neutral somewhat agree disagree agree 1. Many of our faculty do not seem to be very interested in doing research. 2. The faculty does what it can to make research requirements such as the thesis and dissertation as rewarding as possible. 3. My advisor understands and accepts that any piece of research will have its methodological problems. 4. I have felt encouraged during my training to find and follow my own scholarly interests. 5. Statistics courses here are taught in a way that is insensitive to students' level of development as researchers. 6. The statistics courses we take do a good job, in general, of show ing students how statistics are actually used in psychological research. 7. There is a sense around here that being on a research team can be fun, as well as intellectually stimulating. 8. Faculty members in my program use an extr emely narrow range of research methodologies. 9. Generally, students in my trai ning program do not seem to have intellectually stimulating and interpersonally rewarding relationships with their research advisors. 10. It is unusual for first-year students in this program to collaborate with advanced students or faculty on research projects.
53 11. I have the feeling, based on my training, that my thesis (or dissert ation) needs to be completely original and revolutionary for it to be acceptable to the faculty. 12. Our faculty seems interested in understandi ng and teaching how research can be related to counseling practice. 13. Most faculty do not seem to really care if st udents are genuinely in terested in research. 14. During our coursework, graduate students are taught a wide ra nge of research methodologies, e.g., field, laboratory, survey approaches. 15. Students in our program feel that their pers onal research ideas ar e squashed during the process of collaborating with facu lty members, so that the fini shed project no longer resembles the student's original idea. 16. Students here seem to get involved in thinki ng about research from the moment they enter the program. 17. Students in this program are ra rely taught to use research findi ngs to inform their work with clients. 18. The faculty members of my gr aduate program show excitement about research and scholarly activities. Reprinted with permission from Kahn, J. H ., & Miller, S. A. (2000). Measuring global perceptions of the research training envir onment using a short form of the RTES-R. Measurement and Evaluation in Counseling and Development, 33, 103.
54 APPENDIX B SELF-EFFICACY IN RESEARCH MEASURE (SERM) Instru ctions: The following items are tasks relate d to research. Please indicate your degree of confidence in your ability to succes sfully accomplish each of the following tasks on a scale of 0 9 with 0 representing no confidence a nd 9 representing to tal confidence. 0 1 2 3 4 5 6 7 8 9 no total confidence confidence 1. Keeping records during a research project. 2. Designing an experiment using traditional methods (e.g., experimental, quasi-experimental designs) 3. Writing the introduction and litera ture review for a dissertation 4. Writing the introduction and discussion sectio ns for a research paper for publication 5. Formulating hypotheses 6. Writing the method and results sections of a thesis 7. Utilizing resources for needed help 8. Understanding computer printouts 9. Defending a thesis or dissertation 10. Using multivariate statistics (e.g., multip le regression, factor analysis, etc.) 11. Using statistical packages (e.g., SPSS-X, SAS, etc.) 12. Operationalizing variables of interest Reprinted with permission from Kahn, J. H., & Scott, N. A. (1997). Predictors of research productivity and science-related career goa ls among counseling psychology doctoral students. The Counseling Psychologist, 25, 38.
55 APPENDIX C THE RESEARCH OUTCOME EXPECTAT IONS QUES TIONNAIRE (ROEQ) Directions: Using the 5-point scale provided, pl ease indicate the degree to which you agree with each statement. 1 2 3 4 5 Strongly Disagree Agree Strongly Agree 1. Involvement in research will e nhance my job/career opportunities. 2. People I respect will approve of my involvement in research. 3. Involvement in research will allow me to contribute to practitioners knowledge base. 4. Doing research will increase my sense of self-worth. 5. Becoming involved in a research project will lead to the kind of career I most want. 6. Research involvement is valued by significant people in my life. 7. My peers will think high ly of me if I become involved in research. 8. Pursuing research involvement will enable me to associate with the kind of people I value most. 9. Involvement on a research team can lead to close personal connections. 10. Research involvement will lead to a sense of satisfaction. 11. Being involved in research will contribute to my development as a professional. 12. I believe research skills will be fruitful for my career. 13. My involvement in research will lead to meaningful contribu tions to the field. 14. If I get involved in research it will take tim e away from my signif icant relationships.* 15. Involvement in research will take time from leisure activities.* 16. Involvement in research will help me to unde rstand the current issues in my profession. 17. My analytical skills will beco me more developed if I am invo lved in research activities. 18. I believe that research involvement will l ead to becoming well-known and respected in the field. 19. Research involvement will lead to increased financial opportunities. 20. Involvement in research will posi tively influence my applied skills. Reprinted with permission from Bishop, R. M ., & Bieschke, K. J. (1998). Applying social cognitive theory to interest in research among counseli ng psychology doctoral students: A path analysis. Journal of Counse ling Psychology, 45, 182.
56 APPENDIX D SCIENCTIST-PRACTITIONER INVENTORY The following questions ask about interest in activities often performed by psychologists. Indicate how interested you are in each of the following activities. Please provide your answers on the blank line to the right of each item. The response categories are as follows. ____________________________________________________________________________ very low low medium high very high interest interest interest interest interest 1 2 3 4 5 ____________________________________________________________________________ Your response 1. Designing a new treatment method fo r a mental health agency. ________ 2. Writing a scientific book for psychologists. ________ 3. Conducting couples and family therapy ________ 4. Collecting data on a rese arch project you designed ________ 5. Reviewing journal articles ________ 6. Presenting a report during a case conference. ________ 7. Applying for research grants ________ 8. Interpreting a test battery for a client ________ 9. Serving as an editor for a scientific journal ________ 10. Learning new strategies to d eal with psychological problems ________ 11. Reading a book on innovative research designs ________ 12. Receiving therapy to make yourself a better person ________ 13. Learning about new statistical procedures ________ 14. Reviewing an agency's intake form for a new client ________ 15. Developing new explanations of well accepted empirical studies ________ 16. Reading a book written by a famous psychotherapist ________ 17. Serving on a thesis or dissertation committee ________ 18. Planning a behavior modifica tion program for a client ________ Reprinted with permission from Leong, F.T.L. & Zachar, P. (1993). Presenting two brief versions of the Scientist Practitioner Inventory. Journal of Career Assessment, 1 162-170.
57 APPENDIX E RESEARCH MENTORING EXPE RIENCES SCALE (RMES) Faculty often play an im portant role in student s' research training and research experiences. Some students receive their most significant research experien ces with their formally assigned advisor, while others receive their most impor tant research mentoring through more informal faculty relationships. If you do not have anyone that you consider as a faculty mentor, please consider the faculty relationship th at has been most important in your research training while in your current doctoral program, and use the followi ng items to describe your current perceptions of this relationship. It is important that you consider your relationship with only one faculty member in completing this survey. Not all of these behaviors are important to all students or faculty, so please indicate "N/A" for those behavi ors that are not presen t in your relationship. You will need to provide a response to the stem in each column, circling the appropriate number in each column. Research Task Functions IN YOUR RESEARCH RELATIONSHIP WITH A SPECIFIC FACULTY MEMBER, TO WHAT EXTENT DOES HE OR SHE PAY ATTENTION TO THE FOLLOWING: A Great Very Not Deal Some Little Applicable 1. discussing your research-related goals? 5 4 3 2 1 N/A 2. helping you develop research ideas? 5 4 3 2 1 N/A 3. involving you in one or more specific research projects? 5 4 3 2 1 N/A 4. exposing you to different research methods? 5 4 3 2 1 N/A 5. reminding you that flaws in research projects are inevitable? 5 4 3 2 1 N/A 6. suggesting additional resources, such as people or literature, you can consult to improve your research? 5 4 3 2 1 N/A 7. helping you organize a review of the literature? 5 4 3 2 1 N/A 8. helping you to identify weaknesses in a research project? 5 4 3 2 1 N/A 9. helping you develop a realistic timetable for research projects? 5 4 3 2 1 N/A 10. encouraging you to apply for researchrelated grants? 5 4 3 2 1 N/A
58 Research Task Functions IN YOUR RESEARCH RELATIONSHIP WITH A SPECIFIC FACULTY MEMBER, TO WHAT EXTENT DOES HE OR SHE PAY ATTENTION TO THE FOLLOWING: A Great Very Not Deal Some Little Applicable 11. encouraging you to attend important professional conferences? 5 4 3 2 1 N/A 12. introducing you to her/his professional colleagues who have similar research interests? 5 4 3 2 1 N/A 13. encouraging you with presentations of research at professional conferences? 5 4 3 2 1 N/A 14. collaborating with you on joint research projects? 5 4 3 2 1 N/A 15. encouraging you to express your ideas in research meetings? 5 4 3 2 1 N/A 16. using his/her power to motivate you to complete research tasks? 5 4 3 2 1 N/A 17. offering positive feedback about your research work? 5 4 3 2 1 N/A 18. constructively critic izing your research work? 5 4 3 2 1 N/A 19. encouraging you to talk openly about anxieties or fears that interfere with research? 5 4 3 2 1 N/A 20. providing advice about how to manage feelings of frustration with research? 5 4 3 2 1 N/A 21. communicating interest in your ideas when you talk about research? 5 4 3 2 1 N/A 22. communicating respect regarding cultural differences in your relationship? 5 4 3 2 1 N/A 23. expressing appreciation for your contributions to research? 5 4 3 2 1 N/A 24. modeling competence in researchrelated skills? 5 4 3 2 1 N/A 25. observing connections between research and practice? 5 4 3 2 1 N/A
59 Research Task Functions IN YOUR RESEARCH RELATIONSHIP WITH A SPECIFIC FACULTY MEMBER, TO WHAT EXTENT DOES HE OR SHE PAY ATTENTION TO THE FOLLOWING: A Great Very Not Deal Some Little Applicable 26. describing research as rewarding? 5 4 3 2 1 N/A 27. discussing his/her research dilemmas with you? 5 4 3 2 1 N/A 28. expressing enthusiasm for research? 5 4 3 2 1 N/A Reprinted with permission from Hollingsworth, M. A., Fassinger, R. A. (2002). The role of faculty mentors in research training of counseling psychology doctoral students. Journal of Counseling Psychology, 49(3), 324-330.
60 APPENDIX F PAST ATTITUDES TO WARDS RESEARCH Please rate y our agreement with each of the following statements on a 5-point Likert scale, ranging from 1 (strongly disagr ee) to 5 (strongly agree). 1. I would have preferred to have the option of completing my doctoral training without being required to complete research projects. 1 2 3 4 5 2. I had a strong interest in doing research. 1 2 3 4 5 3. I placed a high value on the place of research in my future career. 1 2 3 4 5 4. Participating in research activities after graduation was not a ma jor priority for me. 1 2 3 4 5 Reprinted with permission from Royalty, G. M., Gelso, C. J., Mallingckrodt, B., & Garrett, K. D. (1986). The environment and the student in counseling psychology: Does the research training environment influence graduate students' attitudes toward research? The Counseling Psychologist, 14, 9.
61 APPENDIX G SCHOLARLY ACTIVITY SCALE (SAS) Instru ctions : The following items assess research accomp lishments and current involvement in research activities. Please answer the follo wing questions based on your past and current research involvement. _____ 1. How many published manuscripts (either empirical or otherwise) have you authored or coauthored in a refereed j ournal? (include manuscripts in press) _____ 2. How many unpublished empirical ma nuscripts have you auth ored or coauthored (not including your thesis or dissertation)? _____ 3. How many articles have you submitted to refereed journals? _____ 4. How many manuscripts are you curren tly in the process of preparing to submit for publication (i.e., writing the manuscript)? _____ 5. How many presentations have you made at local, regional, or national conventions? _____ 6. How many presentations are you current ly in the process of preparing to submit for presentation (i.e., writing an abstract)? _____ 7. How many local, regional, or natio nal research conventions have you attended? Y N 8. Are you currently involved in gathering data (do not include your thesis or dissertation)? Y N 9. Are you currently conducting stat istical analyses on data ( do not include your thesis or dissertation)? Reprinted with permission from Kahn, J. H., & Scott, N. A. (1997). Predictors of research productivity and science-related career goals among counseling psychology doctoral students. The Counseling Psychologist, 25, 38.
62 LIST OF REFERENCES Academ y of Psychological Clinical Science. (2004). APCS homepage. Retrieved July, 2008 from http://psych.arizona.edu/apcs/purpose.html. American Psychological Association & Office of Program Consultation and Accreditation (1996). 1995 summary of characteristics and outcomes of accredited counseling psychology doctoral programs Washington, DC: Author. Atkinson, D. R., Neville, H., & Casas, A. (1991). The mentorship of ethnic minorities in professional psychology. Professional Psychology: Research and Practice, 22, 336-338. Baron, R. M., & Kenny, D. A. (1986). The moderato r-mediator variable di stinction in social psychological research: Con ceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51 1173-1182. Belar, C. D. (2000). Scientistpractitioner science + practice: Boulder is bolder. American Psychologist, 55, 249-250. Belar, C. D. & Perry, N. W. (1992). The nationa l conference on scientistpractitioner education and training for the professi onal practice of psychology. American Psychologist, 47(1), 71-75. Bentler, P. M., & Chou, C. P. (1987). Pr actical issues in structural modeling. Sociological Methods and Research, 16, 78. Betz, N. E. (1997). Increasing research involveme nt and interests among graduate students in counseling psychology. The Counseling Psychologist, 25, 88-93. Bieschke, K. J., Bishop, R. M., & Herbert, J. T. (1995). Research interest among rehabilitation doctoral students. Rehabilitation Education, 9, 51-66. Bieschke, K. J., Herbert, J. T., & Bard, C. ( 1998). Using a social cognitive model to explain research productivity among rehabil itation counselor education faculty. Rehabilitation Education, 12, 1-16. Bishop, R. M., & Bieschke, K. J. (1998). Applying so cial cognitive theory to interest in research among counseling psychology doctora l students: A path analysis. Journal of Counseling Psychology, 45, 182. Brems, C., Johnson, M. E., & Gallucci, P. (1996) Publication productivity of clinical and counseling psychologists. Journal of Clinical Psychology, 52, 723. Cherry, D. K., Messenger, L. C., & Jacoby, A. M. (2000). An examination of training model outcomes in clinical psychology programs. Professional Psychology: Research and Practice, 31, 562.
63 Clark, R. A., Harden, S. L., & Johnson, W. B. (2000). Mentoring relationships in clinical psychology doctoral training: Results of a national survey. Teaching of Psychology, 27(4), 262-268. Cronan-Hillix, T., Gensheimer, L. K., Crona n-Hillix, W. A., & Davidson, W. S. (1986). Students' view of mentors in psychology graduate training. Teaching of Psychology, 13(3), 123-127. Elman, N. S., Illfelder-Kaye, J., & Robiner, W. N. (2005). Professional development: Training for professionalism as a foundation for competent practice in psychology. Professional Psychology: Research and Practice. 36(4), 367-375. Fitzgerald, L. F., & Osipow, S. H. (1988). We ha ve seen the future, but is it us? The vocational aspirations of graduate st udents in counseling psychology. Professional Psychology: Research and Practice, 19(6), 575-583. Gaddy, C. D., Charlot-Swilley, D., Nelson, P. D., & Reich, J. N. (1995). Selected outcomes of accredited programs. Professional Psychology: Re search and Practice, 26, 507. Galassi, J. P., Brooks, L., Stoltz, R. F., & Trexle r, K. A. (1986). Research training environments and student productivity : An exploratory study. The Counseling Psychologist, 14, 31-36. Gelso, C. J. (1979). Research in counselin g: Methodological and pr ofessional issues. The Counseling psychologist, 8, 7-35. Gelso, C. J. (1993). On the making of a scientistpractitioner: A theory of research training in professional psychology. Professional Psychology: Research and Practice, 24, 468. Gelso, C. J., & Lent, R. W. (1997). The making of a scientist in applied psychology: An attribute by treatment conception. The Counseling Psychologist, 25, 307. Gelso, C. J., & Lent, R. W. (2000). Scientific training and scholarly pr oductivity: The person, the training environment, and their interacti on. In S. D Brown & R. W. Lent (Eds.), Handbook of Counseling Psychology (3rd ed., pp. 109-139). New York: Wiley. Gelso, C. J., Mallinckrodt, B., & Judge, A. B. (1996). Research training environment, attitudes toward research, and research self-efficacy: The revised Research Training Environment Scale. The Counseling Psychologist, 24, 304. Hill, C. E. (1997). The effects of my research training enviro nment: Where are my students now? The Counseling Psychologist, 25, 74-81. Hollingsworth, M. A. (2000, August). Counseling ps ychology students' perceptions of research mentoring relationships. In J. H. Kahn (C hair), Research training in counseling psychology: New advances and directions. Symposium conducted at the 108th Annual Convention of the American Psychol ogical Association, Washington, DC.
64 Hollingsworth, M. A., Fassinger, R. A. (2002). The role of faculty mentor s in research training of counseling psycholog y doctoral students. Journal of Counseling Psychology, 49(3), 324-330. Hoshmand, L. T. & Polkinghorne, D. E. (1992). Redefining the science-practice relationship and professional training. American Psychologist, 47, 55-66. Huwe, J. M., & Johnson, B. W. (2003). On being an excellent protg: What graduate students need to know. Journal of College Student Psychotherapy, 17(3), 41-57. Kahn, J. H. (2001). Predicting the scholarly activity of counseling psychology students: A refinement and extension. Journal of Counseling Psychology, 48, 344. Kahn, J. H., & Miller, S. A. (2000). Measuring global perceptions of the research training environment using a short form of the RTES-R. Measurement and Evaluation in Counseling and Development, 33 103. Kahn, J. H., & Scott, N. A. (1997). Predictors of research productivity and science-related career goals among counseling psychology doctoral students. The Counseling Psychologist, 25, 38. Krebs, P. J., Smither, J. W., & Hurley, R. B. (1991). Relationship of vocational personality and research training environment to research productivity of counseling psychologists. Professional Psychology: Re search and Practice, 22, 362. Jones, J., & Mehr, S. (2007). Foundations and as sumptions of the scient ist-practitioner model. American Behavioral Scientist, 50(6), 766-771. Lark, J. S., & Croteau, J. M. (1998). Lesbian, ga y, and bisexual doctora l students' mentoring relationships with faculty in counseling psychology: A qualitative study. The Counseling Psychologist, 20, 754-777. Landino, R. A., & Owen, S. V. (1988). Self-efficacy in university faculty. Journal of Vocational Behavior 33(1) 1-14. Leong, F.T.L. & Zachar, P. (1993). Presenting two brief versions of the Scientist Practitioner Inventory. Journal of Career Assessment, 1, 162-170. Luna, G., & Cullen, D. (1998). Do graduate students need mentors? College Student Journal, 32, 322-331. Mallinckrodt, B. (1997). Discoveri ng research training environmen ts that fit the students we select: The challenge of providing for diverse needs. The Counseling Psychologist, 25, 68-73. Mallinckrodt, B., Gelso, C. J., & Royalty, G. M. (1990). Impact of the research training environment and counseling psychology students' Holland personality type on interest in research. Professional Psychology: Re search and Practice, 21, 26-32.
65 Mallinckrodt, B., & Gelso, C. J. (2002). Impact of research training environment and Holland personality type: A fifteen year follow-up of research productivity. Journal of Counseling Psychology, 49, 60-70. McHolland, J. D. (1992). National Council of Schools of Professional Psychology core curriculum conference resolutions. In R. L. Pe terson et al. (Eds.), The core curriculum in professional psychology (pp. 155). Wash ington, DC: American Psychological Association. Murdock, N. L., Alcorn, J., Heesacker, M., & St oltenberg, C. (1998). Model training program in counseling psychology. The Counseling Psychologist, 26, 658-672. Neimeyer, G. J., Saferstein, J., & Rice, K. G. (2005). Does the model matter? The relationship between science-practice emphasis and outcomes in academic training programs in counseling psychology. The Counseling Psychologist, 33, 635. Neimeyer, G. J., Rice, K. G., & Keilin, W. G. (2007). Does the Model Matter? The Relationship Between Science-Practice Emphasis in Clinic al Psychology Programs and the Internship Match. Training and Education in Professional Psychology, 1(3), 153 Neimeyer, G. J. & Taylor, J. M. (in press). Counselling Psychology Quarterly. Noe, R. A. (1988). An investigation of the determinants of successful assigned mentoring relationships. Personnel Psychology, 41, 457-479. Norcross, J., Sayette, M., & Mayne, T. (2008). Insider's guide to graduate programs in clinical and counseling psychology, 2008/2009 ed New York, NY, US: Guilford Press. O'Brien, K. M. (1995). Enhancing research training for counseling students: Interuniversity collaborative research teams. Counselor Education and Supervision, 34, 187-198. O'Sullivan, J., & Quevillon, R. (1992). 40 years later: Is the Boulder model still alive? American Psychologist, 47(1), 67-70. Osborne, Jason W. & Amy Overbay (2004). The pow er of outliers (and why researchers should always check for them). Practical Assessmen t, Research & Evaluation, 9(6). Retrieved June 9, 2009 from http://PAREonline.net/getvn.asp?v=9&n=6 Phillips, J. C., & Russell, R. K. (1994). Re search self-efficacy, the research training environment, and research productivity among graduate students in counseling psychology. The Counseling Psychologist, 22, 628. Ragins, B. R. & McFarlin, D.G. (1990). Perceptio ns of mentor roles in cross-gender mentoring relationships. Journal of Vocational Behavior, 37, 321-339. Royalty, G. M., Gelso, C. J., Mallingckrodt, B., & Garrett, K. D. (1986). The environment and the student in counseling psychology: Does th e research training environment influence graduate students' attitudes toward research? The Counseling Psychologist, 14, 9.
66 Royalty, G. M., & Magoon, T. M. (1985). Correla tes of scholarly produ ctivity among counseling psychologists. Journal of Counseling Psychology, 32, 458-461. Royalty, G. M., & Reising, G. N. (1986). The re search training of couns eling psychologists: What the professionals say. The Counseling Psychologist, 14, 49-60. Schlosser, L. Z., & Gelso, C. J. (2001). Meas uring the working alliance in advisor-advisee relationships in graduate school. Journal of Counseling Psychology, 48(2), 157-167. Schlosser, L. Z. & Gelso, C. J. (2005). The Advisory Working Allianc e Inventory Advisor Version: Scale Development and Validation. Journal of Counseling Psychology, 52 650654. Smith, E. P., & Davidson, W. S.(1992). Mentor ing and the development of African-American graduate students. Journal of College Student Development 33(6), 531-539. Stoltenberg, C. D., Pace, T. M., Kashubeck-West, S., Beiver, J. L., Patterson, T., & Welch, I. D. (2000). Training models in counseling psyc hology: Scientist practitioner versus practitioner-scholar. The Counseling Psychologist, 28 622-640. Vespia, K (2006). Integrating professional id entities: Counselling psychologist, sciencepractitioner, and undergraduate educator. Counselling Psychology Quarterly, 19(3) 265280. Szymanski, D.M., Jovanovic Ozegovic, J., Phillips, J.C., & Briggs-Phillips, M. (2007). Fostering scholarly productivity through academic and internship research training environments. Training and Education in Professional Psychology, 1(2), 135-146. Watkins, C., Lopez, F., Campbell, V., & Himmell, C. (1986). Contemporary counseling psychology: Results of a national survey. Journal of Counseling Psychology, 33(3), 301309.
67 BIOGRAPHICAL SKETCH Geoffrey A. Lee was born in San Antonio, Texas, in 1981. In 1994 his fam ily moved to Key West, Florida, where he reside d until he was eighteen years old. He attended the University of Florida in 1999 majoring in psychology as an undergraduate. In 2003, he graduated earning highest honors wi th a Bachelor of Science in psychology. He joined the Department of Psychology at the University of Florida as a counseling psychology graduate student in August of 2004. He completed his Master of Science degree in June of 2005 and his Doctor of Philosophy in August of 2009.