Self-Efficacy Regarding Social Work Competencies

Abstract

Purpose:

The need for psychometrically sound measurement approaches to social work educational outcomes assessment is increasing.

Method:

The research reported here describes an original and two replication studies of a new scale (N = 550) designed to assess an individual’s self-efficacy regarding social work competencies specified by the Council on Social Work Education as part of the accreditation of social work programs.

Results:

This new measure, the Self-Efficacy Regarding Social Work Competencies Scale (SERSWCS), generally performed in line with our expectations.

Discussion:

The SERSWCS is a measure that is based on substantial theoretical and empirical work, has preliminary evidence regarding the psychometric properties of the data it produces, can be used with large numbers of students in an efficient manner, is neither expensive or subject to user restrictions, and provides views of outcomes that have utility for pedagogical considerations at multiple curricular levels.

Keywords

self-efficacy scale competency assessment higher education educational program accreditation replicate measurement learning outcome reliability validity

Demands for greater accountability at all levels of education echo across the landscape. They follow nearly a century of calls for reform in primary and secondary education (Ravitch, 2000). The “foundation of the accreditation or professional judgment approach to evaluation can be traced directly to the establishment of the North Central Association of Colleges and Secondary Schools in the late 1800s. The accreditation movement did not, however, gain great stature until the 1930s” (Madaus & Stufflebeam, 2000, p. 6). Currently, accreditation by the Council on Social Work Education (CSWE) is a dominant feature in higher education in social work in the United States. The purpose of this article is to explore an approach that could be used to produce social work higher education outcome assessments related to current accreditation-related accountability demands.

Competencies

An advocate of competency-based education noted that various definitions of competencies exist but they are essentially “the result of integrative learning experiences in which skills, abilities, and knowledge interact to form learning bundles that have currency in relation to the task for which they are assembled” (Voorhees, 2001, p. 9). In this definition, performance is at the top of a hierarchy that is the result of the application of competencies. Underlying competencies are various skills, abilities, and knowledge (e.g., Jones, Voorhees, & Paulson, 2002).

The CSWE has adopted elements of competency-based education in framing social work educational outcomes as competencies. CSWE states in their current Educational Policy and Accreditation Standards:
Competency-based education is an outcome performance approach to curriculum design. Competencies are measurable practice behaviors that are comprised of knowledge, values, and skills. The goal of the outcome approach is to demonstrate the integration and application of the competencies in practice with individuals, families, groups, organizations, and communities. The ten core competencies are listed below … followed by a description of characteristic knowledge, values, skills, and the resulting practice behaviors. (CSWE, 2008, p. 3, emphasis added)
It is these practice behaviors that are the focus of this article. As others have indicated (e.g., Bogo, 2010), despite framing aspects of social work in competency terms, the field has yet to develop psychometrically sound measures of those competencies.

What Is Needed?

How, in an educational outcomes sense, can the performances of groups of faculty (e.g., a curricular area) or of an entire school of social work be fairly assessed over a period of time in the manner that social work accreditation requires? A long-term goal of our research team has been to develop measurement approaches that allow programs to utilize an effective method of gathering data relevant to accreditation requirements. We view these efforts as worthwhile because few measures of social work educational outcomes exist that:
are based on prior theoretical and empirical work,

have evidence regarding the psychometric properties of the data they produce,

are able to be used with large numbers of students in an efficient manner,

are not expensive or subject to user restrictions, and

are able to provide views of outcomes that have utility for pedagogical considerations at many levels (e.g., school, curricular area, course, and course elements).

In addition, CSWE Accreditation Standard 4.0.1 (2013) states that each program needs an assessment plan that uses multiple measures to assess competencies. We concur with this perspective and have always noted that we view our measures as one component of an overall multicomponent assessment of social work educational outcomes (cf. DeLong-Hamilton et al., 2011).

Self-Efficacy

The scales we have developed are based on Bandura’s construct of self-efficacy, which is derived from his Social Cognitive Theory (1977, 1982, 1986, 1997, 2004). Although sometimes considered at the collective level, self-efficacy is usually defined as an individual’s confidence that they will be able to successfully perform a particular sequence of behaviors in the future. Note that the individual’s actual competency is not being measured, but rather their belief in their ability to perform. A selective review reveals that self-efficacy is a truly versatile construct. Not only has it been employed across many other fields, it has been used in a variety of situations in social work education and practice. These include the following:
discussions of the social work educational process in general and outcomes assessment specifically (e.g., Calderon, 2013; Carpenter, 2011; Green, 2003; Montcalm, 1999; Petrovich, 2004; Spitzer et al., 2001);

as a predictor of: intentions to remain employed in child welfare (Ellett, 2009), research activity of social workers (Lynch, Zhang, & Korr, 2009), and for domestic violence screening by social workers (Tower, 2003);

as an outcome in a conceptual model of practice with battered women (Danis, 2004), an HIV risk reduction intervention study (Icard, Schilling, & El Bassel, 1995), a comparison of services for battered women (Mancoske, Standifer, & Cauley, 1994), and a study of a wilderness adventure therapy intervention (Clem, Smith, & Richards, 2012);

as a factor in a model that portrayed bachelor of social work (BSW) and master of social work (MSW) research instructors perceptions of their students (Maschi, Wells, Slater, MacMillan, & Ristow, 2013); and one that examined BSW field instructor attitudes toward evidence-based practice (Wiechelt & Ting, 2012);

in the development of the Substance Abuse Treatment Self-Efficacy Scale (Kranz, 2003; Kranz & O’Hare, 2006, 2011), and as a component of two versions of the Evidence-Based Practice Process Assessment Scale (Parrish & Rubin, 2011; Rubin & Parrish, 2011);

as an outcome measure for an intervention by BSW students with foster youths (Bruster & Coccoma, 2013), training in parental mental illness/child protection for social workers (Carpenter, Patsios, Szilassy, & Hackett, 2011), a developmental program for newly qualified child and family social workers (Newly Qualified Social Workers; Carpenter, Shardlow, Patsios, & Wood, 3), an intervention designed to enhance analytic abilities related to the assessment process for children and families (Platt, 2011), and an intervention designed to disseminate an evidence-supported intervention to community practitioners (Woody, Anderson, & D’Souza, 2015);

as an outcome measure for a meta-analysis of the impact of youth empowerment programs (Morton & Montgomery, 2013), an evaluation of a training course for child protection workers (Maxwell, Scourfield, Holland, Featherstone, & Lee, 2012; Scourfield et al., 2012), and the evaluation of an intervention focused on child safeguarding (Scourfield, Smail, & Butler, 2014);

as educational outcome measures in studies of field instruction (e.g., Cuzzi, Holden, Chernack, Rutter, & Rosenberg, 1997; Cuzzi, Holden, Rutter, Rosenberg, & Chernack, 1996; Holden, Cuzzi, Rutter, Chernack, & Rosenberg, 1997; Holden, Cuzzi, Rutter, Rosenberg, & Chernack, 1996; Holden, Cuzzi, Rutter, Chernack, Spitzer et al., 1997; Parker, 2006), interactive multimedia training for undergraduate social work students (Cauble & Thurston, 2000), an undergraduate writing course (Woody et al., 2014), research/evaluation courses at undergraduate and graduate levels (Holden, Barker, Meenaghan, & Rosenberg, 1999; Holden, Barker, Rosenberg, & Onghena, 2007, 2008; Unrau & Beck, 2004; Unrau & Grinnell, 2005), a comparison of web-based versus face-to-face MSW classes (Lawrence & Abel, 2013), “shadowing” (Parker, Hughes, & Rutter, 2006–2007), a service learning course (Williams, King, & Koob, 2002), a trauma informed evidence-based practice graduate course (Strand, Abramovitz, Layne, Robinson, & Way, 2014), a problem-based learning project (Westhues, Barsen, Freyond, & Train, 2014), and a social work administration course (York, 2008);

as an educational outcome measure for the foundation level of an MSW program (Holden, Anastas, & Meenaghan, 2003, 2005; c.f. Ahn, Boykin, Hebert, & Kulkin, 2012, for an alternative application of this measure);

for assessment of student change in the advanced concentration (Rishel & Majewski, 2009); and

as an educational outcome measure for a complete MSW program (Holden, Meenaghan, Anastas, & Metrey, 2002; Meyer-Adams, Dorsey, Potts, Resales, & Koob, 2011; Rawlings, Townsend, & Gingerich, 2005).

Clearly, the construct has been appealing to a variety of researchers covering a range of social work topics for over 20 years. That said, not all studies involving self-efficacy have produced results that appear consistent with Social Cognitive Theory (the theory within which self-efficacy is embedded). We address this issue next.

Measurement issues and definitional concerns

In some instances, social work researchers have not found theoretically predicted relationships between self-efficacy and educational outcomes (e.g., Fortune, Lee, & Cavazos, 2005; Rawlings, 2012). On the other hand, we have contradictory evidence supporting the ability of self-efficacy ratings to predict future competence in teachers (van Dinther, Dochy, Segers, & Baeken, 2014). From our reading of this literature over the years, a number of factors might be contributing to contradictory results.

First, some of the studies in social work using self-efficacy have employed outcome measures of unknown reliability and validity (e.g., field instructors’ ratings of student performance, ratings of a brief videotaped simulation with a “standardized” client). While superficially appealing as measures, no rigorous evidence regarding psychometrics is typically presented.

Second, is the definition and operationalization of self-efficacy in research studies consistent with the theory? Some authors (including some cited earlier) appear to misunderstand the construct of self-efficacy. For instance, self-efficacy is not exclusively a “belief in one’s own ability to learn” (Drisko, 2014, p. 421). Our definition, based on Bandura’s theory, is that “self-efficacy is usually defined as an individual’s confidence that they will be able to successfully perform a particular sequence of behaviors in the future.” While an individual may (A) have varying levels of belief in their ability to learn, they are also likely to have (B) varying levels of belief in their ability to carry out a near endless array of other performances. A does not subsume B as Drisko is interpreted to imply. Self-efficacy is not restricted to a belief in one’s ability to learn. Furthermore, and in contrast to what Drisko implied, we did not argue in our 2002 article that self-efficacy measures indicated in situ competence. We stated that “[s]elf-efficacy is more than a self-perception of competency. It is an individual’s assessment of his or her confidence in their ability to execute specific skills in a particular set of circumstances and thereby achieve a successful outcome” (Holden et al., 2002, p. 116). The more explicit name of the Self-Efficacy Regarding Social Work Competencies Scale (SERSWCS) is in part an attempt to rectify misunderstandings of the construct that we have encountered in the past. Our scales are designed to measure self-efficacy not competency. Misconstruing self-efficacy or claiming it is synonymous with competence does not move the field forward. In addition, although Bandura (most recently in 2006) has offered guidance on the construction of self-efficacy scales, this advice regarding operationalization does not always appear to have been followed when choosing or creating measures of self-efficacy in social work.

Third, in our own studies of self-efficacy, we have always used totally anonymous survey administrations in an effort to obtain more honest responses from students concerning their confidence (cf. Krumpal, 2013). There are a variety of social desirability pressures that may introduce measurement error when self-efficacy or other outcome variables are measured nonanonymously in situations where students may be sensitive to how they are being perceived (e.g., self-reports by students concerning their skills, ratings of students by field instructors, etc.).

Predictive validity

An important consideration of self-efficacy as a social work educational outcome measure is the issue of predictive validity. Bandura has noted,
Nine large-scale meta-analyses have been conducted on findings from studies with diverse experimental and analytic methodologies applied across diverse spheres of functioning in diverse milieus with diverse populations … The evidence from these varied lines of research verify the explanatory and predictive generality of the theory. (Bandura, 2004, p. 622)
Direct demonstration of the predictive validity of social work self-efficacy measures has not been achieved, in large part, due to the lack of reliable and valid measures of social work practice in general. In prior manuscripts, we made our case logically. If self-efficacy ratings are predictive of future performance across many other domains, it seems logical that the same construct would operate within social work education in a similar manner.

Research Questions and Hypotheses

Given the prior work of others and ourselves, we believe that the construct of self-efficacy continues to have promise as an educational outcome measure. Therefore, we developed and tested a new instrument, the SERSWCS. This article reports the findings regarding the following research questions and associated hypotheses.
Research Question (RQ1): How internally consistent are the data produced using the SERSWC? Are the estimates relatively consistent across the three cohorts in this study? It should be noted that test–retest reliability was not assessed, as self-efficacy is not conceptualized as a trait (Bandura, 1997, p. 45). We expected to find high levels of internal consistency for the scale as we have in the past with our other self-efficacy scales.

Research Question (RQ2): What evidence is there for the content validity of the SERSWCS?

Research Question (RQ3): Is there evidence of convergent validity for the SERSWCS? Based on our studies of the Foundation Practice Self-Efficacy (FPSE) scale and the Social Worker Empowerment (SWE) scale, we expected that a large positive association (i.e., r = .57 mean of prior studies) between students’ level of self-efficacy (SERSWCS) and their level of empowerment (SWE) would emerge. Therefore, we explicitly tested the null hypothesis that the observed SERSWCS–SWE correlation was equal to r = .57.

Research Question (RQ4): Is the SERSWCS sensitive enough to detect a statistically significant pre–post change in these samples? In other words, we tested the null hypothesis that pre–post change on the SERSWCS total scale score was equal to zero. We expected that like other self-efficacy scales we have developed, this scale would be sensitive to pre–post changes. In addition, the magnitude of the change was estimated.

Method

This article describes three concurrent studies (I, II, and III) within one school (New York University: Silver School of Social Work) that employed a pretest–posttest design using the same measures, with three different groups of master’s level students, where the pretest–posttest assessments covered different intervals. Data administrations were not done in the first or last classes of the semesters to minimize disruption. These were voluntary and anonymous surveys, administered by one of the investigators at both pre- and posttest. As soon as the process was explained to students, the administrator left the room to preserve the sense of anonymity. In this series of self-efficacy scale development studies, we have argued that it is important that respondents believe they are anonymous in order to reduce the press for socially desirable responding (Krumpal, 2013; Paulhus, 1991). Therefore, we have used a form of respondent generated identification numbers (aka “Hogben numbers”) to allow matching of pretest and posttest responses (Honig, 1995).

Samples

The overall sample across studies was comprised of 550 students who completed both the pretest and posttest. This overall sample contained three studies with different pre–post time intervals.

Study I

This school has a 16-month accelerated program where students with prior experience in the field complete a full 2-year MSW program in four consecutive semesters (spring, summer, fall, and spring). Two cohorts of these students were assessed near the beginning of their first and near the end of their second semester (6 months later). The sample consisted of 140 students who completed the measures at pretest. Of this sample, 108 students also completed the measure at posttest.

Study II

The majority of students in this school enroll in a typical 2-year sequence (fall, spring, fall, and spring). This cohort was assessed near the beginning of their program and near the end of their second semester (7 months later). In total, 298 students completed the measures at pretest and 238 students at posttest.

Study III

The third cohort of students was also from the typical 2-year sequence but had begun the Study 1 year ahead of the students in Study II. They were assessed near the beginning of their program and near the end of their fourth semester (19 months later). The total sample consisted of 330 students. At posttest, 204 students also completed the measure.

Measures

SERSWCS

The SERSWCS is a 41-item scale, developed in line with Bandura’s (2006) self-efficacy scale development guide, which was administered at pretest and posttest. The 41 items are comprised of the 41 practice behaviors that CSWE indicates are representative of the 10 competencies. Respondents indicated their level of confidence by choosing any of the 11 responses from 0 (cannot do) to 100 (highly certain can do). Total scores were computed as the mean of the 41 items. The instructions and scale are written at a Flesch-Kincaid Grade Level of 11.8/Flesch Reading Ease of 33.9. The SERSWCS is available to anyone to use at no cost by simply e-mailing the first author.

SWE Scale

The same comparison scale we had used in earlier studies was used here. Frans’ (1993) SWE was used at pretest in this study to assess the convergent (construct) validity of the Evaluation Self Efficacy. The SWE was chosen because of the conceptual similarity of the self-efficacy and empowerment constructs. The SWE has performed adequately in other self-efficacy scale development studies we have conducted. It is an easy-to-read (Flesch-Kincaid grade-level readability estimate = 8.2), 34-item, self-report measure with support for the reliability (e.g., α values of .88 and .89) and validity (e.g., correlation of .53 with the Torre Empowerment Scale) of the data collected with it (Frans, 1993). Respondents indicate their level of empowerment via a Likert scale from 1 (strongly agree) to 5 (strongly disagree). Total scores were computed as the mean of the 34 reversed-scored items. In this study, SWE yielded good internal consistency with Cronbach α values of α = .89 (Study I), .88 (Study II), and .85 (Study III).

Data Analysis Strategy

As a measure of internal consistency, Cronbach’s α and item–rest correlations were computed. The lowest acceptable values for Cronbach’s α and item–rest correlations were set at .70 and .20, respectively (Nunnally & Bernstein, 1994). In addition, principal component analysis was used to explore whether the SERSWCS items could be reduced to a single dimension. Pearson correlations were used to examine convergent validity with criteria for a small, medium, and large correlation set at .10, .30, and .50, respectively (Cohen, 1988). The Fisher r-to-Z transformation (Cohen & Cohen, 1983) was applied to compare the resulting correlations to those found in previous studies (r = .57 mean of prior studies) with the magnitude of the difference between both correlations computed as q = |Z_r – Z_r0 | which can be interpreted using the same conventions as reported earlier. A repeated-measures analysis of variance (ANOVA) specifying a within-subjects factor (pre- vs. posttest) and a between-subjects factor (study) was used to test the main effects of time point and study as well as the interaction between both. Effect sizes in the repeated-measures ANOVA were computed as η_p ² with values of .01, .06, and .13, indicating a small, medium, and large effect (Cohen, 1988). Follow-up paired sample t-tests were used to gain insight into the nature of a significant interaction effect, with effect sizes computed as Cohen’s d = mean of the difference/SD mean. Values of .30, .50, and .80 were considered to reflect a small, medium, and large difference (Cohen, 1988). The significance level was set at p < .05, unless otherwise stated.

Results

Individual Item-Level Data

We continue to agree with Cronbach’s (1963) view that inspection of individual scale item data is useful for curricular improvement. Tables 1 through 3 present the pre- and posttest means (and SD) for each of the 41 SERSWCS items for all three studies. Across studies, increases in students’ self-reported self-efficacy were observed for all 41 items, with pre-to-posttest differences ranging from 8.41 to 33.33. For Study I (Table 1), the pre–post difference was the smallest for Item 4 and the largest for item 40. For Study II (Table 2), the pre–post difference was the smallest for Item 30 and the largest for Item 1. For Study III (Table 3), Item 5 had the smallest pre–post change, whereas Item 40 had the largest.

Table 1.
Item-Level Statistics for SERSWCS at Both Time Points for Study I.

Item Pretest Posttest Pre–Post Difference Item-Rest Correlations PCA Loadings

n M SD n M SD M SD Pretest Posttest Pretest Posttest

1 140 48.79 26.67 108 76.48 18.26 26.85 27.94 .694 .601 .711 .611

2 139 68.99 20.72 108 85.46 11.63 16.45 20.98 .480 .695 .501 .713

3 139 73.60 19.93 108 86.48 12.10 12.71 19.26 .511 .707 .522 .733

4 139 82.30 16.48 108 91.02 11.91 8.41 15.55 .416 .635 .428 .665

5 139 80.22 18.51 108 89.54 13.21 10.00 17.70 .476 .750 .488 .769

6 139 75.97 18.52 108 86.57 14.86 10.47 20.85 .557 .675 .575 .701

7 139 70.43 20.53 108 85.46 12.85 14.67 19.87 .569 .791 .578 .808

8 138 65.00 26.03 108 83.61 14.43 17.55 24.17 .555 .713 .568 .730

9 139 61.87 20.98 108 80.37 13.04 18.60 20.90 .541 .770 .555 .783

10 139 67.34 20.06 108 83.80 12.66 15.33 19.10 .665 .815 .678 .826

11 138 58.04 24.10 108 79.63 15.22 22.17 24.22 .715 .763 .733 .768

12 138 52.17 24.78 108 76.02 16.74 22.64 26.81 .763 .709 .779 .716

13 138 65.00 21.82 106 82.83 13.43 18.17 22.15 .746 .788 .761 .809

14 138 65.58 21.07 108 84.54 13.14 19.81 20.70 .663 .677 .680 .698

15 138 65.43 19.86 108 82.41 13.10 17.17 20.92 .649 .715 .666 .737

16 137 71.75 20.00 108 86.11 12.29 14.62 21.16 .634 .811 .649 .830

17 136 77.06 18.10 108 88.33 12.86 11.44 19.38 .593 .765 .613 .789

18 138 66.67 21.28 108 85.56 12.02 18.21 20.41 .651 .786 .666 .802

19 138 64.28 24.67 108 81.39 17.59 16.98 22.98 .645 .641 .660 .636

20 138 61.38 24.68 108 80.65 18.56 19.91 24.36 .685 .642 .697 .640

21 139 48.63 24.97 107 72.15 16.60 23.87 26.89 .737 .676 .753 .677

22 139 53.67 25.46 108 77.50 14.98 24.21 27.20 .749 .652 .765 .664

23 138 54.35 23.77 108 79.63 14.53 26.32 27.34 .789 .763 .809 .773

24 138 64.42 20.33 108 85.00 12.72 20.85 22.00 .783 .802 .801 .822

25 138 51.45 25.30 107 75.42 17.39 24.15 25.37 .722 .636 .738 .633

26 139 57.77 24.96 108 77.13 17.98 19.44 25.28 .652 .644 .672 .640

27 138 55.80 22.76 108 77.87 15.17 22.99 24.08 .671 .773 .685 .777

28 139 50.94 25.73 108 76.85 15.20 25.70 26.53 .769 .766 .786 .766

29 138 55.72 21.51 106 80.38 14.47 24.67 22.58 .742 .808 .764 .819

30 138 81.67 16.24 107 91.96 11.28 10.38 17.59 .578 .791 .596 .816

31 137 68.47 19.96 107 86.64 12.28 18.94 20.14 .764 .790 .782 .815

32 138 62.32 21.69 107 82.99 13.40 20.86 21.98 .741 .713 .766 .735

33 138 66.52 20.53 107 86.82 12.33 21.14 20.54 .775 .743 .799 .772

34 138 65.00 21.69 107 85.23 12.46 19.71 22.34 .813 .789 .836 .815

35 138 53.77 23.37 107 80.47 14.03 27.52 24.45 .798 .796 .820 .810

36 138 54.71 22.29 107 80.37 14.00 25.62 24.14 .724 .791 .746 .807

37 138 52.97 23.22 106 80.09 14.90 26.64 25.14 .797 .804 .817 .820

38 138 62.25 22.18 107 82.43 14.46 20.19 21.84 .736 .787 .758 .813

39 138 61.88 23.07 107 83.18 13.36 21.33 22.70 .795 .844 .811 .854

40 138 51.59 23.81 107 79.72 13.21 29.52 24.78 .737 .681 .758 .709

41 138 52.54 25.77 107 80.47 13.69 28.19 25.97 .776 .762 .799 .779

Note. SERSWCS = Self-Efficacy Regarding Social Work Competencies Scale; PCA = principal component analysis. N = 140. Values in boldface indicate items with lowest and highest pre–post difference.

Table 2.
Item-Level Statistics for SERSWCS at Both Time Points for Study II.

Item Pretest Posttest Pre–Post Difference Item-Rest Correlations PCA Loadings

n M SD n M SD M SD Pretest Posttest Pretest Posttest

1 298 51.04 26.21 238 79.45 16.13 28.74 26.47 .653 .677 .669 .689

2 298 67.99 20.97 238 85.25 13.17 17.44 21.12 .572 .620 .589 .640

3 298 70.77 21.11 237 86.62 12.13 16.41 22.06 .594 .650 .608 .670

4 298 81.61 17.51 238 91.51 10.68 10.34 17.93 .499 .602 .514 .621

5 298 80.77 16.63 238 90.97 12.30 10.00 18.14 .549 .507 .563 .521

6 298 77.05 20.47 236 89.36 13.90 11.61 22.27 .518 .476 .533 .494

7 296 70.98 18.80 237 86.20 11.31 15.67 20.25 .655 .557 .668 .575

8 297 66.57 23.04 238 84.33 13.26 17.76 23.28 .634 .647 .648 .668

9 297 62.49 20.18 238 80.21 14.33 17.89 21.78 .617 .642 .632 .663

10 297 66.63 19.99 238 81.51 13.13 15.40 21.18 .688 .703 .701 .719

11 298 58.79 22.61 238 79.20 13.86 21.09 23.86 .732 .769 .749 .781

12 298 55.30 22.92 238 78.03 14.17 23.24 23.31 .754 .750 .769 .767

13 298 64.03 21.82 238 83.15 12.88 20.00 21.84 .675 .729 .693 .750

14 298 67.58 20.31 238 86.13 11.63 18.78 20.43 .674 .592 .691 .612

15 298 66.54 19.81 238 83.78 12.19 17.69 21.74 .669 .586 .691 .608

16 298 71.98 18.45 238 87.48 10.69 16.09 19.69 .693 .636 .714 .657

17 298 79.26 17.04 238 90.42 10.18 11.30 17.39 .544 .542 .563 .563

18 298 68.32 20.11 238 86.30 11.46 18.03 20.17 .694 .646 .712 .663

19 298 66.28 21.19 238 82.61 14.46 16.68 21.35 .643 .665 .660 .677

20 298 64.03 21.00 238 81.01 15.45 16.77 22.22 .676 .723 .691 .732

21 297 51.41 24.51 238 73.95 16.97 22.57 25.36 .715 .768 .730 .781

22 297 54.18 23.27 237 77.59 15.88 22.71 24.57 .674 .762 .690 .775

23 297 54.44 22.82 238 80.08 14.52 25.61 24.15 .764 .816 .778 .831

24 297 64.78 20.25 238 85.42 11.96 21.35 20.68 .803 .792 .816 .809

25 296 51.59 24.09 237 74.47 16.93 23.70 24.74 .791 .743 .804 .750

26 297 57.07 24.67 237 77.64 17.01 21.06 25.62 .695 .686 .711 .695

27 297 56.46 23.54 238 77.44 15.22 21.31 24.64 .768 .732 .782 .742

28 297 52.76 24.19 238 75.50 17.32 22.83 25.48 .773 .779 .788 .788

29 297 57.41 23.24 233 81.37 12.59 24.03 24.50 .793 .789 .809 .806

30 298 83.56 14.71 235 93.40 8.49 9.28 13.83 .561 .567 .581 .596

31 297 68.96 18.67 233 86.44 11.62 17.24 20.28 .730 .753 .751 .777

32 298 64.36 22.94 235 83.06 13.40 18.81 23.77 .686 .680 .710 .708

33 298 65.00 20.63 234 87.00 11.33 22.09 21.38 .767 .734 .787 .760

34 297 64.44 21.06 233 84.72 12.66 20.22 23.27 .802 .708 .821 .736

35 297 55.22 23.47 234 79.79 13.79 24.76 25.36 .792 .767 .812 .788

36 298 57.21 22.76 235 79.45 14.06 22.60 24.54 .776 .627 .794 .649

37 298 54.77 23.30 235 80.00 14.08 25.62 24.91 .816 .752 .832 .774

38 296 62.20 21.59 235 81.68 12.85 19.59 21.77 .714 .675 .736 .696

39 298 63.29 22.49 235 84.79 13.23 21.68 21.98 .767 .738 .784 .754

40 298 53.12 24.15 235 79.19 14.84 26.60 25.71 .757 .681 .776 .703

41 297 53.70 24.21 234 80.43 13.95 27.90 26.09 .817 .715 .833 .736

Note. SERSWCS = Self-Efficacy Regarding Social Work Competencies Scale; PCA = principal component analysis. N = 298. Values in boldface indicate items with lowest and highest pre–post difference.

Table 3.
Item-Level Statistics for SERSWCS at Both Time Points for Study III.

Item Pretest Posttest Pre–Post Difference Item-Rest Correlations PCA Loadings

n M SD n M SD M SD Pretest Posttest Pretest Posttest

1 329 50.91 24.95 204 81.08 15.24 32.21 25.74 .690 .647 .706 .659

2 329 69.24 21.45 204 88.14 11.07 18.28 21.21 .616 .599 .631 .632

3 330 70.91 19.49 204 89.36 11.32 19.17 20.67 .562 .622 .576 .647

4 330 80.48 17.09 204 92.25 11.00 11.96 18.06 .450 .667 .464 .691

5 330 80.91 17.82 204 91.35 11.04 9.68 17.69 .512 .536 .525 .555

6 328 78.23 19.12 202 89.80 10.74 11.44 19.86 .545 .558 .560 .579

7 327 70.24 19.30 204 87.94 9.03 18.57 19.64 .654 .607 .667 .632

8 329 65.23 24.37 204 87.13 10.56 22.97 25.10 .676 .635 .689 .651

9 327 60.28 22.17 204 82.38 11.74 21.97 22.21 .623 .599 .638 .620

10 330 65.42 21.96 204 84.58 11.54 18.55 20.65 .692 .660 .704 .673

11 328 59.60 24.63 203 83.45 12.82 24.36 23.83 .762 .708 .774 .727

12 329 55.96 25.91 203 82.56 13.44 27.48 25.32 .779 .727 .792 .741

13 327 64.46 23.41 203 88.52 11.12 25.55 22.00 .714 .600 .728 .632

14 329 67.42 20.10 203 88.50 10.64 22.30 19.84 .643 .674 .658 .692

15 329 66.20 19.47 203 87.68 10.20 21.24 19.97 .670 .659 .687 .691

16 329 72.13 18.90 203 89.70 10.00 17.72 19.89 .724 .728 .739 .756

17 329 76.84 19.28 203 90.84 10.61 14.70 19.22 .654 .678 .671 .704

18 328 69.12 20.22 203 88.03 10.99 20.00 20.05 .704 .626 .720 .644

19 328 65.21 23.63 203 82.73 14.46 18.34 23.80 .756 .632 .770 .638

20 328 64.30 23.65 203 82.17 14.29 18.71 24.20 .751 .669 .764 .674

21 327 52.11 25.50 204 77.94 14.61 26.50 25.10 .800 .700 .811 .706

22 329 56.14 25.24 204 80.74 14.59 25.20 24.98 .757 .654 .771 .663

23 327 55.60 24.50 204 82.60 13.56 28.22 24.04 .784 .731 .797 .741

24 328 65.09 22.96 204 88.58 10.80 24.88 23.34 .825 .793 .836 .809

25 327 54.46 25.29 204 79.31 15.33 26.58 26.65 .792 .705 .805 .701

26 329 59.36 25.79 203 78.97 16.93 20.34 27.55 .755 .657 .769 .659

27 329 57.54 24.58 204 80.42 14.92 24.19 26.04 .776 .724 .789 .726

28 329 53.22 25.23 204 79.19 15.79 27.97 24.38 .792 .716 .803 .721

29 330 56.00 24.13 201 85.47 11.83 30.75 25.06 .822 .802 .836 .820

30 330 82.61 15.92 203 95.71 7.37 13.00 16.27 .593 .550 .612 .586

31 329 68.57 20.38 203 89.66 10.69 21.28 20.86 .739 .695 .757 .732

32 330 64.76 23.84 202 86.14 13.27 22.43 23.70 .752 .611 .770 .632

33 329 65.59 21.61 203 91.77 9.59 27.14 21.40 .805 .724 .821 .758

34 329 64.19 22.20 203 89.85 10.07 25.76 20.96 .797 .752 .814 .785

35 328 53.90 23.03 202 85.40 11.89 32.39 22.32 .830 .743 .845 .767

36 329 56.38 24.19 203 84.33 13.42 28.87 23.78 .819 .721 .834 .739

37 329 53.68 23.96 203 85.27 12.60 32.41 24.16 .819 .753 .835 .776

38 328 62.47 22.50 203 87.56 10.77 25.15 21.73 .788 .690 .805 .722

39 327 61.77 23.10 203 88.33 11.70 27.38 23.08 .775 .779 .792 .796

40 329 52.89 24.09 203 85.34 12.11 33.33 24.37 .783 .691 .800 .720

41 329 53.43 24.92 203 84.73 13.69 32.71 26.20 .817 .680 .832 .701

Note. SERSWCS = Self-Efficacy Regarding Social Work Competencies Scale; PCA = principal component analysis. N = 330. Values in boldface indicate items with lowest and highest pre–post difference.

Internal Consistency

In terms of our first research question (RQ1), a principal component analysis was conducted on the 41 items SERSWCS, separately for each study and at each measurement point. The scree plots clearly pointed toward extracting a single component. In each analysis, substantial loadings (range .43–.85) were obtained for each item which underpins the unidimensionality of the SERSWCS (Table 1 through 3). As can be seen in the same Tables, all item–rest correlations, ranging from .42 to .84, clearly exceeded the .20 threshold. Across studies, Cronbach’s αs at pre- and posttest ranged from α = .97 to .98 (Table 4), which also noticeably surpassed the .70 threshold at a consistent level. As such, the results for the three study cohorts described here point toward excellent internal consistency of the SERWCS. Although one could even argue that the very high α values may be indicative of item redundancy, an inspection of the factor loadings and item–rest correlations across the different samples and measurement points did not suggest that certain items were consistently overly redundant. Combined with the different content of each item, it seems more likely that the SERWCS items tap into closely related, but distinct, indicators of self-efficacy regarding social work.

Table 4.
SERSWCS Total Score Results.

Study I Study II Study III

Pretest Posttest Pretest Posttest Pretest Posttest

N 140 108 298 238 330 204

Min–Max 20.98–90.75 40.73–99.50 10.73–97.56 40.00–98.54 7.80–99.02 57.56–100

M (SD) 62.74 (15.58) 82.42 (10.52) 63.51 (15.57) 82.86 (9.45) 63.52 (16.75) 86.00 (8.40)

α .974 .980 .976 .974 .979 .972

Note. SERSWCS = Self-Efficacy Regarding Social Work Competencies Scale.

Validity

Content validity

RQ2 asks if there is any evidence regarding content validity. As Holloway (2013) has noted “[u]nless measurement tools measure the specific practice behaviors selected to operationalize the desired program competencies, they do not meet the requirements of program assessment as defined by the 2008 EPAS” (p. 2, emphasis added). This requirement supports the decision we have made for some time, to tailor specific self-efficacy scale items to such external requirements (e.g., accreditation standards, course objectives, etc.). This makes our assertion that the SERSWCS produces data with content validity very simple. In other words, the items on the SERSWCS ask respondents about each of the 41 behaviors that CSWE states represent the 10 competencies. The same wording is used with a minor exception for Item 17 (changed view themselves as learners to see yourself as a learner). The SERSWCS covers the behaviors that CSWE says represent the 10 competencies precisely.

Convergent validity

RQ3 focuses on the convergent validity of the data produced by the SERSWCS. Based on two prior studies (Holden et al., 2003, 2005) using the SWE, we predicted that the SERSWCS would have a large positive relationship with the SWE (specifically, r = .57). Across the three studies reported here, a significant positive correlation of r = .58, .46, and .47 (all ps < .001) was indeed found between SERSWCS and SWE at pretest. This correlation was not statistically significantly different from the expected correlation of r = .57 for Study I (Z = 0.18, p = .861), meaning that the strength of the association was in line with our expectation. Conversely, a statistically significant small difference did emerge for Study II (Z = 2.62, p = .009, q = .15) and Study III (Z = 2.57, p = .010, q = .14), indicating that the obtained correlations were slightly smaller than the expected correlation of r = .57.

Sensitivity to Change

The descriptive statistics for the SERSWCS total scores are presented in Table 4. Across studies, about 28.39% of the cases were missing posttest data. A formal comparison of cases with and without missing data on SERSWCS posttest scores did not indicate significant differences with regard to pretest study variables, SERSWCS: t(354) = −1.05, p = .295; SWE: t(765) = −1.59, p = .113.

The repeated-measures ANOVA performed to answer RQ4, indicated a small significant interaction effect between time point and study, F(2, 547) = 3.281, p = .038, η_p ² = .012. To enhance interpretation, we conducted follow-up paired sample t-tests, which revealed statistically significant pre-to-posttest differences for all three studies (Bonferonni adjusted p < .017). In line with expectations, the SERSWCS was thus sensitive enough to detect increases in self-efficacy over time. All three effects can be considered large, but the pre-to-posttest change was more pronounced for Study III (mean difference = 23.25, t(203) = 20.81, p < .001, d = 1.46) compared to the fairly similar effect sizes obtained in Study I (mean difference = 19.65, t(107) = 12.65, p < .001, d = 1.22) and Study II (mean difference = 19.61, t(237) = 18.81, p < .001, d = 1.22). It should be noted that normality and/or sphericity assumptions associated with these parametric tests were met.

Discussion and Applications to Social Work Education

As Brandt et al. (2014) note, “replications are therefore essential for theoretical development through confirmation and disconfirmation of results” (p. 218). The goal of the studies reported here was to construct and test a new measure of social work educational outcomes that was based on prior theoretical and empirical work and would be suitable for programs to use for accreditation purposes, as well as program evaluation/curriculum development. The three studies reported here investigate SERSWCS performance with three different samples over three different time periods (N = 550). The SERSWCS generally performed in line with our prior expectations. Differences between the three studies were detailed. More generally, this series of three studies is close to our prior self-efficacy scale development studies in terms of populations, methods, and constructs employed. They are different from our prior work in terms of the time period in which the studies were conducted, as well as the participants, scale directions, pretest–posttest time frame, and behaviors assessed in relation to self-efficacy. The SERSWCS performed in a manner similar to that of our other self-efficacy scales.

In terms of internal consistency, we have repeatedly observed Cronbach’s αs greater than .90 with other self-efficacy scales that we have developed. In the current studies, the six αs all exceeded .96. Moreover, the results of six principal components analyses and six sets of item–rest correlations all provide strong support for the unidimensionality of the SERSWCS. Despite the substantial homogeneity, we argued that items were not overly redundant, but future research could examine this issue more closely.

As noted previously, one critique of self-efficacy (i.e., restricting it to one’s ability to “learn”) is mistaken. In this study, as in prior ones, we asked individuals about their confidence in their ability to successfully perform a task in a particular context. The SERSWCS scale instructions read, in part:
We want to know how confident you are, in your ability to perform specific social work tasks. After you consider each task, please rate your confidence in your ability to perform that task successfully, by circling the number from 0 to 100 that best describes your level of confidence. What we mean here by successfully, is that you would be able to perform the specific task in a manner that a social work supervisor would consider excellent.… We are not interested in how confident you predict you will be when you finish this program but rather we want to know how confident you are that you could successfully perform these tasks NOW (today).
In terms of validity, Drisko (2014) raised the item–competency linkage concern in his discussion of self-report surveys. Coverage of the each of the 41 practice behaviors in the SERSWC (in virtually the same language used by CSWE) provides clear evidence regarding content validity. While a rigorous approach (validity effect size estimation), our convergent validity predictions regarding the direction and size of the correlations between the SERSWCS and the SWE were somewhat equivocal. Across studies, the direction and strength of the association between self-efficacy and empowerment was general in line with our expectations, namely a strong positive correlation. Yet, while the observed correlations were not significantly different than our predicted correlation of r = .57 for Study I (r = .58), they were for Study II (r = .46) and III (r = .47), two things should be noted regarding this finding. First, perhaps making this prediction based on only two prior correlations between a similar self-efficacy scale (the FPSE scale) and the SWE was premature. Second, the obtained differences with the predicted correlation were small in size and it seems likely that the statistical significant effects were partly due to the increased statistical power of Studies II and III given their larger sample size. The question of the practical importance of these small, statistically significant differences remains.

Agreeing with Cronbach (1963), we reported results at the individual item level. The wide range of pre–post differences across SERSWCS items provides data that may inform curricular adjustments. Moreover, this level of detail may be more informative than dichotomous decisions whether or not a result exceeds a preset benchmark. Consider SERSWCS Item 21 (use practice experience to inform scientific inquiry). For instance, suppose we had set a benchmark that students average 80.00 or greater on SERSWCS items at the posttest. In Study 1, the students averaged 72.15 at posttest. Contrast this with Item 4 (demonstrate professional demeanor in behavior, appearance, and communication) for which students averaged 91.02. Students were more confident on Item 4 than they were on Item 21. The finding that the benchmark was exceeded for Item 4 but not for Item 21 has value. Yet, in terms of “value added,” there was an increase in self-efficacy of 8.41 on Item 4 versus an increase of 23.87 for Item 21. That finding has value as well. Of course, there are additional nuances to consider (e.g., ceiling effects, time devoted to the topic in the curriculum, etc.), but our point here is that restricting examination of the data to total scale scores or to dichotomous attainment of benchmarks without recognizing where students began would curtail our understanding of the diversity of changes that are occurring in our students. Information that could suggest curricular changes might be missed. That is not to insinuate that total scale scores are unimportant. As we have consistently found over the years with other self-efficacy scales, the total scale scores for the SERSWCS increased significantly from pre- to posttest across all three studies. In Studies I and II, this significant change was over the first two semesters. Study III had a significant pre–post change (across four semesters) that was greater than the pre–post change in Studies I and II. That said, although the change in Study III was more pronounced, the statistically significant difference in pre–post change between studies should not be overstated since it was small in size.

One of our reviewers remarked on the high levels of student self-efficacy at pretest (i.e., total SERSWCS scale scores > 62). From the perspective of a seasoned social worker, this could be seen as overconfidence relative to the knowledge and skills these respondents possess. In fact, our prior research on response shift bias demonstrated that newer social work students rated themselves significantly higher on pretest surveys than they subsequently thought they should have (after they better understood the behavioral performances involved). If we were claiming that self-efficacy was a perfect equivalent of actual competency, those high pretest ratings would bother us as well. Our view is that if the field ever develops psychometrically sound measures of competency , they will be moderately positively correlated with self-efficacy regarding those competencies. Measurement error in both self-efficacy and competency measures will mitigate whatever the less than perfect correlation there is between self-efficacy and competency true scores. Moreover, as an evaluation aside, if our pretest scores are inflated, then pretest–posttest assessments of change provide more conservative significance tests.

As always there are some caveats, readers should keep in mind in thinking about these results. Despite three studies with consistent findings, these were one group pretest–posttest designs, with nonrandom samples and attrition, carried out in a single school, by a single group of investigators. The number of complete pretest–posttest pairs is typically less than the number of pretests completed and this problem is worse over the four semester interval. Our last five self-efficacy scale development studies (covering different pre–post period lengths) had attrition rates ranging from 21% to 45%, compared to the 28% for this current set of studies. This current rate is somewhat larger than the rates reported by Hanson, Tobler, and Graham (1990) in their meta-analysis of attrition in substance abuse prevention research. They report that 21.7% of subjects (on average across 85 cohorts) were missing 6 months after the pretest, 26.6% were missing after 12 months, and 28.2% were missing after 24 months. There are a variety of explanatory possibilities here (e.g., refusals to participate, late for class or absent at the point of administration, leave of absence, shift into a different program track or to a different campus, dropped out, confused by the identification system, lack of incentives for participation, no threat of being identified as a dropout, participation discouraged by classroom faculty, etc.). Comparison of cases with and without missing data on SERSWCS posttest scores did not reveal significant differences with regard to pretest study variables. Nonetheless, more research on the psychometric performance of the SERSWCS, using more representative samples and taking additional measures to reduce attrition, is needed in order increase generalizability. The larger question of which outcome is more important (i.e., smaller attrition rates or reduced press for socially desirable responding) remains.

One of our reviewers was concerned that some of the behaviors specified by CSWE resulted in double-barreled items in our adaptation. We argue that in least some of those instances, the items were complex as opposed to simply asking about two attitudinal objects (“double-barreled”). In a classic text on asking questions, Payne (1951) referred to double-barreled items as deserving to be split but that there were “special cases where two issues necessarily have to be asked about together” (p. 233). In certain instances, when factors are interrelated, they may be asked about together (Payne, 1951). In select instances, we would have preferred to change the wording of items but decided not to in order to preserve the congruence between the results obtained and the specified behaviors that appear interrelated. It was a considered trade-off that necessitated a psychometric choice between keeping the items as written or substantially increasing the number of items, the latter choice avoiding the double (or more)-barreled problem but also potentially inducing response sets and fatigue that would create new problems. On the whole, there might be utility in CSWE considering this issue in future modifications of the EPAS.

These studies raise a number of questions and do not answer the global question: What is the optimal approach to measuring social work educational outcomes? Until the field has measures that produce reliable and valid data regarding student’s actual competence (in a broadly applicable, efficient manner), other indicators of outcome will remain important. Potential changes in the specific competencies (CSWE, 2014) may require revisions to current measures as well as changes in scale development plans. It makes sense for the field to continue working on these problems from multiple perspectives. For instance, the ongoing work of the multi-university Social Work Education Assessment Project (e.g., Hamilton et al., 2011), the University of Toronto Building Professional Competence group (e.g., Bogo et al., 2013; Logie, Bogo, Regehr, & Regehr, 2013), and the Indiana University’s Social Work Office of Education Assessment (e.g., Pike, Bennett, & Chang, 2004).

The curriculum (very broadly defined) is the responsibility of the faculty. Accreditation-related efforts can be extremely costly to faculty and the schools where they are located. Accreditation outcomes are not necessarily benign or valid indicators of quality education. While we believe in faculty engaging in self-assessment, we also believe that faculty have a right and responsibility to thoughtfully question neoliberal calls to accountability that may not be solely focused on improving (or even maintaining) the education our students receive.

Conclusion

There was empirical support for the use of self-efficacy in social work prior to this study and that body of support is now larger. The SERSWCS is a measure that is based on substantial theoretical and empirical work, has preliminary evidence regarding the psychometric properties of the data it produces, can be used with large numbers of students in an efficient manner, is neither expensive or subject to user restrictions, and provides views of outcomes that have utility for pedagogical considerations at multiple curricular levels. Although more work is needed, we believe this is a promising start.

Item	Pretest	Posttest	Pre–Post Difference	Item-Rest Correlations	PCA Loadings
1	140	48.79	26.67	108	76.48	18.26	26.85	27.94	.694	.601	.711	.611
2	139	68.99	20.72	108	85.46	11.63	16.45	20.98	.480	.695	.501	.713
3	139	73.60	19.93	108	86.48	12.10	12.71	19.26	.511	.707	.522	.733
4	139	82.30	16.48	108	91.02	11.91	8.41	15.55	.416	.635	.428	.665
5	139	80.22	18.51	108	89.54	13.21	10.00	17.70	.476	.750	.488	.769
6	139	75.97	18.52	108	86.57	14.86	10.47	20.85	.557	.675	.575	.701
7	139	70.43	20.53	108	85.46	12.85	14.67	19.87	.569	.791	.578	.808
8	138	65.00	26.03	108	83.61	14.43	17.55	24.17	.555	.713	.568	.730
9	139	61.87	20.98	108	80.37	13.04	18.60	20.90	.541	.770	.555	.783
10	139	67.34	20.06	108	83.80	12.66	15.33	19.10	.665	.815	.678	.826
11	138	58.04	24.10	108	79.63	15.22	22.17	24.22	.715	.763	.733	.768
12	138	52.17	24.78	108	76.02	16.74	22.64	26.81	.763	.709	.779	.716
13	138	65.00	21.82	106	82.83	13.43	18.17	22.15	.746	.788	.761	.809
14	138	65.58	21.07	108	84.54	13.14	19.81	20.70	.663	.677	.680	.698
15	138	65.43	19.86	108	82.41	13.10	17.17	20.92	.649	.715	.666	.737
16	137	71.75	20.00	108	86.11	12.29	14.62	21.16	.634	.811	.649	.830
17	136	77.06	18.10	108	88.33	12.86	11.44	19.38	.593	.765	.613	.789
18	138	66.67	21.28	108	85.56	12.02	18.21	20.41	.651	.786	.666	.802
19	138	64.28	24.67	108	81.39	17.59	16.98	22.98	.645	.641	.660	.636
20	138	61.38	24.68	108	80.65	18.56	19.91	24.36	.685	.642	.697	.640
21	139	48.63	24.97	107	72.15	16.60	23.87	26.89	.737	.676	.753	.677
22	139	53.67	25.46	108	77.50	14.98	24.21	27.20	.749	.652	.765	.664
23	138	54.35	23.77	108	79.63	14.53	26.32	27.34	.789	.763	.809	.773
24	138	64.42	20.33	108	85.00	12.72	20.85	22.00	.783	.802	.801	.822
25	138	51.45	25.30	107	75.42	17.39	24.15	25.37	.722	.636	.738	.633
26	139	57.77	24.96	108	77.13	17.98	19.44	25.28	.652	.644	.672	.640
27	138	55.80	22.76	108	77.87	15.17	22.99	24.08	.671	.773	.685	.777
28	139	50.94	25.73	108	76.85	15.20	25.70	26.53	.769	.766	.786	.766
29	138	55.72	21.51	106	80.38	14.47	24.67	22.58	.742	.808	.764	.819
30	138	81.67	16.24	107	91.96	11.28	10.38	17.59	.578	.791	.596	.816
31	137	68.47	19.96	107	86.64	12.28	18.94	20.14	.764	.790	.782	.815
32	138	62.32	21.69	107	82.99	13.40	20.86	21.98	.741	.713	.766	.735
33	138	66.52	20.53	107	86.82	12.33	21.14	20.54	.775	.743	.799	.772
34	138	65.00	21.69	107	85.23	12.46	19.71	22.34	.813	.789	.836	.815
35	138	53.77	23.37	107	80.47	14.03	27.52	24.45	.798	.796	.820	.810
36	138	54.71	22.29	107	80.37	14.00	25.62	24.14	.724	.791	.746	.807
37	138	52.97	23.22	106	80.09	14.90	26.64	25.14	.797	.804	.817	.820
38	138	62.25	22.18	107	82.43	14.46	20.19	21.84	.736	.787	.758	.813
39	138	61.88	23.07	107	83.18	13.36	21.33	22.70	.795	.844	.811	.854
40	138	51.59	23.81	107	79.72	13.21	29.52	24.78	.737	.681	.758	.709
41	138	52.54	25.77	107	80.47	13.69	28.19	25.97	.776	.762	.799	.779

Item	Pretest	Posttest	Pre–Post Difference	Item-Rest Correlations	PCA Loadings
1	298	51.04	26.21	238	79.45	16.13	28.74	26.47	.653	.677	.669	.689
2	298	67.99	20.97	238	85.25	13.17	17.44	21.12	.572	.620	.589	.640
3	298	70.77	21.11	237	86.62	12.13	16.41	22.06	.594	.650	.608	.670
4	298	81.61	17.51	238	91.51	10.68	10.34	17.93	.499	.602	.514	.621
5	298	80.77	16.63	238	90.97	12.30	10.00	18.14	.549	.507	.563	.521
6	298	77.05	20.47	236	89.36	13.90	11.61	22.27	.518	.476	.533	.494
7	296	70.98	18.80	237	86.20	11.31	15.67	20.25	.655	.557	.668	.575
8	297	66.57	23.04	238	84.33	13.26	17.76	23.28	.634	.647	.648	.668
9	297	62.49	20.18	238	80.21	14.33	17.89	21.78	.617	.642	.632	.663
10	297	66.63	19.99	238	81.51	13.13	15.40	21.18	.688	.703	.701	.719
11	298	58.79	22.61	238	79.20	13.86	21.09	23.86	.732	.769	.749	.781
12	298	55.30	22.92	238	78.03	14.17	23.24	23.31	.754	.750	.769	.767
13	298	64.03	21.82	238	83.15	12.88	20.00	21.84	.675	.729	.693	.750
14	298	67.58	20.31	238	86.13	11.63	18.78	20.43	.674	.592	.691	.612
15	298	66.54	19.81	238	83.78	12.19	17.69	21.74	.669	.586	.691	.608
16	298	71.98	18.45	238	87.48	10.69	16.09	19.69	.693	.636	.714	.657
17	298	79.26	17.04	238	90.42	10.18	11.30	17.39	.544	.542	.563	.563
18	298	68.32	20.11	238	86.30	11.46	18.03	20.17	.694	.646	.712	.663
19	298	66.28	21.19	238	82.61	14.46	16.68	21.35	.643	.665	.660	.677
20	298	64.03	21.00	238	81.01	15.45	16.77	22.22	.676	.723	.691	.732
21	297	51.41	24.51	238	73.95	16.97	22.57	25.36	.715	.768	.730	.781
22	297	54.18	23.27	237	77.59	15.88	22.71	24.57	.674	.762	.690	.775
23	297	54.44	22.82	238	80.08	14.52	25.61	24.15	.764	.816	.778	.831
24	297	64.78	20.25	238	85.42	11.96	21.35	20.68	.803	.792	.816	.809
25	296	51.59	24.09	237	74.47	16.93	23.70	24.74	.791	.743	.804	.750
26	297	57.07	24.67	237	77.64	17.01	21.06	25.62	.695	.686	.711	.695
27	297	56.46	23.54	238	77.44	15.22	21.31	24.64	.768	.732	.782	.742
28	297	52.76	24.19	238	75.50	17.32	22.83	25.48	.773	.779	.788	.788
29	297	57.41	23.24	233	81.37	12.59	24.03	24.50	.793	.789	.809	.806
30	298	83.56	14.71	235	93.40	8.49	9.28	13.83	.561	.567	.581	.596
31	297	68.96	18.67	233	86.44	11.62	17.24	20.28	.730	.753	.751	.777
32	298	64.36	22.94	235	83.06	13.40	18.81	23.77	.686	.680	.710	.708
33	298	65.00	20.63	234	87.00	11.33	22.09	21.38	.767	.734	.787	.760
34	297	64.44	21.06	233	84.72	12.66	20.22	23.27	.802	.708	.821	.736
35	297	55.22	23.47	234	79.79	13.79	24.76	25.36	.792	.767	.812	.788
36	298	57.21	22.76	235	79.45	14.06	22.60	24.54	.776	.627	.794	.649
37	298	54.77	23.30	235	80.00	14.08	25.62	24.91	.816	.752	.832	.774
38	296	62.20	21.59	235	81.68	12.85	19.59	21.77	.714	.675	.736	.696
39	298	63.29	22.49	235	84.79	13.23	21.68	21.98	.767	.738	.784	.754
40	298	53.12	24.15	235	79.19	14.84	26.60	25.71	.757	.681	.776	.703
41	297	53.70	24.21	234	80.43	13.95	27.90	26.09	.817	.715	.833	.736

Item	Pretest	Posttest	Pre–Post Difference	Item-Rest Correlations	PCA Loadings
1	329	50.91	24.95	204	81.08	15.24	32.21	25.74	.690	.647	.706	.659
2	329	69.24	21.45	204	88.14	11.07	18.28	21.21	.616	.599	.631	.632
3	330	70.91	19.49	204	89.36	11.32	19.17	20.67	.562	.622	.576	.647
4	330	80.48	17.09	204	92.25	11.00	11.96	18.06	.450	.667	.464	.691
5	330	80.91	17.82	204	91.35	11.04	9.68	17.69	.512	.536	.525	.555
6	328	78.23	19.12	202	89.80	10.74	11.44	19.86	.545	.558	.560	.579
7	327	70.24	19.30	204	87.94	9.03	18.57	19.64	.654	.607	.667	.632
8	329	65.23	24.37	204	87.13	10.56	22.97	25.10	.676	.635	.689	.651
9	327	60.28	22.17	204	82.38	11.74	21.97	22.21	.623	.599	.638	.620
10	330	65.42	21.96	204	84.58	11.54	18.55	20.65	.692	.660	.704	.673
11	328	59.60	24.63	203	83.45	12.82	24.36	23.83	.762	.708	.774	.727
12	329	55.96	25.91	203	82.56	13.44	27.48	25.32	.779	.727	.792	.741
13	327	64.46	23.41	203	88.52	11.12	25.55	22.00	.714	.600	.728	.632
14	329	67.42	20.10	203	88.50	10.64	22.30	19.84	.643	.674	.658	.692
15	329	66.20	19.47	203	87.68	10.20	21.24	19.97	.670	.659	.687	.691
16	329	72.13	18.90	203	89.70	10.00	17.72	19.89	.724	.728	.739	.756
17	329	76.84	19.28	203	90.84	10.61	14.70	19.22	.654	.678	.671	.704
18	328	69.12	20.22	203	88.03	10.99	20.00	20.05	.704	.626	.720	.644
19	328	65.21	23.63	203	82.73	14.46	18.34	23.80	.756	.632	.770	.638
20	328	64.30	23.65	203	82.17	14.29	18.71	24.20	.751	.669	.764	.674
21	327	52.11	25.50	204	77.94	14.61	26.50	25.10	.800	.700	.811	.706
22	329	56.14	25.24	204	80.74	14.59	25.20	24.98	.757	.654	.771	.663
23	327	55.60	24.50	204	82.60	13.56	28.22	24.04	.784	.731	.797	.741
24	328	65.09	22.96	204	88.58	10.80	24.88	23.34	.825	.793	.836	.809
25	327	54.46	25.29	204	79.31	15.33	26.58	26.65	.792	.705	.805	.701
26	329	59.36	25.79	203	78.97	16.93	20.34	27.55	.755	.657	.769	.659
27	329	57.54	24.58	204	80.42	14.92	24.19	26.04	.776	.724	.789	.726
28	329	53.22	25.23	204	79.19	15.79	27.97	24.38	.792	.716	.803	.721
29	330	56.00	24.13	201	85.47	11.83	30.75	25.06	.822	.802	.836	.820
30	330	82.61	15.92	203	95.71	7.37	13.00	16.27	.593	.550	.612	.586
31	329	68.57	20.38	203	89.66	10.69	21.28	20.86	.739	.695	.757	.732
32	330	64.76	23.84	202	86.14	13.27	22.43	23.70	.752	.611	.770	.632
33	329	65.59	21.61	203	91.77	9.59	27.14	21.40	.805	.724	.821	.758
34	329	64.19	22.20	203	89.85	10.07	25.76	20.96	.797	.752	.814	.785
35	328	53.90	23.03	202	85.40	11.89	32.39	22.32	.830	.743	.845	.767
36	329	56.38	24.19	203	84.33	13.42	28.87	23.78	.819	.721	.834	.739
37	329	53.68	23.96	203	85.27	12.60	32.41	24.16	.819	.753	.835	.776
38	328	62.47	22.50	203	87.56	10.77	25.15	21.73	.788	.690	.805	.722
39	327	61.77	23.10	203	88.33	11.70	27.38	23.08	.775	.779	.792	.796
40	329	52.89	24.09	203	85.34	12.11	33.33	24.37	.783	.691	.800	.720
41	329	53.43	24.92	203	84.73	13.69	32.71	26.20	.817	.680	.832	.701

	Study I	Study II	Study III
N	140	108	298	238	330	204
Min–Max	20.98–90.75	40.73–99.50	10.73–97.56	40.00–98.54	7.80–99.02	57.56–100
M (SD)	62.74 (15.58)	82.42 (10.52)	63.51 (15.57)	82.86 (9.45)	63.52 (16.75)	86.00 (8.40)
α	.974	.980	.976	.974	.979	.972

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Ahn

Boykin

Hebert

Kulkin

(2012). Exploring baccalaureate social work students’ self-efficacy: Did it change over time? Journal of Teaching in Social Work, 32, 487–501. doi:10.1080/08841233.2012.724046

Bandura

(1977). Self-efficacy: Toward a unifying theory of behavior change. Psychological Review, 84, 191–215.

Bandura

(1982). Self-efficacy mechanism in human agency. American Psychologist, 37, 122–147.

Bandura

(1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall.

Bandura

(1997). Self-efficacy: The exercise of control. New York, NY: Freeman.

Bandura

(2004). Swimming against the mainstream: The early years from chilly tributary to transformative mainstream. Behaviour Research and Therapy, 42, 613–630. doi:10.1016/j.brat.2004.02.001

Bandura

(2006). Guide for constructing self-efficacy scales. In Pajares

Urdan

T. C.

(Eds.), Self-efficacy beliefs of adolescents (pp. 307–337). Charlotte, NC: Information Age Publishing.

Bogo

(2010). Achieving competence in social work through field education. Toronto, Ontario: University of Toronto Press.

Bogo

Katz

Regehr

Logie

Mylopoulos

Tufford

(2013). Toward understanding meta-competence: An analysis of students’ reflection on their simulated interviews. Social Work Education, 32, 259–273. doi:10.1080/02615479.2012.738662

10.

Brandt

M. J.

Ijzerman

Dijksterhuis

Farach

F. J.

Giner-Sorolla

Grange

J. A.

… van’t Veer

(2014). The replication recipe: What makes for a convincing replication? Journal of Experimental Social Psychology, 50, 217–224. doi:10.1016/j.jesp.2013.10.005

11.

Bruster

B. E.

Coccoma

(2013). Mentoring for educational success: Advancing foster care youth incorporating the core competencies. Journal of Human Behavior in the Social Environment, 23, 388–399. doi:10.1080/10911359.2013.764218

12.

Calderon

(2013). Direct and indirect measures of learning outcomes in an MSW program: What do we actually measure? Journal of Social Work Education, 49, 408–419. doi:10.1080/10437797.2013.796767

13.

Carpenter

(2011). Evaluating social work education: A review of outcomes, measures, research designs and practicalities. Social Work Education, 30, 122–140. doi:10.1080/02615479.2011.540375

14.

Carpenter

Patsios

Szilassy

Hackett

(2011). Outcomes of short course interprofessional education in parental mental illness and child protection: Self-efficacy, attitudes and knowledge. Social Work Education: The International Journal, 30, 195–206. doi:10.1080/02615479.2011.540394

15.

Carpenter

Shardlow

S. M.

Patsios

Wood

(2013). Developing the confidence and competence of newly qualified child and family social workers in England: Outcomes of a national programme. British Journal of Social Work, 45, 153–176. doi:10.1093/bjsw/bct106

16.

Cauble

A. E.

Thurston

L. P.

(2000). Effects of interactive multimedia training on knowledge, attitudes, and self-efficacy of social work students. Research on Social Work Practice, 10, 428–437.

17.

Clem

J. M.

Smith

T. E.

Richards

K. V.

(2012). Effects of a low-element challenge course on abstinence self-efficacy and group cohesion. Research on Social Work Practice, 22, 151–158. doi:10.1177/1049731511423672

18.

Cohen

(1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.

19.

Cohen

(1983). Applied multiple regression/correlation analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.

20.

Cronbach

(1963). Course improvement through evaluation. Teachers’ College Record, 64, 672–683.

21.

Council on Social Work Education. (2008). Educational policy and accreditation standards. Retrieved from http://www.cswe.org/Accreditation.aspx

22.

Council on Social Work Education. (2013). 2008 EPAS handbook (updated 2013). Retrieved from http://www.cswe.org/Accreditation/2008EPASHandbook.aspx

23.

Council on Social Work Education. (2014). Proposed 2015 EPAS revision timeline as of March 2014. Retrieved from http://www.cswe.org/File.aspx?id=72115

24.

Cuzzi

L. C.

Holden

Chernack

Rutter

Rosenberg

(1997). Evaluating social work field instruction: Rotations versus year-long placements. Research on Social Work Practice, 7, 402–414. doi:10.1177/104973159700700308

25.

Cuzzi

L. C.

Holden

Rutter

Rosenberg

Chernack

(1996). A pilot study of fieldwork rotations vs. year long placements for social work students in a public hospital. Social Work in Health Care, 24, 73–91. doi:10.1300/J010v24n01_05

26.

Danis

(2004). Factors that influence domestic violence self-efficacy: Implications for social work. Advances in Social Work, 5, 150–162.

27.

DeLong-Hamilton

Buchan

V. V.

Hull

G. H.

Christenson

Gerritsen-McKane

Rodenhiser

R. W.

Smith

M. L.

(2011). Responding to the 2008 EPAS: Baccalaureate education assessment direct and indirect measurement. The Journal of Baccalaureate Social Work, 16, 17–31.

28.

Drisko

(2014). Competencies and their assessment. Journal of Social Work Education, 50, 414–426. doi:10.1080/10437797.2014.917927

29.

Ellett

(2009). Intentions to remain employed in child welfare: The role of human caring, self-efficacy beliefs, and professional organizational culture. Children and Youth Services Review, 31, 78–88. doi:10.1016/j.childyouth.2008.07.002

30.

Fortune

A. E.

Lee

Cavazos

(2005). Achievement motivation and outcome in social work field education. Journal of Social Work Education, 41, 115–129. doi:10.5175/JSWE.2005.200300318

31.

Frans

D. J.

(1993). A scale for measuring social worker empowerment. Research on Social Work Practice, 3, 312–328. doi:10.1177/104973159300300305

32.

Green

D. M.

(2003). Self-efficacy: A communication model for the development of self-efficacy in the classroom. Journal of Teaching in Social Work, 23, 107–116. doi:10.1300/J067v23n03_09

33.

Hamilton

T. D.

Buchan

V. V.

Hull

G. H.

Christenson

Gerritsen-McKane

Rodenhiser

R. W.

Smith

M. L.

(2011). Responding to the 2008 EPAS: Baccalaureate education assessment direct and indirect measurement. Journal of Baccalaureate Social Work, 16, 17–31.

34.

Hanson

Tobler

N. S.

Graham

J. W.

(1990). Attrition in substance abuse prevention research: A meta-analysis of 85 longitudinally followed cohorts. Evaluation Review, 14, 677–685. doi:10.1177/0193841X9001400608

35.

Holden

Anastas

Meenaghan

(2003). Determining attainment of the EPAS foundation program objectives: Evidence for the use of self-efficacy as an outcome. Journal of Social Work Education, 39, 425–440. doi:10.1080/10437797.2003.10779147

36.

Holden

Anastas

Meenaghan

(2005). EPAS objectives and foundation practice self-efficacy: A replication. Journal of Social Work Education, 41, 559–570. doi:10.5175/JSWE.2005.200404118

37.

Holden

Barker

Meenaghan

Rosenberg

(1999). Research self-efficacy: A new possibility for educational outcomes assessment. Journal of Social Work Education, 3, 463–476. doi:10.1080/10437797.1999.10778982

38.

Holden

Barker

Rosenberg

Onghena

(2007). Assessing progress towards accreditation related objectives: Evidence regarding the use of self-efficacy as an outcome in the advanced concentration research curriculum. Research on Social Work Practice, 17, 456–465. doi:10.1177/1049731506297474

39.

Holden

Barker

Rosenberg

Onghena

(2008). The evaluation self-efficacy scale for assessing progress toward CSWE accreditation-related objectives: A replication. Research on Social Work Practice, 18, 42–46. doi:10.1177/1049731507303954

40.

Holden

Cuzzi

L. C.

Rutter

Rosenberg

Chernack

(1996). The hospital social work self-efficacy scale: Initial development. Research on Social Work Practice, 6, 353–365. doi:10.1177/104973159600600306

41.

Holden

Cuzzi

L. C.

Rutter

Chernack

Rosenberg

(1997). The hospital social work self-efficacy scale: A replication. Research on Social Work Practice, 7, 490–499. doi:10.1177/104973159700700404

42.

Holden

Cuzzi

L. C.

Rutter

Chernack

Spitzer

Rosenberg

(1997). The hospital social work self-efficacy scale: A partial replication and extension. Health & Social Work, 22, 256–263. doi:10.1093/hsw/22.4.256

43.

Holden

Meenaghan

Anastas

Metrey

(2002). Outcomes of social work education: The case for social work self-efficacy. Journal of Social Work Education, 38, 115–133. doi:10.1080/10437797.2002.10779086

44.

Holloway

(2013). Some suggestions on educational program assessment and continuous improvement for the 2008 EPAS. Retrieved from http://www.cswe.org/File.aspx?id=31582

45.

Honig

(1995). When you can’t ask their names: Linking anonymous respondents with the Hogben number. Australian Journal of Public Health, 19, 94–96. doi:10.1111/j.1753-6405.1995.tb00305.x

46.

Icard

L. D.

Schilling

R. F.

El Bassel

(1995). Reducing HIV-Infection among African-Americans by targeting the African-American family. Social Work Research, 19, 153–163. doi:10.1093/swr/19.3.153

47.

Jones

E. A.

Voorhees

R. A.

Paulson

(Eds.), for the Council of the National Postsecondary Education Cooperative Working Group on Competency-Based Initiatives. (2002). Defining and assessing learning: Exploring competency-based initiatives, NCES 2002–159. Washington, DC: U.S. Department of Education, National Center for Education Statistics.

48.

Kranz

K. M.

(2003). Development of the alcohol and other drug self-efficacy scale. Research on Social Work Practice, 13, 724–741. doi:10.1177/1049731503254105

49.

Kranz

K. M.

O’Hare

(2006). The substance abuse treatment self-efficacy scale. Journal of Social Service Research, 32, 109–121. doi:10.1300/J079v32n03_06

50.

Kranz

K. M.

O’Hare

(2011). Revalidating the Substance Abuse Treatment Self-Efficacy Scale (SATSES): A replication with social work practitioners. Journal of Social Work Practice in the Addictions, 11, 270–285. doi:10.1080/1533256X.2011.595290

51.

Krumpal

(2013). Determinants of social desirability bias in sensitive surveys: A literature review. Quality & Quantity, 47, 2025–2047. doi:10.1007/s11135-011-9640-9

52.

Lawrence

Abel

E. M.

(2013). Comparing outcomes of a web-based MSW course to face-to-face outcomes: Implications for social work education. Social Work Education, 32, 762–772. doi:10.1080/02615479.2012.730140

53.

Logie

Bogo

Regehr

(2013). A critical appraisal of the use of standardized client simulations in social work education. Journal of Social Work Education, 49, 66–80. doi:10.1080/10437797.2013.755377

54.

Lynch

M. T.

Zhang

Korr

W. S.

(2009). Research training, institutional support, and self-efficacy: Their impact on research activity of social workers. Advances in Social Work, 10, 193–210.

55.

Madaus

G. F.

Stufflebeam

D. L.

(2000). Program evaluation: A historical overview. In Stufflebeam

D. L.

Madaus

G. F.

Kelleghan

(Eds.), Evaluation models: Viewpoints on educational and human services evaluation (2nd ed., pp. 3–18). Norwell, MA: Kluwer Academic Publishers.

56.

Mancoske

R. J.

Standifer

Cauley

(1994). The effectiveness of counseling services for battered women. Research on Social Work Practice, 4, 53–63.

57.

Maschi

Wells

Slater

G. Y.

MacMillan

Ristow

(2013). Social work students’ research-related anxiety and self-efficacy: Research instructors’ perceptions and teaching innovations. Social Work Education, 32, 800–817. doi:10.1080/02615479.2012.695343

58.

Maxwell

Scourfield

Holland

Featherstone

Lee

(2012). The benefits and challenges of training child protection social workers in father engagement. Child Abuse Review, 21, 299–310. doi:10.1002/car.2218

59.

Meyer-Adams

Dorsey

C. J.

Potts

M. K.

Resales

A. M.

Koob

J. J.

(2011). How to tackle the shift of educational assessment from learning outcomes to competencies: One program’s transition. Journal of Social Work Education, 47, 489–507. doi:10.5175/JSWE.2011.201000017

60.

Montcalm

D. M.

(1999). Applying Bandura’s theory of self-efficacy to the teaching of research. Journal of Teaching in Social Work, 19, 93–107. doi:10.1300/J067v19n01_08

61.

Morton

M. H.

Montgomery

(2013). Youth empowerment programs for improving adolescents’ self-efficacy and self-esteem: A systematic review. Research on Social Work Practice, 23, 22–33. doi:10.1177/1049731512459967

62.

Nunnally

J. C.

Bernstein

I. H.

(1994). Psychometric theory. New York, NY: McGraw-Hill.

63.

Parker

(2006). Developing perceptions of competence during practice learning. British Journal of Social Work, 36, 1017–1036. doi:10.1093/bjsw/bch347

64.

Parker

Hughes

Rutter

(2006–2007). ‘Shadowing’ and its place in preparing students for practice learning. Journal of Practice Teaching & Learning, 7, 51–70. doi:10.1921/19643

65.

Parrish

D. E.

Rubin

(2011). Validation of the evidence-based practice process assessment scale—short version. Research on Social Work Practice, 21, 200–211. doi:10.1177/1049731510389193

66.

Paulhus

D. L.

(1991). Measurement and control of response bias. In Robinson

J. P.

Shaver

P. R.

Wrightsman

L. S.

(Eds.), Measures of personality and social psychological attitudes (pp. 17–60). New York, NY: Academic Press.

67.

Payne

S. L.

(1951). The art of asking questions (13th ed.). Princeton, NJ: Princeton University Press.

68.

Petrovich

(2004). Using self-efficacy theory in social work teaching. Journal of Social Work Education, 40, 429–443. doi:10.1080/10437797.2004.10672298

69.

Pike

C. K.

Bennett

Chang

(2004). Evaluation of two interviewing skills measures: An instrument validation study. Advances in Social Work, 5, 61–75.

70.

Platt

(2011). Assessments of children and families: Learning and teaching the skills of analysis. Social Work Education, 30, 157–169. doi:10.1080/02615479.2011.540381

71.

Rawlings

M. A.

(2012). Assessing BSW student direct practice skills using standardized clients and self-efficacy theory. Journal of Social Work Education, 48, 553–576. doi:10.5175/JSWE.2012.201000070

72.

Rawlings

M. A.

Townsend

A. L.

Gingerich

W. J.

(2005). A psychometric analysis of the social work self-efficacy scale. Paper presented at the 2005 Society for Social Work Research Conference. Retrieved from http://sswr.confex.com/sswr/2005/techprogram/P1415.HTM

73.

Ravitch

(2000). Left back: A century of battles over school reform. Parsippany, NJ: Simon & Schuster.

74.

Rishel

C. W.

Majewski

(2009). Student gains in self-efficacy in an advanced MSW curriculum: A customized model for outcomes assessment. Journal of Social Work Education, 45, 365–383. doi:10.5175/JSWE.2009.200800101

75.

Rubin

Parrish

D. E.

(2011). Validation of the evidence-based practice process assessment scale. Research on Social Work Practice, 21, 106–118. doi:10.1177/1049731509347851

76.

Scourfield

Smail

Butler

(2014). Improving the engagement of fathers in child safeguarding. Child Abuse Review, 24, 77–153. doi:10.1002/car.2333

77.

Scourfield

Tolman

Maxwell

Holland

Bullock

Sloan

(2012). Results of a training course for social workers on engaging fathers in child protection. Children & Youth Services Review, 34, 1425–1432. doi:10.1016/j.childyouth.2012.03.022

78.

Spitzer

Holden

Cuzzi

L. C.

Rutter

Chernack

Rosenberg

(2001). Edith Abbott was right: Designing fieldwork experiences for contemporary health care practice. Journal of Social Work Education, 37, 79–90. doi:10.1080/10437797.2001.10779038

79.

Strand

V. C.

Abramovitz

Layne

C. M.

Robinson

Way

(2014). Meeting the critical need for trauma education in social work: A problem-based learning approach. Journal of Social Work Education, 50, 120–135. doi:10.1080/10437797.2014.856235

80.

Tower

L. E.

(2003). Domestic violence screening: Education and institutional support correlates. Journal of Social Work Education, 39, 479–494. doi:10.1080/10437797.2003.10779150

81.

Unrau

Y. A.

Beck

A. R.

(2004). Increasing research self-efficacy among students in professional academic programs. Innovative Higher Education, 28, 187–204.

82.

Unrau

Y. A.

Grinnell

R. M.

(2005). The impact of social work research courses on research self-efficacy for social work students. Social Work Education, 24, 639–651. doi:10.1080/02615470500185069

83.

van Dinther

Dochy

Segers

Baeken

(2014). Student perceptions of assessment and student self-efficacy in competence-based education. Educational Studies, 40, 330–351. doi:10.1080/03055698.2014.898577

84.

Voorhees

R. A.

(2001). Competency-based learning models: A necessary future. New Directions for Institutional Research, 110, 5–13.

85.

Westhues

Barsen

Freyond

Train

(2014). An outcome evaluation of a problem-based learning approach with MSW students. Journal of Social Work Education, 50, 472–489. doi:10.1080/10437797.2014.917897

86.

Wiechelt

S. A.

Ting

(2012). Field instructor’s perceptions of evidence-based practice in BSW field placement sites. Journal of Social Work Education, 48, 577–593. doi:10.5175/JSWE.2012.201000110

87.

Williams

N. R.

King

Koob

J. J.

(2002). Social work students go to camp: The effects of service learning on perceived self-efficacy. Journal of Teaching in Social Work, 22, 55–70. doi:10.1300/J067v22n03_05

88.

Woody

J. D.

Anderson

D. K.

D’Souza

H. J.

(2015). Dissemination of trauma-focused cognitive-behavioral therapy with community practitioners: Focus on self- efficacy. Journal of Evidence-Informed Social Work, 12, 323–335. doi:10.1080/15433714.2014.950128

89.

Woody

J. D.

Zeleny

M. G.

D’Souza

H. J.

Harder

Reiser

Szto

(2014). Student progress in a social work writing course: Self-efficacy, course objectives, and skills. Journal of Social Work Education, 50, 507–524. doi:10.1080/10437797.2014.91789

90.

York

R. O.

(2008). Comparing three modes of instruction in a graduate social work program Journal of Social Work Education, 44, 157–172.