An examination of school counselor involvement in school-family-community partnerships.

The study in this article investigated school counselor involvement in school-family-community partnerships and factors that influence such involvement. Participants were 235 members of the American School Counselor Association. Factor analyses of responses to the survey designed specifically for this study defined a set of factors that were used to examine variations in school counselor involvement in partnerships. Regression analyses revealed that (a) collaborative school climate, (b) school counselor role perceptions, (c) school counselor confidence in ability to build partnerships, and (d) school counselor attitudes about partnerships were significantly related to the counselors' involvement in school-family-community partnerships. Implications for school counselor training, practice, and research are discussed.

**********

School-family-community partnerships are collaborative relationships in which school professionals partner with family and community members and community-based organizations, including businesses, churches, libraries, and social service agencies, to implement programs and activities to help students succeed (Bryan, 2003, 2005; Epstein, 1995). Over the past decade, the participation of professional school counselors in the development and implementation of school-family-community partnerships has been endorsed and strongly encouraged in the literature (Atkinson & Juntunen, 1994; Bemak, 2000; Bryan, 2005; Colbert, 1996; Dedmond, 1991; Hobbs & Collison, 1995; Keys & Lockhart, 1999; Lapan, Osana, Tucker, & Kosciulek, 2002; Mitchell & Bryan, 2007; Taylor & Adelman, 2000; Walsh, Howard, & Buckley, 1999). These authors believe that school-family-community partnerships provide integral system support for the school counseling program and that school counselors, if involved in such partnerships, may better meet the personal/social, academic, and career needs of larger numbers of students. Furthermore, they believe that school counselors have the necessary skills (e.g., coordination, consultation, collaboration) to carry out tasks related to promoting, developing, and implementing school-family-community partnerships (Bemak; Bryan, 2005; Colbert; Keys, Bemak, Carpenter, & King-Sears, 1998; Mitchell & Bryan; Porter, Epp, & Bryant, 2000; Taylor & Adelman).

In addition to possessing the skills necessary to develop and implement school-family-community partnerships, the consensus among many school counseling professionals seems to be that school counselors already engage in roles that are consistent with facilitators and initiators of school-family-community partnerships (Atkinson & Juntunen, 1994; Bemak, 2000; Bryan, 2005; Bryan & Holcomb-McCoy, 2004, 2006; Taylor & Adelman, 2000). These include leadership, collaboration, advocacy, and liaison roles. Role is a term used to describe a set of behaviors expected of people as a result of their social or institutional position or profession (Katz & Kahn, 1966). Role behaviors are tasks that individuals carry out in response to their perceptions of their role. As a result of an in-depth examination of the school counseling literature (e.g., Atkinson & Juntunen; Lapan et al., 2002; Taylor & Adelman), we found that 18 partnership role behaviors are being promoted for school counselors. These include helping parents, family, and community members organize support programs for students, collaborating with community agency professionals, providing parent education workshops and seminars, collaborating with local businesses and industries, and conducting home visits to families, to name a few (see Table 1). It is not clear whether these 18 role behaviors are unique or whether they can be subsumed under more general roles (e.g., collaboration, leadership, advocacy, liaison), because considerable overlap exists in the way in which they are described in the literature.

There is no previous theoretical framework explaining what factors influence school counselors' involvement in school-family-community partnerships. Several studies have highlighted school level (i.e., elementary, middle, and high school) as a main variable influencing teacher, school psychologist, and school counselor involvement in school-family-community partnerships (Bryan & Holcomb-McCoy, 2004, 2006; Dauber & Epstein, 1993; Epstein & Dauber, 1991; Pelco & Ries, 1999; Pelco, Ries, Jacobson, & Melka, 2000). Consistently, school professionals in elementary school settings reported higher involvement in school-family-community partnerships than those in middle and high schools. Based on the limited amount of research that examines factors related to school-family-community partnership involvement, there is a need to explore factors that promote school counselor involvement in school-family-community partnership roles.

An examination of the empirical research in the school-family-community partnership, school-community collaboration, and parent involvement literature across education and psychology disciplines led to the identification of two types of factors that might affect school counselor involvement in partnerships: (a) school counselor factors, and (b) school-related factors. The five school counselor factors are role perceptions (Bryan & Holcomb-McCoy, 2004; Pelco et al., 2000); confidence in one's ability, skills, and knowledge to implement partnerships (Hiatt-Michael, 2001; Shartrand, Weiss, Kreider, & Lopez, 1997); commitment to advocacy (Mawhinney & Smrekar, 1996); attitudes about partnerships and families (Christenson & Sheridan, 2001; Dauber & Epstein, 1993); and perceived barriers (Bryan & Holcomb-McCoy; Sanders, 2001; Sanders & Harvey, 2002). The three school-related factors are school climate (Christenson & Sheridan; Sanders & Harvey), principal support (Mawhinney, 1994; Rutherford, Anderson, & Billig, 1995; Sanders & Harvey; Smylie, Crowson, Chou, & Levin, 1994), and school practices of partnerships (Christenson & Sheridan). In several previous studies, findings suggested that school-related factors would have a greater effect on involvement in school-family-community partnerships than individual-level factors, like the school counselor factors (Mawhinney & Smrekar; Smrekar & Mawhinney, 1999; Smylie et al.).

In spite of the aforementioned endorsement of partnership roles for school counselors, there is a paucity of research about school counselor involvement in these partnerships. A recent study examined school counselors' perceptions of the importance of their involvement in nine school-family-community partnership programs and eight partnership roles (Bryan & Holcomb-McCoy, 2004, 2006). Their results indicated that school counselors perceive their involvement in partnerships as important. At the same time, however, these authors suggested that there is a need for further research that explores what roles school counselors are playing in school-family-community partnerships and what factors are related to their involvement in such partnerships. The suggestion that school counselors should be involved in leadership, advocacy, collaboration, and liaison roles in partnerships among schools, families, and communities does not guarantee that school counselors will routinely enact such partnership roles.

A study of the factors that influence school counselor involvement in partnerships will help identify ways to foster school counselors' involvement in these partnerships. Indeed, understanding such factors is a precursor to asking how can practicing school counselors, other school professionals, counselor educators, and district, state, and federal education policymakers facilitate school counselor involvement in such partnership roles. Furthermore, an understanding of the factors will help to facilitate the development of training strategies and practices designed to increase such involvement.

PURPOSE OF THE STUDY

With that said, the primary purpose of this study was to examine what factors are related to school counselor perceived involvement in school-family-community partnership roles. This study also sought to shed light on whether specific roles (e.g., collaboration, leadership, advocacy, liaison) can serve as broad descriptors for the 18 partnership role behaviors being prescribed for school counselors. The primary research question guiding the study was this: What are the relationships of school counselor factors, school-related factors, and school level to school counselors' involvement in school-family-community partnership roles?

We hypothesized that school level would be related to school counselor involvement in partnership roles with elementary school counselors having a significantly higher level of involvement relative to middle and high school. We also hypothesized that the school-related factors (i.e., school climate, principal support, and school practices of partnerships) would be related to involvement after controlling for school level. Additionally, we hypothesized that after controlling for school level, the school counselor factors (i.e., role perceptions, confidence in ability to build partnerships, commitment to advocacy, attitudes about partnerships and families) would be positively related to involvement with the exception of perceived barriers, which would be negatively related to involvement. Finally, we hypothesized that the school counselor factors would contribute to explaining school counselor involvement in school-family-community partnership roles above and beyond school-related factors. Prior to answering the primary research question, we needed to determine whether there was a single dimension or multiple dimensions underlying the survey items measuring school counselor involvement in school-family-community partnership roles. It was equally necessary to determine whether there was a single dimension or multiple dimensions underlying the items measuring the hypothesized school counselor and school-related factors.

METHOD

Participants

The 235 participants in this study were primarily female (84.3%), and White (93.2%). This sample is representative of the American School Counselor Association (ASCA), which has a membership consisting of a majority of White females (ASCA, personal communication, July 17, 2006). In addition, 26.4% of the participants worked in elementary schools, 24.7% in middle schools, and 31.5% in high schools. The remaining 17.4% worked in combination schools (i.e., K-8 and 6-12 schools). Most schools were in suburban areas (45.5%), 34% in rural areas, and 20% in urban areas. A significant proportion of the participants had 6 to 10 years of school counseling experience (42.6%), while 19.6% had between 1 and 5 years, 14% had 11 to 15 years, 8.5% had 16 to 20 years, and 15.3% had over 20 years of school counseling experience.

Procedures

A packet containing the School Counselor in Partnerships Survey (SCIPS), a cover letter, a self-addressed stamped envelope, and an announcement of a drawing for prizes to be won by two participants was mailed to each prospective participant, randomly selected by ASCA from its membership database. The incentive was offered in the hope of increasing the response rate, which is usually about 30% in survey research (Alreck & Settle, 1995). A definition of school-family-community partnerships was provided along with directions for completion of the survey. Participants were informed of the anonymous and volunteer nature of the study and that returning the completed survey indicated their informed consent. Follow-up e-mails were sent to the participants on two occasions as reminders to return the survey: at 1 week and 2 weeks after the initial mailing. In addition, a second survey packet was mailed out to non-respondents 3 weeks after the initial mailing. A total of 271 surveys were returned. Of these, 235 surveys were usable (37% response rate); the remaining 36 were returned because there was no counselor at the school, the counselor declined the offer to participate, or the participant was no longer working at the school.

The School Counselor in Partnerships Survey

The SCIPS was designed for this study to measure school counselors' involvement in 18 partnership role behaviors. In addition, the SCIPS was designed to measure the school-related and school counselor factors that were believed to influence school counselor involvement in partnership role behaviors. The survey items were generated on the basis of a review of empirical studies in the school-family-community partnership, collaboration, and parent involvement literature. The SCIPS was piloted on 10 second-year master's level interns and on 10 professional school counselors to obtain feedback on word clarity, comprehension, format, and length of the survey. On the basis of the feedback, a few items were reworded and changed. The survey then was administered to 12 doctoral students from counselor education, college student personnel, and higher education programs to further assess clarity, readability of items, and time taken to complete the survey. The survey was simultaneously presented to seven experts based on their research experience on parent involvement and school-family-community partnerships. The experts provided feedback on content validity as well as the clarity and appropriateness of items. After these extensive reviews of the survey, no further changes or revisions were suggested.

The SCIPS consists of five sections. The first section contains a definition of school-family-community partnerships and includes items pertaining to demographic characteristics of the participants (e.g., gender, years of experience, race/ethnicity, and school setting). The second section consists of four items that ask participants to indicate their involvement in school-family-community partnerships, their involvement in advocacy, the extent they believe partnerships to be a part of their role, and the amount of training they have had related to partnerships on a 3-point Likert scale (1 = not at all, to 3 = very much). Responses to these four items were not analyzed for the purpose of this study.

The third section comprises the 18 school-family-community partnership role behaviors (Table 1). For each role behavior, participants used a 5-point Likert scale (1 = never, 2 = seldom, 3 = sometimes, 4 = frequently, 5 = always) to indicate (a) their perceived level of involvement in each partnership role behavior, and (b) their perceptions about the extent that they should perform each partnership role behavior as part of their school counseling role (i.e., their role perceptions).

The fourth section consists of items that assess the school-related factors (i.e., school climate, principal support, and school practices of partnerships) and the remaining four school counselor factors (i.e., confidence in ability to build partnerships, commitment to advocacy, attitudes about partnerships and families, and perceived barriers). Examples of items used to measure the school-related factors include the following:

* This school has a friendly, warm atmosphere.

* The principal encourages counselor participation in planning partnerships.

* Parents and family members play many different roles in this school.

* Partnerships with family and community members are valued in this school.

Examples of items used to measure the school counselor factors include the following:

* Partnerships with families provide support for the counseling program.

* I have received the training to implement school-family-community partnerships.

* I would actively advocate for children even if it were not considered part of my role.

* I do not have time to get involved in partnerships.

Participants indicated their agreement with each of the items using a 5-point Likert scale (1 = strongly disagree, 2 = disagree, 3 = agree somewhat, 4 = agree, 5 = strongly agree).

The final section of the survey consists of four open-ended questions addressing the kinds of school-family-community partnerships in which participants are involved, the things that encouraged them to get involved in partnerships, the obstacles that hindered their involvement in partnerships, and the strategies they used to overcome barriers to their involvement in partnerships. These open-ended questions provided qualitative data that were not analyzed for the purpose of this study. See Bryan (2003) for the entire survey.

RESULTS

Exploratory Principal Components Analyses

Prior to conducting the statistical analyses related to the primary research question, we conducted four principal components analyses (PCAs) with varimax rotation to determine the factor structure of participants' responses to each set of items measuring (a) perceived involvement in the 18 partnership role behaviors; (b) role perceptions, that is, participants' perceptions about the extent that they should perform each role behavior as part of their school counseling role; (c) the remaining school counselor factors; and (d) the school-related factors. The goal of PCA is to identify a set of uncorrelated factors (also called components) that are smaller in number than the original set of items. The factors are derived so that they explain nearly all of the total variance in the original items (Pett, Lackey, & Sullivan, 2003; Tabachnick & Fidell, 2001).

The Kaiser-Meyer-Olkin measure of sampling adequacy was greater than .80 and the Bartlett's test of sphericity was nonsignificant for each of the sets of items to be factor analyzed. This indicated that each set of items was an excellent candidate for PCA. The three criteria used to determine the number of factors to retain were Kaiser's criterion (eigenvalues over 1.0), Catell's scree test, and the interpretability of the factor solution (Pert et al., 2003). For each scale (i.e., set of items that load on a factor), Cronbach alpha coefficients were calculated to determine the internal consistency of the scale and to provide final information about items that should be excluded from the scale. Items that increased the reliability of the scales if deleted were excluded from the final scales. Final scales consisted of items that loaded strongly (r [greater than or equal to] .40) on the factor. These final scales were used to calculate factor scores for each factor. Factor scores are linear combinations of the items that load on a factor so that each respondent has a score on the factor of interest. Variables that are created from factor scores are continuous standardized scores with a mean of zero and can be used to represent the factors in subsequent analyses (Hair, Anderson, Tatham, & Black, 1998).

The following factors or components resulted from the PCAs. Their factor scores then were used as the dependent and independent variables in a hierarchical regression analysis to test the research hypotheses. Below is a review of the PCA.

Involvement in school-family-community partnership roles. After the eigenvalues over 1.0 are examined, the involvement items appeared to consist of four factors. However, the scree test suggested one factor. All items had loadings above .43 on the first factor. While one or two items loaded on the other factors, their loadings were all stronger on the first factor so that only the first factor was clearly defined. Hence, a single factor was retained with loadings ranging from .44 to .78. The factor accounted for 39.1% of the item variance and had a coefficient alpha of .90. Factor scores were calculated to create the dependent variable, perceived involvement in school-family-community partnership roles.

Role perceptions. Initially, a four-factor solution was examined because four factors had eigenvalues over 1.0. However, the scree test suggested that there was a single factor. All items had loadings above .50 on the first factor. Only one or two items loaded on the other components, but these items had higher loadings on the first factor. Hence, a single factor was retained with loadings ranging from .50 to .75. The factor accounted for 38.9% of the variance in the items and had a reliability coefficient of .90. Factor scores were calculated to create a composite variable, role perceptions, which was used as an independent variable in subsequent hierarchical regression.

Other school counselor factors. The Kaiser criterion suggested a nine-factor solution. The scree test was not conclusive, suggesting either a four- or five-factor solution. A four-factor solution and a five-factor solution were inspected with the intent of maintaining conceptual meaningfulness. We retained a five-factor solution because it was the most interpretable. These five factors accounted for 52.9% of the item variance. We labeled the five factors: (a) confidence in ability to build partnerships (9.5% variance), consisting of seven items with loadings ranging from .59 to .73 and a coefficient alpha of .84; (b) commitment to advocacy (8.9% variance), consisting of seven items with loadings ranging from .49 to .72 and a coefficient alpha of .77 after one item was deleted; (c) attitudes about partnerships (16.5% variance), consisting of nine items with loadings ranging from .70 to .84 and a coefficient alpha of .93 after two items were deleted; (d) attitudes about families and communities (7.5% variance), consisting of six items with loadings ranging from .48 to .65 with a coefficient alpha of .76 after one item was deleted; and (e) perceived barriers (10.5% variance), consisting of eight items with loadings from .51 to .78, with a coefficient alpha of .82. Factor scores were calculated in order to create five composite variables that were used as independent variables in the subsequent hierarchical regression. Sample items for each factor and their loadings are found in Table 2.

School-related factors. The Kaiser criterion suggested a seven-factor solution while the screen test suggested a single factor. Two-factor and three-factor solutions were inspected. However, these factor solutions were poorly defined with most of the items loading fairly strongly on the first factor and weakly on the other factors (r < .40), suggesting one unique dimension. Hence, we retained one factor that we labeled collaborative school climate. To assess the internal consistency of the items, six items were reversed so that high scores on the items meant high levels of collaborative school climate. Seven items were deleted as a result of their low item-total correlations, which ranged from .12 to .40. After these items were deleted, the final scale had a coefficient alpha of .96. Factor loadings ranged from .40 to .80 and the factor accounted for 47% of the item variance in the remaining 30 items. Factor scores produced a composite variable, collaborative school climate, which was used as an independent variable in the subsequent hierarchical regression. Sample items and their loadings for this factor are included in Table 2. See Bryan (2003) for a full report of the survey items and their loadings.

Descriptive Analysis

The means and standard deviations (based on the original Likert scale) for perceived level of involvement in each of the 18 partnership role behaviors are shown in Table 1. On average, school counselors reported being more involved in some partnership role behaviors than in others. For example, they reported frequent involvement in locating services and resources for students and their families in the community (M = 4.01, SD = .95), collaborating with community agency professionals (M = 3.95, SD = .97), and working with a team of school staff, family, and/or community members and professionals (M = 4.02, SD = 1.07). They reported that they were seldom involved in conducting home visits to families (M = 1.98, SD = 1.07) and training the staff on how to build effective school-family-community partnerships (M = 2.31, SD = 1.09).

On average, school counselors reported being somewhat involved in all other partnership role behaviors. Considerable proportions of the 235 participants reported being never or seldom involved in conducting home visits to families (65%), collaborating with local businesses and industries (53.9%), implementing programs to help school staff understand families and the community (46.6%), providing parent education workshops and seminars (37.7%), and implementing programs to help family and community members understand the school (29.6%).

Hierarchical Regression Analysis

Hierarchical (or blocked) regression analysis was used to answer the primary research question: What are the relationships of school counselor factors, school-related factors, and school level to school counselors' involvement in school-family-community partnership roles? In hierarchical regression analysis, predictors are entered in blocks with the "sole purpose of noting whether blocks entered at later stages add meaningfully to the prediction of the criterion" (Pedhazur, 1997, p. 229). The dependent variable, perceived involvement in school-family-community partnership roles, was created from factor scores and was a continuous standardized variable with a mean of zero. The entry of the independent variables into the hierarchical regression model was guided by a priori research that suggested that school-related factors would have a greater effect than individual-level factors on school professionals' involvement in school-family-community partnerships (Mawhinney, 1994; Mawhinney & Smrekar, 1996; Smrekar & Mawhinney, 1999; Smylie et al., 1994) and that school level would be related to involvement in partnerships (Bryan & Holcomb-McCoy, 2004, 2006; Dauber & Epstein, 1993; Epstein & Dauber, 1991; Pelco & Ries, 1999; Pelco et al., 2000).

School level was entered as a control variable in the first block of the model, collaborative school climate in the second block, and the school counselor variables (i.e., role perceptions, confidence in ability to build partnerships, attitudes about partnerships, attitudes about families and communities, and perceived barriers) were entered as a set in the final block. Elementary, middle, and combination schools were entered as three dummy variables while high school was the reference category. Collaborative school climate and the school counselor variables were all created from factor scores and were continuous standardized variables with a mean of zero.

First, Pearson's correlations were calculated to determine the intercorrelations among the variables. Correlations between perceived involvement and the independent variables ranged from -.30 to .71. All correlations were significant except for the correlation between attitudes about families and confidence in ability to build partnerships. Role perceptions had the highest correlation to perceived involvement, r (235) = .71, p < .005. Correlations among the independent variables ranged from -.39 to .53. The highest correlation was between role perceptions and attitudes about partnerships, r (229) = .53, p = .001. In spite of moderate correlations among some of the independent variables, they were not high enough to suggest a threat of multicollinearity (Tabachnick & Fidell, 2001).

Table 3 shows the summary of the hierarchical regression analysis for variables predicting school counselor perceived involvement in school-family-community partnerships. School level accounted for a significant amount of the variability in school counselor involvement in school-family-community partnerships, [R.sup.2] = .04, F (3, 205) = 3.01, p < .05. More specifically, school counselors in elementary and combination schools had significantly higher involvement in school-family-community partnerships than did counselors in high schools. After school level was controlled for, collaborative school climate accounted for a significant proportion of the variability in school counselor perceived involvement in school-family-community partnerships, [DELTA][R.sup.2] = .13, F (1, 204) = 32.84, p < .001. Elementary schools were no longer a significant predictor of school counselor involvement in partnerships after collaborative school climate was taken into account. However, combination schools continued to be a significant predictor of involvement.

After school level and collaborative school climate were accounted for, school counselor variables accounted for a significant proportion of the variability in school counselor involvement in school-family-community partnerships, [DELTA][R.sup.2] = .47, F (6, 198) = 43.32, p < .001. In particular, school counselor role perceptions, confidence in ability to build partnerships, and attitudes about partnerships were all significantly related to perceived involvement above and beyond collaborative school climate. Once collaborative school climate and the school counselor variables were taken into consideration, school level no longer explained any variability in school counselor perceived involvement in school-family-community partnerships. Role perceptions had the largest positive relationship to involvement, followed by collaborative school climate and confidence in ability to build partnerships. Attitude about partnerships had a negative relationship to school counselor involvement in partnerships, [beta] = - .14, t = -2.44, p = .016. This result was surprising because a high score on this attitudinal variable indicated high positive attitudes about partnerships. Perceived barriers and school level were not significant predictors of school counselor involvement in partnerships.

Post Hoc Analysis

A post hoc hierarchical regression analysis with four steps was conducted in an attempt to explain the significant negative relationship between attitudes to partnerships and school counselor involvement in school-family-community partnerships (Table 3). The decision to enter role perceptions in the last step of the model following entry of the other school counselor variables was made due to its significant moderate correlation with attitudes about partnerships and its strong correlation to the dependent variable. This analysis revealed that, after school level and collaborative school climate were taken into account, confidence in ability, commitment to advocacy, and attitudes about families were all significantly related to school counselor perceived involvement in school-family-community partnerships (Table 3). Only after role perceptions was entered in the final step of the model was attitudes about partnerships significantly and negatively related to perceived involvement; at this point, commitment to advocacy and attitudes about families were no longer significantly related to involvement. Confidence in ability to build partnerships and collaborative school climate continued to be significantly positively related to involvement at the final step of the model.

The confounding negative relationship of attitudes about partnerships to involvement, as well as the changes in the relationships of commitment to advocacy and attitudes about families to involvement after role perceptions entered the analysis, may be due to common method variance. Common method variance is variance that results due to the measurement method used. Some of the sources of common method variance are common rater effects, item characteristic effects, item context effects, and measurement context effects (Podsakoff, Mackenzie, Lee, & Podsakoff, 2003). The common method variance in this current study may be due to the use of the same items and scale format for measuring both the dependent variable and role perceptions.

DISCUSSION

The purpose of this study was to examine what factors are related to school counselors' perceived involvement in school-family-community partnership roles. This study also explored the factor structure of the SCIPS. This research has several implications for school counselor training, research, and practice. Prior to discussing the relationships of the factors to school counselors' perceived involvement in partnership roles and how these results can inform counselor training and practice, we discuss the results of the exploratory principal components analyses.

The SCIPS appears to be a promising means of assessing school counselors' perceived involvement in school-family-community partnerships. Components analyses of participants' ratings on the survey suggest that there is one factor underlying the measures of perceived involvement, one factor (i.e., collaborative school climate) underlying the items purported to measure the school-related factors, one factor underlying the measures of role perceptions, and five factors (i.e., confidence in ability to build partnerships, commitment to advocacy, perceived barriers, attitudes about partnerships, attitudes about families and communities) underlying the items measuring the other school counselor factors.

Regarding partnership involvement, because no separate roles emerged, it is possible that partnership roles may overlap; for example, locating or brokering community resources and services for students may, at the same time, be considered an advocacy, collaboration, and liaison role behavior. However, it is also possible that the items and scale format used to measure involvement may not accurately capture each of the partnership roles discussed in the literature. Therefore, further effort should be given to defining the sets of role behaviors that may comprise advocacy, collaboration, leadership, and liaison roles related to school-family-community partnerships and to understanding the nature of these partnership roles.

Conversely, the overlap of the hypothesized school climate, principal support, and school practices of partnerships themes on the collaborative school climate factor seems to be supported by the school climate literature (Griffith, 2001; Hoy, 1990; Hoy, Smith, & Sweetland, 2002). School climate is % general concept that captures the atmosphere of a school; it is experienced by teachers and administrators, describes their collective perceptions of routine behavior, and affects their attitudes and behaviors in the school" (Hoy et al., p. 38). Given the nature of school climate, it is not surprising that these themes might be captured by a single factor. For example, principal characteristics, such as principal influence and principal leadership, have been found to be components of school climate because principals play a major part in developing and maintaining a welcoming and collaborative school climate (Griffith; Hoy et al.). This may account for the overlap between principal support and school climate in this study. Replication of this study would help to clarify these findings.

In contrast to the partnership involvement and school-related items, this study presents evidence for the multidimensionality of the items measuring the school counselor factors. As expected, all of the hypothesized school counselor factors emerged, with the attitudes factor emerging as two separate factors, attitudes about partnerships and attitudes about families and communities. These results are promising in that they lend support to the framework used in this study to explain why school counselors may get involved in partnership roles.

In general, the results indicate that, on average, school counselors perceive themselves as being somewhat involved in most of the 18 school-family-community partnership role behaviors. This is encouraging given the importance attributed to school counselor involvement in partnerships (Bemak, 2000; Colbert, 1996; Keys & Lockhart, 1999; Lapan et al., 2002; Taylor & Adelman, 2000; Walsh et al., 1999). This finding is also consistent with previous findings that school counselors perceive their involvement in school-family-community partnerships to be important (Bryan & Holcomb-McCoy, 2004). Nevertheless, the proportion of school counselors who reported being never or seldom involved in some of the partnership behaviors suggests that there still may be a considerable number of school counselors who are not involved in some potentially beneficial system support activities, such as conducting parent workshops and collaborating with businesses.

Regarding the relationships between school counselors' perceived involvement in school-family-community partnerships and the school-related and school counselor factors, collaborative school climate was significantly positively related to school counselors' perceived involvement after controlling for school level. In addition, school counselor role perceptions, confidence in their ability to build partnerships, and attitudes about partnerships were Significantly related to school counselors' perceived involvement in partnerships above and beyond collaborative school climate and school level. School level was not related to perceived involvement after taking into account the school-related and school counselor variables. Hence, this study's findings may help to clarify the relationship between school level and school professionals' partnership involvement found in previous studies (Dauber & Epstein, 1993; Epstein & Dauber, 1991; Pelco & Ries, 1999; Pelco et al., 2000).

The relationship of collaborative school climate to school counselors' perceived involvement in partnerships corroborates findings that institutional variables are key influences in school professionals' efforts to create and sustain school-family-community collaboration (Mawhinney, 1994; Mawhinney & Smrekar, 1996; Sanders & Harvey, 2002; Smrekar & Mawhinney, 1999; Smylie et al., 1994). However, findings support the view that individual-level variables such as role perceptions, confidence in ability to build partnerships, and attitudes about partnerships also provide impetus for school professionals' involvement in partnerships (Mawhinney & Smrekar).

One important finding from this study was the significant positive relationship between school counselors' perceived involvement in partnerships and their role perceptions. This corroborates previous research by Pelco and Ries (1999) and Bryan and Holcomb-McCoy (2004) that found that school professionals are more likely to be involved in school-family-community partnerships when they perceive their roles to be congruent with partnership activities. However, one must interpret these results with caution because of the possible confounding effect of role perceptions due to possible bias from common method variance.

Another compelling finding of this study was the fact that school counselors' confidence in ability to collaborate was consistently positively related to their perceived involvement in partnerships in both the initial and post hoc analyses. This finding supports previous research suggesting that there is a positive relationship between teachers' perceived involvement in partnerships and their perceptions that they were equipped with the skills and knowledge for building partnerships (Hiatt-Michael, 2001; Pelco & Ries, 1999; Pelco et al., 2000; Shartrand et al., 1997). It is possible that in this study, the scale that measures confidence in ability to build partnerships may be capturing school counselors' self-efficacy beliefs about partnerships, that is, school counselors' beliefs that they are effective in building partnerships and that they have the skills and competencies to build such partnerships. However, further exploration of the convergent validity between the confidence-in-ability scale in the SCIPS and instruments that measure self-efficacy is necessary before such conclusions can be drawn.

LIMITATIONS AND IMPLICATIONS

The self-report nature of the study and the selection of participants from one professional association limit the generalizability of the study. The study's response rate of 37%, although typical in survey research, is a further limitation. One cannot assume that participants who did not respond to the SCIPS survey would have responded in the same way as those who did respond. Finally, as mentioned earlier, common method variance is a limitation of this study. Nevertheless, this study creates a basis for further exploration of the factors that are related to school counselor involvement in school-family-community partnerships. It is the first step in building a model for explaining school counselor involvement in school-family-community partnerships.

Results from this study have several implications for training and practice in school counseling. One such implication is that school counselor education programs examine the content and activities in their curricula and practica that promote the partnership role behaviors discussed in this article. Previous research suggests that school counseling students' role perceptions, confidence in ability to build partnerships or self-efficacy beliefs about partnerships, and attitudes about partnerships could all be influenced by intentional curricula and practica (Garcia, 2004; Hoover-Dempsey, Walker, Jones, & Reed, 2002). In order to promote and increase school counselor involvement in school-family-community partnerships, it will be necessary to intentionally train school counselors to develop effective partnerships. Intentional training should include involving school counseling students on interdisciplinary teams to conduct school-family-community collaboration experiences and projects. In order to facilitate this interdisciplinary collaboration, counselor education programs will have to include coursework that allows school counseling students to work collaboratively in the classroom with students from family counseling, community counseling, educational administration, teacher education, school nurse education, school psychology, school social work, and other school-related majors. This would be most feasible in an interdisciplinary course on school-family-community consultation and collaboration.

Furthermore, in order to train school counselors for school, family, and community collaboration, it will be necessary for counselor educators to go beyond traditional modes of instruction to engage students in experiential activities and field experiences that involve them in partnerships (Kirschenbaum, 2001). Accordingly, school counseling trainees should be intentionally placed in internships where practicing school counselors already engage in various types of school-family-community partnerships.

Another implication is that school counselors examine how they impact the school climate through their roles as advocates, collaborators, leaders, and liaisons in partnerships with school, family, and community stakeholders. Practicing school counselors can be leaders in shaping and sculpting the climate, norms, and culture of a school. Previous research indicated that school counselors help to create a collaborative school climate when they make efforts to partner with school, family, and community stakeholders (Littrell, Peterson, & Sunde, 2001; Mawhinney & Smrekar, 1996; Peterson & Deal, 1998). As school counselors collaborate with other school professionals, families, and community members and organizations to advocate for systems change, implement systemic interventions, and remove systemic barriers for students, they communicate core values around collaboration. This action helps to bring a shift in the way schools, families, and communities relate and partner, thereby helping to create a more collaborative school climate.

In order to be leaders in helping to shape a collaborative school climate, school counselors must understand the dynamics of collaboration, school climate, and organizational change. They must possess strong collaboration skills and skills in assessing the school's climate and culture and the impact of their own practice on that climate and culture. Therefore, it is important that counselor educators provide school counseling students with rich experiences and curricula that foster these competencies as well as the vision, identity, values, and core beliefs that are synchronized with school-family-community collaboration and that empower school counselors to think and act systemically.

In addition to preservice training and preparation, professional development opportunities that focus on the role of practicing school counselors in building partnerships could be beneficial in fostering school counselors' role perceptions, confidence or self-efficacy, and attitudes about partnerships. Professional development should include time for practicing school counselors to hear and share how they can overcome school-related barriers to building partnerships. Likewise, professional development should be implemented with other school personnel (e.g., administrators, psychologists, teachers) in order to educate them on the role school counselors can play in developing partnerships with families and communities. This is important because some school personnel may not embrace the importance of partnerships to student success.

Given that school counselors consider their involvement in school-family-community partnerships to be important (Bryan & Holcomb-McCoy, 2004, 2006), they must provide strong advocacy for the time and resources to implement such professional roles. Such advocacy would decrease the likelihood of school counselors being burdened with responsibilities outside of their role perceptions. To help advocate for school-family-community partnerships, school counselors should use the student data (e.g., achievement, behavior-related, students' needs) that come across their desks on a daily basis and the success stories about the effects of partnership programs.

The limitations of this study pave the way for future research. Future research should focus on refining the SCIPS, determining its construct validity and reliability, and clarifying the relationships between the school-related and school counselor variables and involvement in partnerships. Such research should control for common method variance by improving scale items and using different items and scale formats for the predictor and criterion variables (e.g., role perceptions). In addition, it would be important to replicate this study with a participant sample drawn from a nationally representative sample of school counselors rather than only those who belong to professional organizations in order to determine if the factor structure and relationships hold across different samples of school counselors.

In addition to research focusing on the SCIPS, future research could address a variety of related questions: How do school counselors develop the role perceptions, confidence or self-efficacy, and attitudes to build partnerships? Which school-family-community partnership programs work best in enhancing student achievement and provide the most support to students and to school counseling programs? What strategies work best in building effective partnerships with various target populations (e.g., immigrant families, low-income or wealthy families) and in various settings (e.g., rural communities, inner-city communities)? Do school-family-community partnerships differ among school counselors in urban, rural, and suburban settings? How do school counselors overcome barriers to partnership building? It is important that counselor educators and researchers collaborate with school counselors to implement and study school-family-community partnerships. Such research will advance the understanding of partnerships in the school counseling context in much the same way that it is being done in the early childhood education, special education, general education, and school psychology fields.

CONCLUSION

School-family-community partnerships provide valuable system support services that help school counselors bridge cultural gaps among schools, families, and communities, build educational resilience in children, and empower families. Given the value of partnerships to students and the school counseling program, it is important that further attention be given to understanding and facilitating school counselors' capacity to build these partnerships. Therefore, counselor educators, school counseling professionals, school leaders, and policymakers face a number of tasks that go beyond simply recommending that school counselors build school-family-community partnerships.

One such task is for federal, district, and school-level policymakers and administrators to provide the support for school counselors to initiate and help implement school-family-community partnerships. Another task is for practicing school counselors and counseling researchers to implement collaborative field research to study what partnerships school counselors are building, why and how they build them, and the outcomes of such partnerships. Finally, there is a need for counselor educators to develop and test training curricula and strategies for equipping school counselors with the role perceptions, confidence in their ability to build partnerships, and attitudes necessary for facilitating their partnership involvement and knowledge for impacting school climate.

References

Alreck, P. L., & Settle, R. B. (1995). The survey research handbook (2nd ed.). Chicago: Irwin.

Atkinson, D. A., & Juntunen, C. L. (1994). School counselors and school psychologists as school-home-community liaisons in ethnically diverse schools. In P. Pedersen & J. C. Carey (Eds.), Multicultural counseling in schools: A practical handbook (pp. 103-119). Needham Heights, MA: Allyn & Bacon.

Bemak, F. (2000).Transforming the role of the counselor to provide leadership in educational reform through collaboration. Professional School Counseling, 3, 323-331.

Bryan, J. (2003). An investigation of school counselor involvement in school-family-community partnerships: Exploring the gap between current and prescribed practice. Dissertation Abstracts International, 64 (12), 4367A. (UMI No. AAT-3115701)

Bryan, J. (2005). Fostering educational resilience and achievement in urban schools through school-family-community partnerships: School counselors' roles. Professional School Counseling, 8, 219-227.

Bryan, J. A., & Holcomb-McCoy, C. H. (2004). School counselors' perceptions of their involvement in school-family-community partnerships. Professional School Counseling, 7, 162-171.

Bryan, J. A., & Holcomb-McCoy, C. H. (2006). School counselors' training and involvement in school-family-community partnership roles: An exploratory study. Journal of School Counseling, 4(13). Available online at http://www.jsc. montana.edu/pages/articles.html

Christenson, S. L., & Sheridan, S. M. (2001). Schools and families: Creating essential connections for learning. New York: Guilford Press.

Colbert, R. D. (1996).The counselor's role in advancing school and family partnerships. The School Counselor, 44, 100-104.

Dauber, S. L., & Epstein, J. L. (1993). Parents' attitudes and practices of involvement in inner-city elementary and middle schools. In N. Chavkin (Ed.), Families and schools in a pluralistic society (pp. 53-71). Albany, NY: State University of New York Press.

Dedmond, R. M. (1991). Establishing, coordinating school-community partnerships. NASSP Bulletin, 75(534), 28-35.

Epstein, J. L. (1995). School/family/community partnerships: Caring for the children we share. Phi Delta Kappan, 76, 701-712.

Epstein, J. L., & Dauber, S. L. (1991). School programs and teacher practices of parent involvement in inner-city elementary and middle schools. Elementary School Journal, 91, 289-305.

Garcia, D. C. (2004). Exploring connections between the constructs of teacher efficacy and family involvement practices: Implications for urban teacher preparation. Urban Education, 39, 290-315.

Griffith, J. (2001). Principal leadership of parent involvement. Journal of Educational Administration, 39, 162-186.

Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate data analysis (5th ed.). Upper Saddle River, NJ: Prentice Hall.

Hiatt-Michael, D. (2001). Preparing teachers to work with parents: ERIC Digest. Washington, DC: ERIC Clearinghouse on Teaching and Teacher Education. (ERIC Document Reproduction Service No. ED460123)

Hobbs, B. B., & Collison, B. B. (1995). School-community collaboration: Implications for the school counselor. The School Counselor, 43, 58-65.

Hoover-Dempsey, K.V., Walker, J. M., Jones, K. P., & Reed, R. P. (2002). Teachers Involving Parents (TIP): An in-service teacher education program for enhancing parental involvement. Teaching and Teacher Education, 18, 843-867.

Hoy, W. K. (1990). Organizational climate and culture: A conceptual analysis of the school workplace. Journal of Educational and Psychological Consultation, 1, 149-168.

Hoy, W. K., Smith, P. A., & Sweetland, S. R. (2002). The development of the Organizational Climate Index for high schools: Its measure and relationship to faculty trust. High School Journal, 86, 38-49.

Katz, D., & Kahn, R. L. (1966). The social psychology of organizations. New York: Wiley and Sons.

Keys, S., Bemak, F., Carpenter, S. L., & King-Sears, M. E. (1998). Collaborative consultant: A new role for counselors serving at-risk youths. Journal of Counseling and Development, 76, 123-134.

Keys, S. G., & Lockhart, E. J. (1999). The school counselor's role in facilitating multisystemic change. Professional School Counseling, 3, 101-107.

Kirschenbaum, H. (2001). Educating professionals for school, family and community partnerships. In D. Hiatt-Michael (Ed.), Promising practices in school, family and community partnerships (pp. 185-208). Greenwich, CT: Information Age Publishers.

Lapan, R. T., Osana, H. P., Tucker, B., & Kosciulek, J. F. (2002). Challenges for creating community career partnerships: Perspectives from practitioners. Career Development Quarterly, 51, 172-190.

Littrell, J. M., Peterson, J. M., & Sunde, J. (2001).Transforming the school culture: A model based on an exemplary counselor. Professional School Counseling, 4, 310-319.

Mawhinney, H. B. (1994). Institutional effects of strategic efforts at community enrichment. Educational Administration Quarterly, 30, 324-341.

Mawhinney, H. B., & Smrekar, C. (1996). Institutional constraints to advocacy in collaborative services. Educational Policy, 10, 480-501.

Mitchell, N. A., & Bryan, J. A. (2007). School-family-community partnerships: Strategies for school counselors working with Caribbean immigrant families. Professional School Counseling, 10, 399-409.

Pedhazur, E. J. (1997). Multiple regression in behavioral research: Explanation and prediction (3rd ed.). Orlando, FL: Harcourt Brace College Publishers.

Pelco, L. E., & Ries, R. R. (1999).Teachers' attitudes and behaviors toward family-school partnerships: What school psychologists need to know. School Psychology International, 20, 265-277.

Pelco, L. E., Ries, R. R., Jacobson, L., & Melka, S. (2000). Perspectives and practices in family-school partnerships: A national survey of school psychologists. School Psychology Review, 29, 235-250.

Peterson, K. D., & Deal, T. E. (1998). How leaders influence the culture of schools. Educational Leadership, 56, 28-30.

Pett, M. A., Lackey, N. R., & Sullivan, J. J. (2003). Making sense of factor analysis: The use of factor analysis for instrument development in health care research. Thousand Oaks, CA: Sage.

Podsakoff, P. M., Mackenzie, S. B., Lee, J., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879-903.

Porter, G., Epp, L., & Bryant, S. (2000). Collaboration among school mental health professionals: A necessity, not a luxury. Professional School Counseling, 3, 315-322.

Rutherford, B., Anderson, B., & Billig, S. (1995). Studies in education reform: Parent and community involvement in education. Final technical report, U.S. Department of Education. Denver, CO: RMC Research Corp. Retrieved June 16, 2003, from http://www.ed.gov/pubs/SER/ ParentComm/

Sanders, M. (2001). The role of community in comprehensive school, family, and community partnership programs. Elementary School Journal, 102, 19-34.

Sanders, M. G., & Harvey, A. (2002). Beyond the school walls: A case study of principal leadership for school-community collaboration. Teachers College Record, 104, 1345-1368.

Shartrand, A. M., Weiss, H. B., Kreider, H. M., & Lopez, M. E. (1997). New skills for new schools: Preparing teachers in family involvement. Cambridge, MA: Harvard Family Research Project.

Smrekar, C. E., & Mawhinney, H. B. (1999). Integrated services: Challenges in linking schools, families, and communities. In J. Murphy & K. S. Louis (Eds.), Handbook of research in educational administration (pp. 443-461). San Francisco: Jossey-Bass.

Smylie, M. A., Crowson, R. L., Chou, V., & Levin, R. A. (1994). The principal and community-school connections in Chicago's radical reform. Educational Administration Quarterly, 30, 343-364.

Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics (4th ed.). Boston: Allyn and Bacon.

Taylor, L., & Adelman, H. S. (2000). Connecting schools, families, and communities. Professional School Counseling, 3, 298-307.

Walsh, M. E., Howard, K. A., & Buckley, M. A. (1999). School counselors in school-community partnerships: Opportunities and challenges. Professional School Counseling, 2, 349-356.

Julia Bryan is an assistant professor, Department of School Psychology and Counselor Education, College of William and Mary, Williamsburg, VA. E-mail: [e-mail] Cheryl Holcomb-McCoy is an associate professor, School Counseling Program, University of Maryland at College Park.

Table 1. Means and Standard Deviations of School Counselors'
Perceived Involvement in 18 Partnership Role Behaviors
(N = 235)

School-Family-Community
Partnership Role
Behaviors (a)                          Involvement

                                       M      SD

Working with a team of school         4.02   1.07
staff, family, and/or
community members and
professionals (e.g.,
school mental health team)

Locating services and resources       4.01    .95
for students and their families
in the community (e.g., mental
health, nutrition, social
services, clothing)

Collaborating with community          3.95    .97
agency professionals

Working with school staff,            3.83   1.12
family, and/or community
members to advocate
for equity, access, and
academic success for students

Teaching parents and students         3.56   1.11
how to access services in
the school and community

Involving parents, family, and        3.54   1.15
community members in the
delivery of guidance activities
and services to students
(e.g., career guidance)

Using school and student data         3.29   1.19
to advocate for change in
the school on behalf
of students

Helping parents, family, and          3.23   1.15
community members organize
support programs for students
(e.g., tutoring and
mentoring programs)

Implementing programs to              3.20   1.18
help family and community
members understand the
school (e.g., parent
and family seminars)

Coordinating
school-community                      3.13   1.17
outreach efforts

Providing parent education            2.94   1.25
workshops and seminars

Serving on advisory councils,         2.85   1.41
or committees in the community

Facilitating the integration          2.82   1.37
of community services within
the school (e.g., mental
health and social services
housed in the school)

Implementing programs to help         2.69   1.15
the school staff understand
families and the community
(e.g., in-service training on
culturally diverse families)

Collaborating with local              2.54   1.15
businesses and industries

Using data to show the benefits       2.47   1.20
of school-family-community
partnership programs for students

Training the staff on how to          2.31   1.09
build effective school-family-
community partnerships

Conducting home visits to families    1.98   1.07

Overall                               3.13    .72

Note. Items were rated on a Likert scale: 1 = never,
2 = seldom, 3 = sometimes, 4 = frequently, 5 = always.
(a) These items were

Table 2. Sample Items and Factor Loadings for School
Counselor and School-Related Factors

Sample Items and Loadings for School Counselor Factors

Attitudes About Partnerships

School-family-community partnerships are very important
  for helping children succeed                                     .84
School-family-community partnerships are important
  for a good school                                                .81
School-family-community partnerships help counselors
  be more effective with students                                  .78

Perceived Barriers

There is a lack of resources for building                          .78
  partnerships
It is difficult to get teachers involved                           .71
  in partnerships
No one is available to lead in developing                          .69
  partnership activities

Confidence in Ability to Build Partnerships

I am confident in my ability to initiate school-family-            .76
  community partnerships
I lack the training necessary to build partnerships with          -.73
  the community
I lack the training necessary to build partnerships               -.72
  with families
I have the skills to collaborate with families
  and community members                                            .68

Commitment to Advocacy

I advocate for culturally and ethnically diverse children in       .72
  my school to ensure that they get equal access to rigorous
  classes
I believe that all children can succeed academically               .67
I would actively advocate for children even if it                  .66
  were not considered part of my role
I am a voice for all children in my school to ensure that          .61
  the school meets their needs

Attitudes About
Families and Communities

Most parents are interested  in their children's education        -.64

There are many students in my school who face severe economic,     .64
  social, and emotional needs
In our school families are disinterested in their                  .62
  children's education
Families in our school are hard to reach                           .62
Sample Items and Loadings for School-Related Factor
Collaborative School Climate

Most administrators and staff in this school support
  school-family-community partnerships                             .78
The principal is skillful in building relationships
  with parents and community members                               .78
The principal supports those who lead partnership
  activities and programs in the school                            .78
Involving family members in this school is part
  of regular practice                                              .76
This school's climate is conducive to fostering
  partnerships with families and the community                     .75
Partnerships with families and community members
  are valued in this school                                        .75
Parents and family members feel welcome in our school              .72
The school has a warm, friendly atmosphere                         .65

Table 3. Hierarchical Regression and Post Hoc Analyses of Factors
Related to School Counselor Involvement in School-Family-Community
Partnerships

Step and Predictor Variable            B      SEB        [beta]

Step 1
  Elementary school                    .427   .177        .193 *
  Middle school                        .072   .181        .032
  Combination school                   .445   .202        .171 *

Step 2
  Elementary school                    .283   .167        .128
  Middle school                        .045   .168        .020
  Combination school                   .408   .188        .157 *
  Collaborative
    school climate                     .365   .064        .369 ***

Step 3
  Elementary school                    .146   .115        .066
  Middle school                        .061   .113        .027
  Combination school                   .099   .128        .038
  Collaborative
    school climate                     .287   .052        .291 ***
  Role perceptions                     .580   .054        .583 ***
  Attitudes about
    partnerships                      -.137   .056       -.138 *
  Perceived barriers                   .031   .053        .031
  Confidence in ability                .233   .052        .234 ***
  Commitment to advocacy               .093   .050        .097
  Attitudes about families            -.063   .049        -.066

Post Hoc Analysis (a)

Step 3
  Elementary school                    .014   .144        .006
  Middle school                       -.030   .142       -.013
  Combination school                   .139   .161        .053
  Collaborative
    school climate                     .303   .066        .307 ***
  Attitudes about
    partnerships                       .039   .067        .039
  Perceived' barriers                 -.014   .066       -.014
  Confidence in ability                .324   .065        .325 ***
  Commitment to advocacy               .211   .061        .220 **
  Attitudes about families            -.164   .061       -.166 *

Step 4
  Elementary school                    .146   .115        .066
  Middle school                        .061   .113        .027
  Combination school                   .099   .128        .038
  Collaborative
    school climate                     .287   .052        .291 ***
  Attitudes about
    partnerships                      -.137   .056       -.138
  Perceived barriers                   .031   .053        .031
  Confidence in ability                .233   .052        .234 ***
  Commitment to advocacy.              .093   .050        .097
  Attitudes about families            -.065   .049       -.066
  Role perceptions                     .580   .054        .583 ***

                                               [DELTA]
Step and Predictor Variable       [R.sup.2]   [R.sup.2]

Step 1                              .042       .042 *
  Elementary school
  Middle school
  Combination school

Step 2                              .175       .133 ***
  Elementary school
  Middle school
  Combination school
  Collaborative
    school climate

Step 3                              .643       .468 ***
  Elementary school
  Middle school
  Combination school
  Collaborative
    school climate
  Role perceptions
  Attitudes about
    partnerships
  Perceived barriers
  Confidence in ability
  Commitment to advocacy
  Attitudes about families

Post Hoc Analysis (a)

Step 3                              .434       .259 ***
  Elementary school
  Middle school
  Combination school
  Collaborative
    school climate
  Attitudes about
    partnerships
  Perceived' barriers
  Confidence in ability
  Commitment to advocacy
  Attitudes about families

Step 4                              .643        .209***
  Elementary school
  Middle school
  Combination school
  Collaborative
    school climate
  Attitudes about
    partnerships
  Perceived barriers
  Confidence in ability
  Commitment to advocacy.
  Attitudes about families
  Role perceptions

Note: The reference category is high school.

(a) Steps 1 and 2 are the same as in the initial analysis.

* p < .05. ** p < .01. *** p < .001.