Predictors of Effectiveness of Collaborative Relationships of the USDA Youth At Risk Coalitions

Since 1914, Extension professionals have sought to bring together individuals and groups to address complex community issues. Collaboration and Coalitions are words that have been used to describe this process. Shrinking budgets and shortages of resources have made working together even more important. Making effective use of our resources to have real impact in our communities should be our primary reason for building effective relationships with other individuals and groups.

In 1991, 69 sites throughout the country were funded through the U.S. Department of Agriculture (USDA) to conduct innovative programs for Youth-At-Risk (YAR). The programs were funded through competitive grants to Extension Service field offices and were made up of three types: Science Technology and Literacy, School Aged Child Care, and Building and Developing Coalitions. These programs all had one common requirement, they must have a collaborative component.

This study examined selected situational factors and structural characteristics of the coalitions to determine if relationships exist between the perceived effectiveness of the collaborative relationships within the coalition and those situational factors and structural characteristics. The situational factors selected appeared to assess the conditions under which individual members try to build collaborative relationships.

In this study coalition size, coalition type, resource dependence, awareness, consensus and domain similarity were used. Coalition size referred to the number of active collaborators in the coalition. Coalition type was the type of Youth-At-Risk site studied (science, technology, and literacy; school aged child care; or developing and building coalitions). Resource dependence is the extent to which members needed resources from the other members of the coalition in order to accomplish their goals. Awareness referred to prior knowledge of or relationships with other agencies or individual members of the coalition. Consensus was the extent to which coalition members agreed on needs, problems, solutions and methods. Domain Similarity was the extent to which members shared like goals, funding sources, staff skills, clientele and services.

Structural characteristics were characteristics of a coalition which appeared to capture its structural essence. In this study resource flows, formalization, communication frequency, and communication quality were used. Resource flows referred to the direction and intensity that resources were shared among coalition members. Formalization was the extent to which written, verbal or legally binding agreements were used by the coalition. Communication frequency was how often members communicated; either written, as a group, individual face-to-face discussions, or by telephone. Communication quality referred to the members' perception of the quality of communication between themselves and other members of the coalition.

Perceived effectiveness referred to members' perception of the effectiveness of the coalition relationship. The extent to which members carry out their commitments and believe the relationships are worthwhile, productive and satisfying.

The findings from this study may help the coalition builders to maximize those situational factors and structural characteristics that promote effectiveness of the collaborative relationships, resulting in better services for youth and families. The findings may also help decision-makers in funding those projects that have the greatest likelihood of success. Many have touted coalitions and collaboration as methods that can increase efficiency and effectiveness when addressing complex societal issues (Quick, Flashman & Gibeaux, 1981; O'Connor, Albrecht, Burton & Newquist-Carroll, 1984; Lippett & Van Til, 1981; Stevens, 1990; Rossi, Gilmartin & Dayton, 1982; Black, 1983).

Different situational factors and structural characteristics have been identified which may influence the effectiveness of the relationships. Different levels of these factors and characteristics could positively or negatively affect the effectiveness of the relationships (Van de Ven, 1974; Astroth, 1990; Bennard, 1991). Characteristics of members and coalitions was also described as being an important component of effective collaborative relationships (Black, 1983; Miller, 1983; Hord, 1986; Lieberman, 1986; Johnson, Bruininks & Thurlow, 1987; Yukl, 1989; Bennard, 1991).

The structural framework for this study was developed using characteristics of members and coalitions and the structural characteristics and situational factors identified by Van de Ven and Ferry (1980). Van de Ven (1974) suggested that there were structural characteristics and process dimensions to interorganizational relationships. Van de Ven's work with child service agencies and their partners piloted the use of these situational factors and structural characteristics.

The study sought to understand if situational factors and structural characteristics were associated with perceived effectiveness. It also examined the nature and strengths of any relationships. The following research objectives were formulated to guide the study:

1. To describe how situational factors (coalition size, coalition type, resource dependence, agency or personal awareness, consensus and domain similarity) were experienced by the active collaborators.

2. To describe how the structural characteristics (frequency of communications, resource flows, and formalization of agreements) were experienced by the active collaborators.

3. To determine the degree to which the active collaborators perceive the collaborative relationships to be effective.

4. To determine if and to what extent, selected situational factors are associated with perceived effectiveness of the collaborative relationships.

5. To determine if, and to what extent, selected structural characteristics are associated with perceived effectiveness of the collaborative relationships.

6. To determine the best predictors, from the variables studied, of the perceived effectiveness of the collaborative relationships.

The study is limited in that the results are generalized only to the population studied. This study sought to explore the effectiveness of the collaborative relationships and in no way sought to examine the effectiveness of the YAR sites, their projects or the overall USDA/YAR initiative. No such inferences should be made.

This study was ex post facto in nature and was designed to gather data concerning the nature and strength of relationships between variables. The independent variables for the study were the Situational Factors and Structural Characteristics of the coalitions and characteristics of the coalitions and its members. The dependent variable was the effectiveness of the relationships among coalition members as perceived by the active collaborators.

The population of this study consisted of 58 Youth-At-Risk grant sites funded for their second year in 1992. The active collaborators involved in each site were surveyed.

This study was a census, thus the major threats to external validity were frame error and non-response error. The frame was constructed using a list of USDA-funded YAR sites and was considered to be complete and accurate. A comparison of early respondents to late respondents was conducted; the t-test yielded no significant difference on any of the variables selected (Miller and Smith, 1983). In an attempt to control measurement error as a threat to internal validity, a panel of experts reviewed the mail questionnaire for content and face validity.

The instrument used was adapted from a similar instrument used by Van de Ven and Ferry (1980). The instrument was pilot tested with a YAR site not included in the study. Chronbach's alpha was used to measure the internal Consistency of the survey instrument. A .50 Chronbach's alpha (Nunnally, 1967) was established a priori as being a satisfactory level of internal consistency. Calculated Chronbach's alpha ranged from .60 to.93. Therefore the instrument was judged to be reliable.

Data were collected using a variation of the mailing procedures recommended by Dillman (1978). Useable data was received from 39 of the 58 YAR sites for a 67% response rate. The design of the study was a one shot case study. Data were analyzed using the SPSSPC+ computer program using descriptive statistics and correlations.

Stepwise multiple regression was used to determine the best predictor(s) of the dependent variable--perceived effectiveness of the collaborative relationship. The independent and extraneous variables entered into the regression model included: resource dependence, personal or agency awareness, consensus, domain similarity, frequency of communications, resource flows from respondent, formalization of agreements, quality of communications, race, sex, education level, employment status, length of time employed, and community type. The total R2 was computed to determine the amount of variance accounted for by the linear combination of the independent and extraneous variables.

The study included a description of the population, the active collaborators, and how the structural dimensions and situational factors were experienced by the respondents. The respondents reported that they experienced effective collaborative relationships. The item asking if the time and effort spent maintaining the relationships was worthwhile, was rated highest among the effectiveness measures (M = 4.18 on a 5 point scale).

YAR site size had a negligible association with the dependent variable (r = -.07). The type of YAR site and domain similarity had low positive associations (r = .23 and .27), resource dependence and agency or personal awareness both had moderate positive associations (r = .35 and .43) and consensus had a substantial positive association (r = .52) with the dependent variable. Relationships between the structural dimensions and the dependent variable ranged from low to moderate. Frequency of communications had a low positive association (r = .23), resource flows to and from the members, quality of communications, and formalization of agreements all had moderate positive associations (r = .49, .45, .43, and .31 respectively) with the dependent variable.

Stepwise multiple regression was performed to determine the best predictor(s) of the dependent variable. All of the independent and extraneous variables described in the objectives were entered into the regression model. Sixty-four percent of the variance in the dependent variable was explained by the linear combination of the variables consensus, formalization of agreements, and resource flows from the respondents to the coalition. These variables uniquely explained 46%, 13%, and 5% of the variance in the dependent variable respectively. The regression indicates that for every one point change in consensus there will be a .70 point change in the dependent variable. Likewise a one point change in formalization will be a corresponding change of .23 in the dependent variable, and a one point change in resource flows from the respondent will have a corresponding .10 change in perceived effectiveness of the collaborative relationship.

The best predictors of perceived effectiveness of the collaborative relationships were found to be consensus, followed by formalization of agreements and resource flows from the respondent. If these factors are known, perceived effectiveness of the collaborative relationships can be predicted. By maximizing these variables, coalition builders can increase their likelihood of building effective collaborative relationships.

A model was constructed to indicate which situational factors and structural characteristics had the most influence on perceived effectiveness of the collaborative relationships. The model suggests maximizing resource dependence, awareness and consensus as situational factors. The model also suggests maximizing resource flows between members and the coalition, formalization of agreements, and quality of communications as structural characteristics.

YAR coalitions should be built around specific issues, developing mutual dependence on resources among members, making efforts to increase their awareness of the goals and objectives of the other members. When YAR coalitions are formed, members should reach agreement on needs, goals, objectives, target audiences, services to be provides, methods and roles of each of the members. Communication among members should be frequent, with follow up in writing outlining decisions reached and action items. The magnitude and diversity of resources available from each member should be maximized and members should be encouraged to share those resources to the greatest extent practical. Agreements, agendas, minutes, etc. should be documented and available to all members. Members who enter into YAR coalitions should commit themselves to fully supporting the coalitions's efforts. Partial support may not lead to the desired outcomes.

Other studies should be conducted on YAR sites and among members who did not have effective collaborative relationships in order to identify how they experienced these situational factors and structural dimensions. This study should be replicated with other coalition types in and out of Extension in order to give broader meaning to the findings. This study should also be replicated with those YAR sites not included in this study. A research study should also be conducted to examine the relationship between collaborative effectiveness and project effectiveness among the YAR sites.

Astroth, K. A. (1990, October). Strategic alliances: The 4 C's process. Paper presented at the Idaho Annual Extension Conference, Idaho Annual Exension Conference, Moscow, Idaho.

Bennard, B. (1991, March). Collaboration fosters creative problem solving. Western Center News, 4(2), 4, 9.

Black, T. R. (1983, May/June). Coalition building - Some suggestions. Child Welfare, LXII(3), pp. 263-268.

Dillman, D. A. (1978). Mail and telephone surveys: The total design method. New York: Wiley.

Hord, S. M. (1986). A synthesis of research on organizational collaboration. Educational Leadership, 43(5), pp. 22-26.

Johnson, D. R., Bruininks, R. H., & Thurlow, M. L.(1987). Meeting the challenge of transition service planning through improved interagency cooperation. Exceptional Children, 53(6), pp. 522-530.

Lieberman, A. (1986). Collaborative work. Educational Leadership, 43(5), pp. 4-8.

Lippett, R., & Van Til, J. (1981). Can we achieve a collaborative community? Issues, imperatives, potentials. Journal of Voluntary Action Research, 10, 7-17.

Miller, L. E., & Smith, K. L. (1983). Handling nonresponse issues. Journal of Extension, 21, 45-50.

Miller, S. M. (1983). Coalition etiquette: Ground rules for building unity. Social Policy, 4(2), pp. 47-49.

Nunnally, J. C. (1967). Psychometric theory. New York: McGraw-Hill

O'Connor, R., Albrecht, N., Burton, C., & Newquist-Carroll, L. (1984). New directions in youth services: Experiences with state-level coordination. U.S. Department of Justice & U.S. Department of Health and Human Services. Washington, D.C., U.S. Government Printing Office.

Quick, S., Flashman, R., & Gibeaux, A. (1981). SOS learning networks: A model of interorganizational cooperation. Journal of Voluntary Action Research, 10, 67-77.

Rossi, R. J., Gilmartin, K. J., & Dayton, C. W. (1982). Agencies working together: A guide to coordination and planning. Beverly Hills: Sage.

Stevens, G. L. (1990). A process for building coalitions. University of Nebraska-Lincoln, Cooperative Extension, Institute of Agriculture and Natural Resources.

Van de Ven, A. H. (1974). On the nature, formation, and maintenance of relations among organizations. Academy of Management Review, 5(1), pp. 113-129.

Van de Ven, A. H., & Ferry, D. L. (1980). Measuring and assessing organizations. New York: Wiley.

Yukl, G. A. (1989). Leadership in organizations (2nd ed.). Englewood Cliffs, NJ: Prentice Hall.