MSU RETURNING MATERIALS: Place in book drop to LIBRARIES remove this checkout from “ your record. FINES will be charged if book is returned after the date stamped below. 13%qu- Ln, 9 wagqomw?) 54ml.- £,;; 3, E Q 93 1h \ MOVE“ *W/ém‘" 99% 3'89' _ ' mx Ans-EC flwasaqm anf'aaab, ‘ - r” - m. 19911 mm W” . I15 g WNW m5 ‘ A w; , £33049; . 3— 34% $332719” ‘ INNOVATION ADOPTION DECISIONS IN ORGANIZATIONS: AN EMPIRICAL INVESTIGATION by Rand Jeffery Gottschalk A MASTERS THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1982 B TRA (I) T 3,3- (J) INNOVATION ADOPTION zECISIOFS IN ORGANIZATIONS: U A ’1‘ F4 ('3 Z ff) AN EMPIRICAL INVESTI By Rand Jeffery Gottschalk O F.) The adoption innovations is a popular area for study among social scientists in a variety of disciplines. This study investi- gated factors in the adoption decision as perceived by organiza- tional members involved in the decision. Three hundred and eighteen organizations that had and had not adopted one of eight innovative programs in the criminal justice area and educational area wer contacted through conversational phone interviews. These reasons for adeption were rationally and empirically scaled into four scales: Expected Smooth Implementation, Cost and Financial Support. Changes in Roles and Role Relationships, and Support from Organiza- tional Members. Program and content area differences were found on the Expected Smooth Implementation scale. Program differences were also found on the Support from Organizational Members scale. Orga- nizational demographics did not differ between adopting organiza— tions and organizations that had never heard of the program. Educa- a tional and criminal justice organizations did differ in terms of th- number of people involved in decision making. However, the corre— lation between participation in decision making and adoption status was low. Reasons for the findings and possible implications are discussed. ACKNOWLEDGEMENTS 4. I would like thank all my committee members for their help. My chair, Neal Schmitt, has been extr mely patient with me at all times during my time at the university. Without Neal's patience, availability, and willingness to help at any time I would not have been able to do this thesis. Bill Davidson was also extremely helpful. His incisive comments, power hitting, and above all his sabbatical gave me the push that I had been lacking. Mike Cook was thrust into the difficult position of being the only committee member who was not involved in the research team. Mikes's comments were always helpful and he handled this difficult position extremely gracefully. I would also like to thank both Craig Blakely and Dave Roitman for their help at all stages in this process. They have both helped me to think better as a function of working in a group with them. Craig, as project director, has removed many of the barriers that I always seemed to find when it came to this project. That's not trivial. Dave's tolerance for my different style and his hard working example were very motivating for me. I would also like to thank Ray, Wes, Rick, and Martha for ii their tolerance and friendship over the last four years. I never would have lasted without them. Finally, I would like to thank my parents for their patience and help over the years. In their lives, they have provided me with a tough act to follow. Yes Mom, I'm now one step closer to a job. TABLE OF CONTENTS Fag LIST OF TABLES. . . . . . . . . . . . . . . . . . . . . . . . vi LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . 1x LIST OF APPENDICES. . . . . . . . . . . . . . . . . . . . . . X Chapter :- INTRODUCT:[(FNO . O C C . C O D C O O O C O O I O O I O O 1' -J Statement of the Problem. . . . . . . . . . . . . . . . POlicy MOdels O O O O O O O O O O O O O O O I O O O O 0 Individual Influences on the Adoption of Innovations. . GUI“ Organizational Influences on Innovation Adoption. . . . 1 Methodological Criticisms and Considerations. . . . . . 2O Exemplary Studies . . . . . . . . . . . . . . . . . . . 26 Implications of the Exemplary Studies . . . . . . . . . 91 Rationale for the Present Study and Research Questions . . . . . . . . . . . . . . . . . . . . . . . 31 II. IIIETHOD. O O I O O I O O O i O O O O O O O O I O O O O O 34 Sample. . . . . . . . . . . . . . . . . . . . . . . . . 34 Innovations . . . . . . . . . . . . . . . . . . . . . 34 Unit of analysis. . . . . . . . . . . 39 Random sampling of organizations withi innovations . . . . :3 Purposive sampling of organizations . . . . . . . . . 41 Switching from random to purposive lists. . . . . . . 41 Respondents . . . . . . . . . . . . . . . . . . . . . 42 Measures. . . . . . . . . . . . . . . . . . . . . . . . 48 AdOption Decision Questionnaire .-. . . . . . . . . . 48 Organization Profile. . . . . . . . . . . . . . . . . 49 Procedure . . . . . . . . . . . . . . . . . . . . . . . 49 Interviewers. . . . . . . . . . . . . . . . . . . . . 49 Interviewer training. . . . . . . . . . . . . . . . . 49 Interview administration. . . . . . . . . . . . . . . SO Reliability and validity. . . . . . . . . . . . . . . 52 iv (1) Chapter III. RESULTS Reliability an d Validity. Reasons for Adoption. . . Number of re Scaling of t Differences Organizational Participation Other Analys 3 (V1- \ SIG (1) IV. DISCU . . Limitations of APPENDICES . . . . REFERENCES . . . . he reason 330113 0 o 0 Demographic in DeCision to Explicat f d K L adeption on the 8 nd Envir ing. . . Drogram D 0 FJ. m :3 b5. i...) .— (—1- in (I) (D "d .__\ n 3) Ci) \F'l p. 0\ \fl LT! U1 U1 —- w ._. m CO 03 C1343 (DNJ \l C) \ KN d1 (II TABLE -4 \JJ U1 0 13 LIST OF TABLEL Innovations Selected in the Present Study . . . . . . Number of Adapters and Unaware NonuAdOpters bv Proggram I I I I I I I I I I I I I I I I I I I I I I I rU Md ’0 rcentage, Number, and Current Content Coded Job C; L o_ition b Program for Adepters. . . . . . . . . . . Percentage, Number, and Current Content Coded Job \J Position by Program for Unaware Non-Adopters. . . . . Inter-Coder Agreement on the Adoption Decision Questionnaire and Organization Profile by Program . . Inter—Respondent Agreement on the Adeption Decision Questionnaire and Organizational Profile by Program I I I I I I I I I I I I I I I I I I I I I I I Means and Standard Deviations of the Number of Reasons Given by each of the Programs . . . . . . . . . . . . Scale Correlations of the Reasons for Adoption. . . . Scale Item Means and Standard Deviations by Program I I I I I I I I I I I I I I I I I I I I I I I Results of Univariate Analysis of Variance on Expense and Financial Support Scale, Changes in Roles and Role Relationships Scale, Expected Smooth Implementation Scale, and Support Scale by Program . . . . . . . . . Scheffe Procedure for Comparisons Among all Pairs of Means on Expected Smooth Implementation Scale . . . . Scheffe Procedure for Comparisons Among all Pairs of Means on Support Scale. . . . . . . . . . . . . . . . Number and Percentage of Adopters (A) and Unaware Non—Adopters (NA) who have Increasing, Decreasing, and Stable Numbers of Clients, Administrative Staff, and quont-‘Line Staff. I I I I I I I I I I I I I I I I I I vi i PACE 4:. 46 47 «I Q 73 14 17 19 22 23 LE Results of Analysis of Variance on Numbers of Clients, Administrative Staff, and Front-Line Staff (Increasing, Decreasing, or Stable) by Program and AdOption Status. . . . . . . . . . . . . . . . . . . . . . . . . Means and Standard Deviations for Adepters on Number of Levels Participating in the Adoption Decision by Program Uncollapsed and Collapsed to Control for Possible Bias. I I I I I I I I I I I I I I I I I I I I I I I I I Means and Standard Deviations for AdOpters on Number of Levels Participating in Decisions in General by Program Uncollapsed and Collapsed. . . . . . . . . . . Results of Analysis of Variance of Collapsed Number of Levels Involved in Decision Making in General by Program and Adoption Status for both AdOpters and Non—Adopters. . . . . . . . . . . . . . . . . . . . . . Results of Analysis of Variance of Collapsed Number of Levels Involved in the AdOption Decision by Program for Adopters. . . . . . . . . . . . . . . . . . . . . . Means and’Standard Deviations for AdOpters on Number of Levels Involved in the Adoption Decision — Number of Levels Involved in Decisions in General Collapsed and UflCOllaps.edo I I I I I E I I I I I I I I I I I I I I I Results of Analysis of Variance of Collapsed Number of Levels Participating in the AdOption Deci ion — Number of Levels Participating in Decisions in General by Program for Adopters . . . . . . . . . . . . . . . . . Correlations Among Four Scales, Number of Levels Participating in both the AdOption Decision and Decisions in General (Collapsed), Number of Levels in Adoption Decision - Number of Levels in General (Collapsed), Length of Time Program has been in Use in the Organization, nd Demographic Variables for Adopters. . . . . . . . . . . Results of Analysis of Variance of How Long the Program has been Used by Program. . . . . . . . . . . . . . . . Scheffe Procedure for Comparisons Among all Pairs of Means on How Long the Program has been in Use. . . . . vii *U :x» C) L1?! 74 -\l «Q -J ‘43 80 82 e4 85 87 TABLE 24 PAGE Results of Analysis of Variance on Expense and Financial Support Scale, Changes in Roles and Role Relationships Scale, Expected Smooth Implementation Scale, and Support Scale by Program and Age. . . . . . 88 Scale Average Item Means, Variances, and Multivariate Analysis of Variance for Respondents' Job Position . . 90 vi? PO Reference Items in Domain Scales LIST OF FIGURES Categoriztion 1‘ 0; -nd Scale Names. ix Organizations. *U :2) C} [13 APPENDIX lid Coding form for the Adoption Decison Questionnair . . . . . . . . . . . . . . . . . . Coding form for the Organization Profile . . . . . Interview Guide. . . . . . . . . . . . . . . . . CHAR P—J E13 '3 H INTRODUCTION Statement of the Problem Innovation diffusion is an area of great importance both practically and theoretically. The practical importance comes into play if new practice is perceived to be better than existing practice. It is then imperative for the sake of both organizational efficiency and survival that this process be carried out quickly. The theoretical interest in this area is very much linked to practicality. Since an innovation is usually defined as any practice or product that is new to the organization (Rogers & Shoemaker, 1971), the only way that organizations change is through the innovation process. This may account for both the multi-disciplinary nature of the study of innovation and the large numbers of studies dealing with this tOpic (Rogers & Eveland, 1975). Policy Models Several policy models of innovation diffusion have been F0 preposed. For example, Datta (1981) discussed beliefs concerning change in educational organizations. This is a question of more than academic interest since the policies of the federal govern- ment are based on these beliefs. The first strategy she dis- cussed has been labeled the Directed DevelOpment approach. It is also known as the Research and DevelOpment model (House, Kerins, & Steele, 1972), or the Research, DeveIOpment, and Diffusion model (Yin, 1978). Characteristic of this model is the use of programs that have been through some kind of effectiveness demon- stration or validation process. Organizations are given incentives, usually in the form of money, to start these programs. This type of model takes a fidelity parapective; that is, since the programs are proven effective it is imperative that they be adopted and implemented as devoloped. This Directed Development approach has been criticized on a number of grounds. The approach assumes a passive user pOpulation; potential users of a program or approach are assumed to be just :waiting passively for a program to come along and solve their prob- lems (House, Kerins, & Steele, 1972). Second, this approach focuses on the environment external to the organization rather than to con- ditions internal to the innovating organization (Yin, 1978). Third, gthis model pays attention only to the determinants of the adoption of innovations, rather than looking at the adoption, implementation, and eventual incorporation of the innovation as an ongoing part of the organization. It therefore takes what has been called the adop- .tion perspective (Berman & McLaughlin, 1974). A second policy model for influencing change suggests that problems should be solved at the local level. The intent of H) ederal policy should be to foster problem solving skills at the local level and to provide programs that are easily adaptable to the needs of the local system. Datta refers to this model as the Local Problem-Solving and Mutual Adaptation view. The support for the second perspective of change and most 0 (+- U) the empirical criticism of the Directed DevelOpment model res on the work of Berman and McLaughlin (1978). This series of reports deals with the impact of four major federal programs on innovative practices in public schools. The most widely known conclusions of this study concern the lack of impact of the federally supported innovations on local school practices. Datta (1981) has pointed out that the conclusions and implications drawn from this report are misleading and not totally based on the data. While Berman and McLaughlin (1974) purport to be studying the impact of major federal expenditures to aid local school systems adopting and implementing well developed innovations, this is not in fact the case according to Datta, who noted that the "massive federal funds" amounted to only about an additional $100 per student served per year. Another widely cited finding of the study was that none of the programs seemed to be implemented as planned. Datta suggests that this was due to the fact that none of the programs were disseminated specifically; the "programs" were in fact only general directives to serve a specific targeted group or to work on a general need. In sum, Datta's basic thesis was that the Directed Development model was never really tested by the Berman and McLaughlin study. Stages of Innovation Although there is disagreement concerning the basic policy model, there is currently a consensus among researchers and policy makers that innovation diffusion should be conceived of as consisting of three major stages: Adoption, Implementation, and Incorporation. This investigation focuses on adoption, although linkages with implementation and incorporation are considered. The linkage between implementation and adoption is estab- lished through the idea of future expectations on the part of users of the program. The process of innovation is a very uncer- tain one (Tornatzky, Roitman, Boylan, Carpenter, Eveland, Hetzner, Lucas, & Schneider, 1979). Organizations make changes on the basis of the predicted effects of the changes. A major class of determinants of the adoption of new programs is predic- tions by relevant organizational actors concerning the effects on the organization of the implementation of the program. For this reason it is imperative to study these predictions or expect— ations. This investigation involved an empirical investigation of these expectations. First, the literature on the adOpotion of innovations is discussed. A discussion of some of the methodological criticisms of this tradition will follow. Three "exemplarary" studies particu— larly relevant to the present investigation are then discussed. The manner in which the methodological criticisms have been taken into account in the present investigation is presented followed by the rationale, research questions, method, and results of the present investigation. Individual Influences on the Adeption of Innovations A seminal work embodying the classical diffusion approacn is that by Rogers and Shoemaker (1971). Rogers and Shoemaker re- viewed over 1500 empirical studies. rhey organized the elements. "Crucial elements in the 0 literature into four basi diffusion of new ideas are (1) the innovation (2) which is communicated through certain channels (3) over time (4) among members of a social system"(p. 18). Rogers and Shoemaker used the voting method of meta-analysis (Jackson, 1980). This strategy involves looking at all the empirical studies dealing with a given proposition and reporting the percentages of studies showing effects in a positive or negative direction as well as reporting the percentage of studies that show no effects for a given proposition. The directional conclusion is based on which alternative was supported by the highest percentage of studies. The innovation category breaks down into five attributes for which Rogers and Shoemaker review the literature. These five attributes of innovation are all considered as per— ceived by the adopter: V,(1) Relative advantage—-the degree to which an innovation is perceived as better than the idea it supersedes. O’\ (2) Compatability-—the degree to which an innovation is perceived as being consistent with the existing values, past experiences, and needs of the receivers. (3) Complexity--the degree to which an innovation is perceived as difficult to understand and use. (4) Trialability——the degree to which an innovation may be experimented with on a limited basis. (5) Observability—-the degree to which the results of an innovation are visible to others (pp. 22-23). Relative advantage, compatibility, trialability, and observ- ability, are all positively related to the rate of adoption of an innovation, while complexity was not found to be related to the rate of adOption. It should be noted that Rogers and Shoemaker (1971) dealt almost exclusively with the adOption of innovations by_ individuals. They compared the characteristics of individuals who adopted early to those who adopted later in the process. Characteristics included variables such as education, social status, dogmatism, empathy, rationality, and intelligence. These findings were taken mostly from the rural sociology tradition which investigates the spread of new farm practices and new crops. The question of the generalizability of these findings to the complex situation of an individual in an organization remains. An example of the limited generalizability of the Rogers and Shoemaker findings was provided in a compelling analysis by Pincus (1974). Pincus likened the public school to a public utility in terms of the bureaucratic structure and incentives operating to adOpt new technologies. Both a public utility and the public schools are non-market oriented with a captive pepulation of non-selected clients. The public schools, however, are likely to be even less innovative than the public utility since there exists little or no consensus concerning the aims of schooling, the technology of schools is unclear, the quality of schooling can vary substantially within a school district, and there is little incentive for schools to be economically com- petitive. Given this situation, the attributes of innovation that Rogers and Shoemaker identified as leading to adoption are likely to be negatively related to adeption in the public schools. Schools are more likely to adopt innovations that are not readily observable in their effects since this raises the specter of evaluation. Schools should, therefore, be less likely to adopt trialable innovations since the decision to use the innovation can be reversed. Given the kinds of incentives pointed out by Pincus, the situation with respect to the deter— minants of the adoption of innovations by public service bureaucracies is likely to be vastly different than the deter- minants of the adeption of a new seed or new farm practice by an individual farmer. Two recent studies looked at the relative contribution of individual-level and organizational-level variables within an empirical framework. Baldridge and Burnham (1975) studied 20 sohools in seven different school districts. They interviewed all district superintendents and principals. In addition, 53 opinion leaders (identified by principals), 309 change partici- pants (participating in the change but not instigating the change), and 775 other teachers (a 50% random sample of the re- maining teachers) were questioned. The dependent variable was the adOption/non-adoption of one curricular innovation and one organizational innovation selected from each school. In order tQ be selected for study, an innovation had to meet two criteria. First, the innovation had to be rated as important by knowledge— able observers, meaning that the observers felt that the change had "real promise" for change in a major educational area. Second, the innovation had to be well established and appear likely to continue for a significant period of time. It was de- termined that individual characteristics, such as sex, age, work experience, education, career satisfaction, social origin, and cosmOpolitanism did not differ among the three groups of teachers studied. The organizational characteristics of district size and complexity were found to be positively related to the adoption of innovations. In addition, indicators of environmental heteroge- neity were found to vary positively with the adoption of innova- tions. Environmental change did not relate to the adoption of innovations. Together size, complexity, and environmental heter- ogeneity accounted for 31% of the variance in the adoption of innovations. Most importantly, the authors concluded that indi- vidual level variables are not important in the adoption of orga- nizational innovations. Rage and Dewar (1973) used the concept of values held by organizational personnel to predict the number of new practices adopted in a three-year period in 16 health and welfare organizations. Two types of Operationalization of values were used. The first referred to the formal elite, the high level personnel and executive director of the organization. The second referred to the behavioral elite, which included the executive directors and those personnel who reported they always or usually participated in strategic decisions. The values concerned the recognition of the need for change in the environment (an individual level construct). In this study, Hage and Dewar (1973) used as a definition of innovation, anything that was new to the organization. Both value explanations emerged as the strongest predictors of the adoption of innovations. Two measures of organizational complexity, the number of occupational specialities and the degree of professional activity, were also significantly related to the adoption of innovations. However. with all the structural variables held constant, the elite values were still better predictors of the adoption of innovations than the organizational characteristics. The structural variables did not effect the predictive power of the values. This study is important because it shows (contrary to Baldridge & Burnham, 1975) that individual variables can be predictive of organizational innovation over and above the influence of struc- tural eXplanations. O The papers discussed in this section indicate why the indi‘ vidual tradition in the classical diffusion approach may not be totally applicable to the tOpic of organizational innovation. Organizational innovation is a complex phenomena involving multi- ple actors embedded in a context of formal and informal authority structures, personal relationships, as well as organizational relationships to clients and the environment (Pincus, 1974). This context may serve to constrain the individual organizational actor or group (Hage & Dewar, 1973) thus attenuating the rela~ tionships among individual variables and innovation adOption (e.g., Baldridge & Burnham, 1975). However, given the finding that innovation adoption was predicted more accurately with indi- vidual level variables (values) than with organizational proper- ties (Hage & Dewar, 1973), totally ignoring the whole class of individual variables seems unwarranted. In sum, the assessment of under what organizational and environmental conditions indi- vidual variables play a role in organizational innovation needs further clarification. Organizational Influences on Innovation Adoption Zaltman, Duncan, and Holbek (1973) used a framework similar to that of Rogers and Shoemaker (1971) to look at the adoption of innovations by organizations. They included the five innovation attributes of Rogers and Shoemaker, but they also expanded the list. Other attributes discussed by Zaltman et. al. included: cost (both financial and social), returns to investment, risk and iuncertainty, communicability, scientific status, point of origin, terminality (whether there exists some point in time after which the adoption of the innovation is less rewarding, useless, or impossible), reversibility, commitment (degree required for successful use of the innovation), impact on interpersonal relationships, and the gateway capacity of the innovation (the extent to which the adoption of one innovation makes the adoption ‘of other innovations easier). Some of these attributes are clearly not attributes of the innovation alone, but depend on the nature of the adepting organization (Downs & Mohr, 1976). In addition to innovation attributes, Zaltman et. al. discussed organizational characteristics related to the adoption of innovations. These are almost the conceptual equivalent of the characteristics of individuals related to early adoption as discussed by Rogers and Shoemaker. Zaltman et. al. (1973) discuss five characteristics of organizations affecting the innovation process: (1) Complexity--the number of occupational specialities in the organization and their professionalism. (2) Formalization--the emphasis placed on following specific rules and proce urea in performing one's job within the organization. (3) Centralization~-the locus of the authority and decision making in the organization. (4) Interpersonal Relations-~the degree of impersonality in interpersonal relationships within the organization. (5) Ability to Deal with Conflict—-how well the organization '1 \D —163). ~ deals with conflict (pp. 1 \ 0) p.) . Q... The effects of these five properties are complex. Some in the adoption of innovations yet hinder the effective implementation of innovations. Other preperties hinder adoption yet aid in the effective implementation of innovations. For example, complex organizations are more aware of new practices and all other things being equal are more likely to adopt innovations. However, this professionalism is likely to lead to disagreements over how the new idea is to be put into practice. Highly formalized organizations are likely to encounter a large amount of resistance to the changing of practices that the ad0ption of an innovation will necessitate. However, once the decision is made to adopt and the practices entailed by the innovation are written into the jobs of the organization's members, there is likely to be little problem in implementation. Zaltman et. al. (1973) emphasized that the effect of the attributes of the innovation are likely to vary depending on the characteristics of the organization. This arguement implies the presence of statistical interaction between innovation attributes and organizational properties. In sum, a major accomplishment of the Zaltman et. al. (1973) work was to emphasize the importance of organizational innovation and to point out some of the applicability of the Rogers and Shoemaker work to this area. Much of the research in the area of organizational innovation has been carried out since the publication of the Zaltman et. al. (1973) work and was not included in their review of the literature. The research of Hage and Aiken (1967) and Aiken and Hage “Q 1 is typical of many of the studies concerning the effects v (19 f organizational properties on innovation adoption. In both 0 studies the same sample and dependent variable were utilized. Executive directors of 16 (10 private and six public) health and welfare offices were interviewed and asked how many new practices they had started in the previous five years. This served as the dependent variable. The organizational prOperties were obtained by interviewing all the directors and department heads, one~half the staff in departments of less than 10 members, and one—third of the staff in departments with more than 10 members. A total of 314 staff were interviewed. These responses were aggregated so as not to give undue weight to the responses of the lower level personnel over the more numerically sparse higher level personnel. The authors concluded that: the number of occupational specialities correlated .48 with innovativeness, the degree of professional activity associated with each occupation correlated .37 with innovativeness, the degree of participation in decision making correlated .49 with innovativeness, and the degree of job codification (formalization) correlated —.47 with innovativeness (Hage & Aiken, 1967). Aiken and Hage (1971) followed the same data collection strategies, except that they asked the executive directors about the number of new practices adepted by the organization in the preceeding five years, three years after the original data collection efforts. Number of occupational specialities (.59) and degree of professional activity (.63) were significantly related to the number of new practices adopted by the organization in the preceeding five years. In addition, the presence of a rule manual (-.60), the number of communications (.46), and the frequency of communication meetings (.53) were all found to be significantly related to the measure of innovation. In toto, these findings are taken to suggest that the more organic form of organization (Burns & Stalker, 1961) will adOpt new practices more readily than the mechanistic organization. These conclusions were causal. However, since the data were correlational in nature, a reversal of the direction of causality cannot be ruled out. In other words, it may be that organizations which adopt innovations become more organic over time because of the adeptions. Emphasizing the effect of; innovation adaption on changes in organizational properties is a no less compelling explanation than the authors' conclusions. Fairweather, Sanders, and Tornatzky (1974) provided additional although somewhat qualified support for the idea that participative decision making leads to the adoption of innovations. However, the results provided strong support for the argument that face—to-face "active" dissemination methods (e.g. demonstrations and site visits) are more effective than "passive" methods (e.g. articles and brochures) in disseminating complex social technologies. This study attempted to persuade state and federal hospitals to adopt the Lodge Program, an innovation which consists of creating autonomous problem solving vb \Yi groups of patients in the hospital and then moving these functioning groups into the community. Two hundred and fifty—five hospitals were randomly assigned into one of three persuasion strategies; sending of a brochure, holding a workshop, or establishing a demonstration ward. The dependent variable had three levels; no change/no persuasion attempt permitted, no change/persuasion attempt permitted, and persuaded to change. It should be noted that of the 255 hospitals contacted, 23 were persuaded to change. It was found that initial entry (persuasion attempt) was easier in the brochure and workshop condition than in the demonstration ward condition. However, the demonstration ward led to more adoptions of the Lodge Program. AdOption of the program did not vary as a function of the setting of the hospital (rural vs. urban), the social status of the person contacted in the hospitial, consultant experience, or whether the hospital was state or federal. Cluster analyses of questionnaire data revealed that the social change process was more dependent on participative decision making in the workshOp and brochure approaches (inactive) than in the more action oriented demonstration ward approach. Tornatzky, Fergus, Avellar, Fairweather, and Fleischer (1980) dealt with the effects of the number of initial contacts, group enhancement, and the involvement of staff and/or administrators in decision making on whether 108 hospitals wanted a workshOp and further implementation assistance on the Lodge “A O". program. The three variables mentioned above were experimentally manipulated. None of the variables affected the consultation decision, and only the level of staff contacted affected the initial workshOp decision. The effect of the level of staff was such that when ward level personnel were included there was a greater likelihood that the hospital would participate in the workshop. It should be noted that 30 of the 108 organizations asked for consultation at a later time. The authors concluded that: (1) participation techniques focusing on structure are more effective in producing actual decisions for the innovation, (2) attitudes such as readiness to change and perceived philos0phical congruity of the program to the hospital are related to decisions to change, and (3) pre-existing organizational characteristics such as overall participation and innovation attitudes were more predictive of adoption of the innovation than any of the brief experimental manipulations. Further analyses by Tornatzky et. al. (1980) suggested that organizational values and norms were more likely to persist as stable characteristics related to innovation adoption rather than processes such as participative decision making. These variables may be conceptualized as part of the underlying organizational climate (Schneider, 1975)- Siegel and Kaemmerer (1978) focused specifically on the organizational climate for innovation. These researchers administered a questionnaire to respondents in six traditional and two alternative high schools. The assumption was that the 17 climate in the alternative schools would be more supportive of innovation than the climate in the traditional schools. Significant differences between teachers and students in both types of schools were observed on the following five a priori specified dimensions: leadership, ownership, norms for diversity, continuous develOpment, and consistency. In all cases the teachers were higher on each of these subscales than the students. In addition, there was a main effect for type of school, with the alternative schools scoring significantly higher on all five subscales than the traditional schools. A factor analysis was performed on the responses of the traditional schools' members. This resulted in three factors. The first factor was labelled Support of Creativity. This was the extent to which members perceived the organization as supportive of their functioning independently and supportive of new ideas. The second factor was Tolerance for Differences, which reflected the members' perceptions of the organization's supportiveness and tolerance of diversity among its members. The final factor was Personal Commitment. This reflected the degree of personal commitment that members felt towards the organization. Alter~ native schools were significantly different from traditional schools on the first two of the above factors. This study high- lights organization member's perceptions in traditional and alternative schools. There are two explanations for these findings. Alternative and traditional schools may select staff with norms and values that match the norms and values of the i o .qulru .5 1)) schools. Alternatively, potential organizational members may self-select (Crites, 1969) into the organization closest to their personal norms and values. Either or both of these processes could have occurred. Another empirical study investigating the influence of organizational—level variables on innovation adoption in schools was conducted by Deal, Meyer, and Scott (1075). These research— ers performed a study involving 188 elementary schools in 34 dif- ferent districts. All principals were interviewed concerning current instructional practices and classroom organization. Two types of innovation were considered in this study, one an in- structional innovation and the other an organizational innova- tion. Instructional innovation was defined as the amount of in- structional differentiation existing in the reading curriculum. Organizational innovation was defined as the organization of teachers into small groups to teach reading. Archival data was collected at both the local school and district levels. Multiple regression was employed to assess the total amount of variance that was explained by all the variables. Normalized beta coeffi- cients were used to indicate the relative importance of each variable. Twenty-three percent of the variance was explained in both instructional and organizational innovation. The most im- portant variables in explaining instructional innovation were number of students in the district (negative), the amount of Open space in the school (positive), and the per student exPenditure in the district (positive). The most important variables in ex— plaining organizational innovation were the amount of open space in the school (positive), the proportion of total district admin- istrative staff in Special administrative positions (positive), and the per-student eXpenditure at the district level (positive). The authors concluded that . . . there is a dis- connected pattern of district and school influences on innovation at the classroom level" (p. 124). They also concluded that the pattern of relationships among district and school level vari- ables are different for instructional and organizational innova- tions. Although the authors conducted no statistical test on the presence of an interaction, the above conclusions suggest that an interaction is present among type of innovation and district and school level variables. At this point it is useful to summarize the major findings and research implications of these studies. The studies reviewed in this section have focused attention on organizational variables in the adoption of innovations. Hage and Aiken (1967) and Aiken and Hage (1971) pointed out the association among variables indicating a more organismic form of organization (high complexity, low formalization, less centralization) and the number of new practices started in a five year period. Unfortéately, questions about what organizational characteristics were associated with what type of new practices and questions about the direction of causality among organizational characteristics and practice change were not addressed in this research. Tornatzky et. al. (1980) demonstrated the impact of an organizational atmosphere favoring change or organizational climate on the adoption of an innovation. The more favorable the climate was toward change, the more likely that follow—up and adoption activities would occur. Siegel and Kaemmerer (1978) demonstrated the existence of differential amounts of these climate dimensions in traditional and alternative schools. In both these studies, the question of what caused this "innovative" type of climate was not addressed. Deal, Meyer, and Scott's (1975) results seemed to indicate an interaction among the type of school innovation (instructional or organizational) and the organizational properties associated with adoption. The innovation variable was not defined on a sufficiently detailed level to assess what organizational property related to the starting of a specific practice. Finally, with the exception of the Fairweather and Tornatzky work, these studies as a whole tend to relate macro—organizational prOperties to macro-definitions of innovation. Questions remain concerning the relationship of these macro-properties to the adoption of specific innovations or practices. Methodological Criticisms and Considerations Responding to the lack of consistent findings concerning both organizational prOperties and innovation attributes from one study to another, Downs and Mohr (1976) pointed out some conceptual problems and methodological pitfalls inherent in much of the previous research. They distinguished among primary and secondary attributes of innovations (and organizations). A to primary attribute of an innovation is an attribute upon which an innovation can be classified without reference to the adoptin organization. This is a property of the innovation that is invariant across all the organizations that may ultimately adopt the innovation. A secondary attribute of an innovation is an attribute whose classification depends upon the organization that is contemplating its adoption. These attributes ther'fore vary depending on the organization. Downs and Mohr (1976) listed four sources of instability across studies of innovation adoption. First, there could be variation in the primary attribute across studies. Although different investigators think they are studying the same att- ribute, the difference in the level of the attribute dictates a necessity for a different theory of innovation for each level of a primary attribute. Downs and Mohr argued that, to the extent that primary attributes do exist, there is no chance for a single unitary theory of innovation adOption. A second source of ins- tability is that the idea of a secondary attribute requires that statistical interactions among innovation attributes and organizational prOperties must be taken into account. Whether an attribute of the innovation or another factor leads to adoption depends on the level of some variable characterizing the organization. This is clearly an interactional hypothesis, and has been ignored by much past research (e.g., Deal, Meyer, & Scott, 1975). A third source of instability concerns the "ecological fallacy" of cross—level inference. Many studies use as a summary measure of innovation a variable such as "number of «J o , new practices" (e.g., Aiken & Hage, 1971; Hage & Aiken, 196 Hage & Dewar, 1973). This variable is not the same as the adop- *J tion of one new practice. The determ'nants of innovativeness in the aggregate may differ from the determinants of the adoption of a single innovation. When a correlation is found between two aggregated variables such as average perceived cost of innovations and total number of new practices, this does not imply a correlation on the individual level between perceived cost and adeption of the innovation. The fourth source of in— stability across studies is related to the third source and deals with the operationalization of the dependent variable of innovation. Downs and Mohr provided three different operation- alizations of this variable. The first was a binary yes/no definition based on the responses to questions dealing with whether the organization had adopted this particular innovation (or innovations). The second definition dealt with the time of adoption or when the organization first adopted the innovation. The third definition had to with the extent of implementation. These three definitions are clearly three conceptually different phengmgna and the determinants of one would seem unlikely to be the determinants of any of the other variables. However, authors often do not limit the generalizability of their findings by the definition of the dependent variable used. Downs and Mohr concluded with seven prescriptions for studying innovation based on their criticisms of the past m \JJ literature. The major prescription was to use what they called the Innova.ion Decision Design. This involves looking at each innovation in relation to each organization. For example, 10 innovations used, adopted, or considered by 100 organizations, yields a sample size of 1000. The unit of analysis is the innovation-organization link. Other prescriptions dealt with the use of the interactive model, not using a summary or aggregate measure of innovation, being aware of the meaning of the dependent variable chosen for study, and using multiple innovations within a study to look at variations in primary attributes. The reasoning and examples given by Downs and Mohr are very compelling. However, the implications of their arguments rule out any general theory of innovation across the content area of the organizations and the innovations under study, and should therefore be viewed critically. To begin with, there are problems with some of the premises of the Downs and Mohr (1976) argument. For instance, a question remains about the mere existence of primary attributes. Downs and Mohr use cost as an example of a primary attribute stating that: "For example, some findings of research into the determinants of high-cost innovations are generalizable only to other high—cost innovation" (p- 703). However whether an innovation is perceived as high—cost will depend on how much money the organization has to spend on the area to which the innovation pertains. In short, there do not seem to be any good examples of real-world "primary" innovation attributeS. In the second place, Downs and Mohr discussed the necessity of a different theory of innovation for each different category of primary attribute. They suggested studying each level of primary attribute separately. However, if this was the strategy, the rimary attribute could not correlate with any other variable given its invariance. A better strategy would seem to be using more than one category or level of a primary attribute, and to then experimentally study the effects (Fairweather et. al., 1974). In addition to the criticisms expressed by Downs and Mohr, there are a number of other methodological issues to be raised concerning the other studies reviewed above. For example, many of the studies used a very small sample of organizations. Hage and Aiken (1967), Aiken and Hage (1971), and Hage and Dewar (1973) all used a sample consisting of only 16 organizations. The study by Baldridge and Burnham (1975) used 20 schools in seven different school districts. To the extent that district- wide variables are important in adoption, the effective sample size was only seven. Siegel and Kaemmerer's (1978) sample con- sisted of eight different schools. Given these small samples, it is little wonder that somewhat inconsistent findings have emerged. However, the samples of Deal et. a1. (1975). Fairweather et. al. (1974), and Tornatzky et. al. (1980) consis— ted of over 100 schools or hospitals. Another criticism has been leveled at the number of innovations studied. In many of the studies reviewed in this ,.. "A .I paper, an aggregate measure of innovation was used (Aiken & Hage, 1971; Hage & Aiken, 1967; Hage & Dewar, 1973). The specific determinants of a single innovation cannot be inferred from these studies. Siegel and Kammerer (1978) did not incorporate a measure of innovation into their study. They assumed a priori that alternative schools were more innovative than traditional schools. Baldridge and Burnham (1975) and Deal et. al. (1975) used variable measures of innovation such that a practice defined as an innovation in one school might not be considered an innovation in another school. Only Fairweather et. al. (1974) and Tornatzky et. al. (1980) dealt with a single innovation. These differing definitions of the dependent variable could easily account for much of the inconsistency in the results across studies. For example, Baldridge and Burnham (1975) only studied extensive innovation with "real promise" that appeared well established and able to continue for a significant period of time. Therefore, the study severely curtailed the range of innovations studied by the selection of only those innovations that had become routinized or part of the standard practice of the organization (Yin, Quick, Bateman, & Marks, 1978). Consequently, their findings apply only to the adoption of innovations that ultimately become routinized. In summary, many methodological problems must be considered when studying the topic of innovation. Differences within and among studies in (1) the specificity of the measure of innovation, (2) the number and level (e.g. school versus dis— “ "m.“ m; n... 4. -... m- 'V—T'P‘ ’ - h-ank-D‘fi'iwmtu—n‘ ' trict) of organizations sampled, and (3) the criteria for sample inclusion for both innovations and organizations can cause vastly different findings. The Innovation Decision Design suggested by Downs and Mohr, which takes into account complex statistical interactions among variables would seem to offer promise in dealing with these issues. Exemplary Studies The studies reviewed below were identified as exemplary for the present research. These studies have applied the greatest amount of empirical attention to the methodological considerations discussed earlier. They also used large samples in general, and issues related to the dependent variables in each of these studies were carefully considered. Duchesneau, Cohen, and Dutton (1979) studied 50 firms in the footwear industry that were randomly selected from a sample stratified on the basis of geographic location, firm size, and type of shoe produced. Only firms above a minimum size standard and who were potential adopters of the innovations being studied were included in the random sample. Fifteen process and product innovations were studied. Presidents in each firm who were identified as the key decision makers were interviewed. In firms where the presidents were not identified as the key decision maker, the key decision maker was interviewed. In addition, presidents of each firm also completed a questionnaire. A sample of managers in each firm also completed a questionnaire. The response rates to the presidential and managerial questionnaires were 72% and 66% respectively. This was a very large scale study in terms of the number of variables studied. For this reason only certain findings will be summarized here. It is of special importance to note that these findings supported the Downs and Mohr (1976) interaction hypothesis in two respects. First, the effect of a given factor on adoption varied as a function of the innovation studied. The authors also considered definitions of innovation both in the macro sense (all 15 innovations) and in the micro sense (yes/no to the adoption of a single innovation). The findings again support Downs and Mohr in that factors predicting the adoption of innovations in the aggregate did not predict the adeption of innovations in the individual sense. Another major finding was that models including both economic variables (research and develOpm nt activity) and organizational variables (number of occupational specialities) were better predictors of the adoption of innovations than either class of these variables alone. When both economic variables and organizational variables were included in the analysis, the size of the firm had no incremental effect on the adoption of innovations. Size was found to be a proxy variable for: the presence of a larger body of professional and technically specialized managers, less reliance on traditional sources of information, presidents less bound by industry tradition, and the utilization of more sephisticated management techniques. One final finding of interest was the relationship between real time and retrospective data. It was found that there was substantial 1\) (D variation between the real time and retrospective values obtained for many variables. Correlations between the values of variables collected in 1975 and 1977 range from .28 to .98. Agreement on responses to two other questions was 86% and 69%. Based on their analysis, the authors concluded that retrospective analysis tends to overstate the values of variables compared to ective ’0' real time measurement and that the reliability of retros data is suspect. Another exemplary study was carried out by Bigoness and Perreault (1980). They used the data of Duchesneau et. al. (1979) to generate a multiple innovation criteria (counter to the recommendations of Downs and Mohr). Bigoness and Perreault specified three domains that are important to consider when designing studies of innovation. The first was the innovativeness domain, i.e., the adoption/non-adoption of a given innovation. The second was the content domain of the innovations sampled. This refers to the type of innovations, e.g., specific single-application innovations versus general innovations. The third was the reference domain, i.e., the comparison sample of the organizations being considered. That is, innovating organizations should be compared to organizations that did not ad0pt a specific innovation with respect to a variety of organizational and member characteristics. These authors argued that the generalizability of the findings of a study are limited by the representativeness of the content and reference domains. Bigoness and Perreault used multiple innovations scored in a 3 x0 dichotomous yes/no fashion to determine the innovativeness of firms in the footwear industry. The independent variables were the possession of an internal technical group and the presence of managers with technical expertise. Item analysis procedures were applied to 12 innovations to arrive at the final criterion measure, resulting in the deletion of six innovations with no variability. The final criterion measure thus included six innovations. The reliability of this measure (internal consistency) was .68, and the authors concluded from this coefficient that the innovations were homogeneous. Summarizing a few of these results, the presence of a technical engineering group was found to be significantly related to the adoption of innovations for firms that did not possess technical managerial eXperience; the technical expertise of the managers was related to innovativeness when firms did not possess a technical engineering group; and firms with both managers with technical eXpertise and a technical engineering group were not more innovative than firms with either of the above groups. A third exemplary study dealt more directly with the Directed DevelOpment (or RD&D) model of change. House, Kerins, and Steele (1972) studied the diffusion of centers for gifted children. Twelve hundred visitors to 20 demonstration centers were asked what they were doing as a result of their visit and their reasons for accepting or rejecting what they had seen two, four, six, or 12 months after their visit to the demonstration center. The authors found that: 29% of the visitors could supply a specific, concrete example of what they were doing as a result of the visit; and the shorter the time period between the visits and the administration of the questionnaire, the more concrete examples the teachers could supply. However, more important than the time interval between the visit and the questionnaire was the time of year in which teachers and administrators received the questionnaires. That is, more specific examples were garnered when teachers were questioned in the spring rather than in the fall. In addition, only 2% of the visitors adopted the program in toto. For administrators, follow-up help from the demonstration center and an administrative judgment of how well the program worked were the most important factors in accounting for changes as a result of the demonstration visit. The most important reasons for attempting the change for teachers were that: time spent would be well used, they were able to adopt parts of the program, administrators would accept change, enough facilities were available, and that cooperation from other teachers could be obtained. Interestingly, these reasons all dealt with how well the change fits into the teachers world; the subject area and grade level of the teacher was only of slight importance. Finally, it should be noted that this study asked innovation adopters directly why they adopted part or all of the program, rather than relying only on organizational or individual variables to explain the reasons for adoption. ‘v - -W... _.._..__‘ v.— 31 Implications of the Exemplary Studies These studies taken together have important implications for the present research. First, the Duchesneau et. al. (1979) study supports Downs and Mohr's (1976) argument concerning the use of aggregated versus single measures of innovation, yet the Bigoness and Perreault (1980) study supports the idea of using a multiple innovation criteria. Therefore, given the present "state of the art" of our understanding of innovation processes, it would seem that the ability to use both the micro (within innovation) and macro (across innovations) definitions of innovation should be built into research designed to study innovation. Second, the framework of reference and content domains (Bigoness & Perreault, 1980) is very useful to consider when sampling innovations and organizations. Finally, a focus on the perceived reasons .(House, Kerins, & Steele, 1972) for adoption would seem to offer promise in explicating factors important to organizational adeption decisions. Rationale for the Present Study and Research Questions The preceding literature review has pointed out the importance of various individual and organizational characteristics in the study of innovation in organizations. Individual values, participative decision-making, organizational climate, and centralization of control all have been related empirically to the adoption of innovations by organizations. In addition, methodological problems such as different levels of aggregation in the measure of innovation, small sample sizes, varying criteria for sample inclusion, and the failure to use statistical tests for interactions cast doubts on many of the findings in the existing literature. As a resolution to the inconsistency of this literature, it is suggested that the Q. 'J "'J H) (D '1 (D :3 C+ perceived reasons for the adeption decision across many innovations may provide a mediating variable between organizational characteristics and innovation adoption. The present research addresses the following questions: 1) What are the reasons that organizations adopt programs? Do these reasons differ as a function of the program that is under consideration? Given that most of the past research in this area has taken place in the field of education, are there differences between education as a field and other fields in terms of the reasons for adopting programs? 2) Do organizations which ultimately adopt a program differ from organizations which have never heard of the program in terms of increasing, decreasing, or stable numbers of clients, administrators, and front—line service providing staff? Are there differences on these variables as a function of either the program under consideration or field that the organization is in? 3) Do organizations that ultimately adopt a program have different patterns of participation in decision making than organizations which have never heard of the program? If so, are these differences a function of either the program or the field? CHAPTER II Sample Innovations. This study focused on organizational social innovations, defined broadly as changes in work, patterns of int- eraction, or practices and procedures within organizations. The reference domain (Bigoness & Perreault, 1980) or comparison group of organizations for this study was created with respect to two dimensions: use/non-use, and awareness/non-awareness of a given innovation at a given point in time. This creates a matrix (see Figure 1) of different organizations in each of the three cells with the fourth cell being not of interest in the present re- search. In the first cell are those organizations that are both aware of the innovation and currently using the innovation. Organizations falling in this cell are sustainers or adopters. The second cell contains those organizations that are aware of the innovation but are not currently using the innovation for one of two reasons. Either these organizations never implemented the 34 (1») C7) Figure 1 Reference Domain Categorization of Organizations Currently Not Using Currently Innovation Using Innovation Sustainers Aware Nonmadopters, Aware of OR OR . do + rs rni. t rs 'dis- Innovation A pte Te ’ na 0 ( continued innovation use} Univar °’ e ---------- Unaware Non-adopters of Innovation O‘\ \ \ innovation (called aware non-adOpters) or these organizations have stOpped using the innovation (called terminators). The third non-empty cell contains those organizations that have never heard of the innovation and for that reason are not using the innovation. The logically empty cell contains those organiza- tions that are unaware of the innovation but are currently using the innovation. Without looking at implementation issues (Hall & Loucks, 1976) it is impossible to determine whether the practices I f and methods that the organization is using in a given area are actually the innovation under study. Since this was a study of issues in the adeption of innovations, it was decided to ignore this cell. Since the field of education has been at the forefront of research on innovation adoption, it was decided to include a second field as well in this study. The field of criminal justice was chosen as the second field. The fields of education and criminal justice at the Federal level, are two of the most visible users of the RD&D policy model. Four programs were chosen for study from each of these two fields. Table 1 contains a brief description of each of the programs. In order to be included in the present study, programs had to meet the following four criteria: a) Validation -—A major criterion for the inclusion of an innovation in the study was a federally approved summative eval- uation or "validation" of the innovation prior to its dissemina- tion. This was necessary in order to insure that the innovations Table 1 Innovations Selected in the Present Study Education Help One Student to Succeed {HOSTSl--A diagnostic, prescriptive, tutorial reading program for children in grades 2-6. Tutors are community volun- teers and high school students. The program includes ”pulling out" stu— dents from their regular classes at least one—half hour per day. ECOS Training Institute (ECOS)~-A training program to help principals and teachensinfuse new content areas into existing curricula or add new con- tent areas. A major part is the formation of a committee composed of administrators, teachers, and students. Deals with all grade levels. Experience Based Career Education (EBCE)--This program provides experience outside of school at volunteer field sites for the student. Systematic career and interest exploration on the part of the student is also encour— aged. The development of an individualized learning plan for each student is carried out. Program concerns high school students. Focus Dissemination Project LEOCUS)--A "school within a school" for dis- affected junior and senior high school students. All students are required to participate in a group of 8-lO students and one leader (called Family). Students take at least one class in the Focus program. Classes in the Focus program involve individualized, self-paced instruction. Criminal Justice One Dayipne Trial (ODOT)--A jury management system that calls in a certain number of potential jurors per day. Potential jurors come in for that day and if not selected to serve in a trial have completed their obligation. Community Arbitration Project LCAP)—-Juvenile offenders are sent to a formal arbitration hearing run by the court intake division, rather than to courts. Juveniles have the specific consequences of their actions explained to them. Youths are then given a number of hours of informal supervision usually involving work in the community. Community Crime Prevention (SCCPP)—-This program is a three phase attack at residential burglary. This involves the setting up of a neighborhood block watch, property marking and inventory, and home security inspections. Pro-Release Center (MCPRC)--Involves the setting up of a residential facility separate from the prison. This facility should be in the community from which most of the inmates are drawn. Inmates are encouraged to work so that they will have a job when they are released. Counseling and social aware- ness instruction is also part of this program. \JJ 0} would not vary considerably as a function of their perceived le- gitimacy to potential adOpters, and to increase the likelihood that the process of disseminating the innovations closely approx— imated the RD&D model. Two federal programs currently in opera— tion require this validation prior to dissemination: the National Diffusion Network (NDN) of the Department of Education and the Exemplary Projects Program of the Department of Justice. To become a part of the NDN diffusion effort, a program must go through the Joint Dissemination Review Panel. This process en- tails the submission of evidence documenting the innovative pro- gram's effectiveness (Emrick et. al., 1977). Criminal Justice programs seeking exemplary project status go through a similar evaluation/application process and are evaluated on site (The National Institute Host Program Report, 1979)- AS a first step in selecting the specific programs to be investigated, written literature on all the NDN programs and the Exemplary Projects was obtained. Program descriptions were read and programs were chosen that met three additional criteria. b) Replications --In order to guarantee a sufficient sample size for analysis of the data, only programs that could reasonably guarantee 20-30 adOpters were chosen for study. This was ascertained through discussions with innovation developers, and NDN and Exemplary Projects Program administrators. c) Age of the innovation --In order to insure the ability to investigate implementation and routinization issues in later phases of the research, it was necessary to select innovations ’\_>l ub— that had been disseminated for at least three years. This allowed a sufficient length of time for adopters of two or more years to be operating programs that have become standard oper- ating procedures in their organizations. d) Organization—wide innovations --Since this study was primarily concerned with organizational influences on innova- tion adeption, a large amount of effort went into choosing inno— vations that were complex and likely to require an organizational decision to adopt. Two sub-criteria were of primary importance in this decision: the number of organizational units involved in the program, and the demands of the organization to interact with its environment. For example, if the innovation could be imple- mented by a single teacher or a single police officer without a large amount of interaction being required with community mem- bers, the innovation was not included in this study. Unit of analysis. The study of organizational innovations requires an appropriate definition of the adOpting "organization". In the field of education, following a number of discussions with NDN administrators, innovation developers, and other researchers concerning the appropriate decision unit, schools were chosen as the unit of analysis. Although school districts frequently have a good deal of influence on the ad0ption decision, they are generally far removed from crucial implementation decisions. Schools and districts are the prime example of what is called "loosely-coupled systems" (Weick, 1976). For example, in the present study, a number of 40 cases were found where district administrators claimed to have adOpted a program, when in fact no such program was in place. Consequently, the influence of the district was measured in this study, but the school remained the unit of analysis. In criminal justice, the decision concerning the appropriate unit of analysis was more difficult. In order to insure compara- bility across social policy areas, the organization was defined as the unit which implemented the innovation, unless the imple— menting unit was created solely to operate the newly adopted pro- gram (i.e., the organizational unit had to exist prior to the adOption decision in order to make the decision). During the course of the research, decisions on what to call various units within the criminal justice field were made with an eye to in- suring compatibility with the decision to treat the school as the adOpting organization. Random sampling of organizations within innovations. In order to allow greater generalizability of research findings, a random selection of organizations was undertaken within each in- novation. In the education area, a 3% random sample of all schools in the continental United States was generated from a source tape purchased from the National Center for Educational Statistics and Market Data Retrieval. Potential adopting schools were then randomly selected from this list for any of the four educational innovations. In the criminal justice area, a 3% ran- dom sample from all the appropriate organizational units was gen- erated for each of the following types of organizations: circuit and district courts, juvenile courts, police departments, and prisons. This sample was generated from a tape of organizations {'1 purchased from the United States Department of the census. Potential adopters of the four criminal justice programs were randomly selected from this source list. Purposive sampling of organizations. As a fail-safe, in case the random identification procedure was too costly in terms of time and effort, lists were obtained from other sources. The source for obtaining lists of potential adopters was the innova~ tion developers. These lists varied considerably in quality from one innovation to another. For example, some developers had very detailed lists of adOpters including data implemented and contact person at the adopting site. Other develOpers only had lists of peOple who had requested literature about the program. As a result, for some of the innovations it was difficult to identify a sufficient number of adopters from the developer supplied lists. In an attempt to alleviate this problem, various state planning agencies in criminal justice and state facilitators in education were contacted. Switching from random to purposive lists. After several weeks of data collection using the randomly generated lists of organizations, data was compiled concerning the utility of this method for identifying adapters, terminators, and aware non-adOpters. Combining the 115 organizations contacted across the eight innovations, the following breakdown existed within the following categories: terminators N=O (0%) aware non-adOpters N=4 (3.4%) unaware non—adopters N—lO3 (89.7%) The number of organizations contacted within each innovation ranged from 12 to l9 with a mean of approximately l4. If it were to be assumed that t.e true number of adopters and aware non~adopters in the pepulation was seven out of every 100 organizations contacted, then the number of contacts required to obtain a sample of 25 of each category would be approximately 357 per innovation. For this study, 2856 contacts would be required. Due to time and financial constraints, the random lists were abandoned as a means of identifying adopters in favor of the purposive lists obtained primarily from the innovation develOpers. The idea of comparing the various reference groups with respect to their status on an innovation was scaled down to compare just the adOpters with the unaware non-adepters. The random lists continued to be used in order to identity additional unaware non-adOpters. Respondents. One hundred and sixty adopters and 158 unaware a non—adepters were contacted over the phone. Table a shows the breakdown of adopters and unaware non-adopters by each of the eight programs. As can be seen from the table, somewhat more of the adopters were from the four educational programs and two more unaware non-adopters were from the criminal justice programs. The decision to treat schools as the "organization" led to J) (J) Table 2 Number of Adopters and Unaware Non-adopters by Program Unaware Adopters Non—adopters HOSTS 32 ‘ 18 ECOS 24 20 Education EBCE 28 20 FOCUS 25 20 ODOT l6 20 Criminal CAP 9 20 Justice SCCPP 18 20 MCPRC 8 20 Total 160 158 44 some problems. The framework for this study was erected with educational organizations in mind and was stretched to fit the situation in criminal justice. multiple schools are found within a single school district. It should be noted that in some cases multiple schools within a single district had adopted the same program. For the purposes of this research, only one school within the district was treated as a program adopter. The term "super-ordinate organization" was used to refer to districts in this study. Various actors at the district level were catego— rized into three different groups. Policy/Budget SUpereordinate organization (SO), which included members of the school board; Administrative 80, including various district—level administra- tors such as superintendents and assistant superintendents; Specialized SO staff, which included district heads of depart- ments such as Vocational Education, curriculum coordinators, and Personnel. On the school level, the various actors were categor- ized as follows: Organization (O) administrators, which included principals and assistant principals; Specialized 0 staff, in- cluding guidance counselors and school level department heads; 0 Front-Line staff or teachers. In addition a category was used to signify when the respondent had left the organization or the field. The above framework was used for the criminal justice prog- rams as well. The distinction between the O and the SO was much more difficult in criminal justice than it was in education. Some of the actors coded into the various categories include: Ls \ 7. county commisioners (SO Policy/Budget); city council, city v?” manager, director of public works, chiefs of he Department of Corrections, and Juvenile Service Administration heads (80 Admin- istrative); special project or grant coordinators, state level special consultants (SO specialized staff); judges, court admin- istrators, police chiefs, wardens, mayors, independent agency directors, and chief probation officers (O Administrators,; work release coordinators, lawyers or detectives (Specialized 0 staff); intake workers, court clerks, baliffs, probation officers, correction officers and patrolman (O Front-Line staff). Difficulties with this framework for the organizations in criminal justice were mainly concerned with defining the approp- riate super—ordinate organization. All decisions were made with an eye to keeping consistency with the situation in education. Table 3 shows a breakdown of the adopters by the program and job position. Across all the programs, the two major categories of respondents were 0 Administrators (n=71, 44%) and O Front-Line staff (n=37, 23%). Table 4 shows the same breakdowns for unawre non-adopters. Across all the programs the two major categories of respondents were again 0 Administrators (n=90, 58%) and 0 Front-Line staff (n=38, 25%). The job positions in Tables 3 and 4 were content coded after all the data had been collected. The number of unaware non-adopters in Table 4 is not the same as the number presented in Table 2 due to missing or incomplete data. m AoVso AoVa AoVso AoVs AmVso.ooA AoVso AoVao seas: wA AAVaa.m AeVem.mm AaVeo AoVso AAAVaA.Ae AoVao AoVao sauna a AoVs ANVaN.mN AoVao AoVao AAVaw.aA AoVao AoVao aau 0A ANVE..NA AmVsm.NA AAVsN.s AcVa AAAVaw.we AoVao AoVs cogs in mm AlV a. e AAVso.am AAVso.wN AmVso.NA AeVao.sN AAVso.e 8 Va meson mm AmV A o. AeVam.sA AAAVam.am ANVaA.A AAVao. mm AAVsa.m AoVao mums em AAVaN.s AAV N. am AmVam.NA ANVsm.m : Vaw me AoVso AoVao moon mm AoVa AaVaA.wN AeVaw. mA AmVaa.mA 8 %V ANVam.e 8 Va memo: .Hmmmw eAmAu ccsem ceasm.u ccaum.qmi mispaAWmAcAsea m>wpa+wmA=A2e< pmmesmnn EarmOLa we» also econJW=oru ewN_AmAumam emNAAaAomam 0 ca AsoaAoa om o Auxwe as» cw emcmmpaxm meowpmw>mnnnAnmlwmAcA5e< o Apxoe we“ as umcwmpaxm mcoAueA>ana.05). Scaling of the reasons for adoption. The reasons for adop— tion were scaled using rational and empirical methods (Jackson, 1971). The first step was the elimination of items that had less than a 10% endorsement frequency across the 160 adopters. The five point scale used for the Adoption Decision Questionnaire was actually two separate scales. The negative end of the scale was designed specifically to be used with aware non-adopters. The negative end of the scale refers to factors that were reasons not to adOpt. Since the reference group of aware non—adopters with respect to the innovation was so hard to locate given either the random or purposive sampling strategy coupled with resource con- straints, it was decided not to pursue identification of these organizations. In the scaling of the reasons for adoption, only adopters were considered. Because of the discontinuous nature of the scale, only those reasons which were coded on the scale points: not mentioned or unimportant, less important reasons to ad0pt, and important reasons to adept, were considered in scaling. Zero-order correlations were then computed for all items. Based on item content, coding protocol, and negative cor— 50 ”I Table 7 Means and Standard Deviations of the Number of Reasons Given by Each of the Programs Standard Program Mean Deviation HOSTS l7.84 4.90 ECOS 15.67 4.51 EBCE l7.07 4.ll FOCUS l5.00 4.l2 DDOT l4.88 4.27 CAP l4.ll 5.2l SCCPP 15.55 5.ll MCPRC l3.50 4.24 60 relations, some items addressing very specific aspects of the same general construct were combined into one item. For example, combining of items took place on two items referring to the per" ceived eXpense of the program. One item referred to the program being inexpensive due to grant support brought in by the program while the other item referred to the program being inexpensive 1 without any referent. sle CO rly, both these items are getting at the perceived expense of the program to the organization. However, due to coding procedures these items would never both be coded for the same organization. Therefore there was a negative correlation between these two items. For the purposes of buil— ding reliable scales, in instances such as the above, items were added together and the mean'of the two items was taken to repre— sent the constructs. Given the fact that some of the specific items had to be combined into a single item, the fairly low inter—rater agree- ment was less of a concern than it might ordinarily have been. A good deal of the disagreement among coders consisted of differences in coding on items that were later combined. Given both the combining of some of the items and the conservative nature of the inter-rater agreement calculation, the reliability obtained was considered adequate. Figure 2 shows the scale names and item composition of each of the four scales that resulted from the process outlined above. The first scale was called Expense and Financial Support and its items refer to both the perceived expense of the program and the ADQ Item #* 67 68 115 119 123 52 54 55 59 Figure 2 tems in Scales and Scale Names Scale Name Expense and Finan- cial Support Changes in Roles and Role Re1a- tionships Item Program would be relatively ineXpen— sive for the organization. Program would be relatively inexpen- sive for the organization due to grant support brought in by the program. Federal financial support was avail- able. State financial support was avail- able. Local financial support was avail- able. Program would involve large change in the organization's client roles or role behaviors. Program would require a large change in the organization‘s member roles or role behaviors. Program would involve a large change in the role relationships (inter- action) between any organization actors. Program would improVe the interper- sonal relationships in the organiza- tion. *The following items were combined due to coding procedures for scaling purposes (see text): 67 and 68; 115, 119 and 123; 52 and 54. AinItem #* 43 39 41 107 111 72 74 76 78 8O 62 Figure 2 (cont.) Items in Scales and Scale Names Scale Name ggpected Smooth Implementation Sumac: Item Program would increase the efficiency of the organization (broadly inter- preted). Program would not take a lot of staff time to execute. Program would be likely to function smoothly in the organization (WORK- ABLE; organization member-organiza- tion, i.e., administrative). Program would be likely to function smoothly in the organization (WORK- ABLE; organization member-client, i.e., services process). Appropriate materials for the program were available before adoption. ApprOpriate facilities were available for the program before adoption. Members of the policy super-ordinate organization were supportive of the program. Members of the administrative super- ordinate organization were supportive of the program. Administrators in the organization were supportive of the program. Specialized super-ordinate organization staff were supportive of the program. Specialized organizational staff directly involved with the program’s implementation were supportive of the program. Front-line staff (potentially) directly involved with the program's implemen- tation were supportive of the program. *The following items were combined due to coding procedures for scaling purposes (see text): 39 and 41; 107 and 111. perceived availability of financial support. The second scale 1 was called hanges in Roles and Role Relationships and its items (— refer to changes in either client or member roles and inter- actions or interpersonal relationships among organizational actors. The third scale was called EXpected Smooth Implemen— tation and had items referring to the workability of the program, lack of staff time to execute the program, expecrations that the program would increase the efficiency of the organization, and the belief that the organization posessed apprOpriate facilities or materials to carry out the program before adoption. The fourth scale was called Support and had items referring to sup- port from various actors at either the super-ordinate organiza- tion or organizational level. Table 8 shows the correlations among the scales and the in- ternal consistency (alpha) of each of the scales. It should be noted that no internal consistency estimate was carried out for the Support scale. Due to the coding protocol, multiple actors were sometimes put into the same item. Also, there is little reason to expect that what one level of the organization supports another level would also support. Given the saliency in the literature of support as a factor in the adoption process (e.g., Berman & McGlaughlin, 1978) this scale was retained on strictly a priori, rational grounds. A look at Table 8 shows that all the internal consistency estimates are .50 or above. Table 8 also shows that three out of the six possible correlation among the scales were significantly different from zero. Table 8 Scale Correlations of the Reasons for Adoption Changes in Roles Expense and and Role Expected Smooth Financial Support Relationships Implementation Support Expense and Financial (~54)* Support Changes in Roles and Role .23*** (.54) Relationships Expected Smooth Implementation .20*** .l7*** (.51) Support .07 .10 -.09 ** *Diagonals are coefficient alphas **Rational Scale (no a computed) ***p < .05 Differences among programs on the scales. Table 9 shows the scale means and standard deviations for each of the four scales on by program. Due to significant correlations among the scale , a multivariate analysis of variance was performed. The scales showed significant multivariate heterogeneity of variance. A square root transformation on each of the scales was carried out to attempt to control for the heterogeneity since the hetero_ geneity could have been due to the non-orthogonality or unbal— anced nature of the design. Significant heterogeneity of variance still existed on the Change in Roles and Role Relation- ships scale and the Expected Smooth Implementation scale. The multivariate analysis of variance did reveal significant dif- ferences among the scales by program (Wilk's F approximation= 17.26 p<.OOOOi). On this basis, univariate analyses of variance were performed. Table 10 shows the results of the univariate analyses of variance. As can be seen, all the scales differed significantly as a function of the program. In order to determine whether these differences occurred as a function of whether the program was in education or criminal justice, planned contrast analyses of variance were performed. No significant differences were found beween areas on the Expense and Financial Support scale (T (152) = .82, p>.05), the Changes in Roles and Role Relationships scale (T (152) = .75, p>.05), and the Support scale (T (152) = 1.09, p>.09), however on the Expected Smooth Implementation scale education and criminal justice were significantly different from 66 Table 9 Scale Item Means and Standard Deviations by Program* SCALE Changes in Roles Expense and and Role Expected Smooth Financial Support Relationships Implementation Support Ease-i .52 M66 r- §2 than .529. Mesa 5.2 HOSTS .63 .22 .45 .50 .42 .40 .42 .36 £005 .45 .32 .l7 .28 .l3 .l6 .63 .29 £805 .46 .24 .lO .2l .04 .08 .77 .34 FOCUS .35 .24 .l4 .2l .l9 .24 .53 .33 000T .52 .28 .08 .23 .43 .48 .54 .25 CAP .42 .30 .l3 l6 .56 .39 .50 .32 SCCPP .5l .24 .32 .50 .23 .29 .59 .32 MCPRC .28 .30 .l5 .l6 .3l .38 .46 .25 *0 = not a factor in adoption decision, l = somewhat of a factor in the adoption decision, 2 = strong factor in adoption decision 67 PO. v as mo_. 00. F_. No. Nm_ eoeem . - . . . . . . . . . . . Seem we emm N on em «me o mm P” ewe m me mo em” m mm N -oee meosq N3 d m: N3 m m: N3 m mi N3 m m: wt mugzom .mummmmw Imeeuepcm2m_maH mmwgmcowuepmm peoamwm Feeucecwd :poosm umeomaxm mpom use new mmcmaxm mmpom cw mmocecu anemone an mpeum peoagsm use .m_eom :owpeecmEmpasH gpoosm emuumaxm .mFeum ma_;mcowpepmm m_om see mmpom :w mmmcezu .m—eom ueoqaam _ewocecwd new mmcmgxm co mo:e_ee> co mwmxpmc< mumwee>eca eo mopsmmm OH wreck one another KT (152) = 3.34, p(.OOl) with criminal justice being significantly higher than education. The expected ease of the implementation was more of a reason to adept the program for criminal justice organizations than for educational organiza- tions. (D F- 0 <1 (n [,4 :1 (D (‘f‘ (D "i O I It should be noted that on the univariat geneity of variance was found. Both the Changes in holes and Role Relationships scale (Bartlett's Box F = 7.21, p<.e01) and the Expected Smooth Implementation scale (Bartlett's Box F = 10.64, p<.001) had significaltly heterogeneous variances. The Expense and Financial Support scale and the Support scale had homogeneous variances. Post-hoe Scheffe pairwise comparisons among programs were calculated for all scales. There were no significant pairwise differences among programs on either the Expense and Financial Support scale or the Changes in Roles and Role Relationships scale. Tables 11 and 12 show the significant pairwise compari- sons among programs. As can be seen in Table 11, the HOSTS prog- ram is significantly higher than the EBCE program, and the ODOT and CAP programs were significantly higher than the EBCE program on the EXpected Smooth Implementation Scale. Table 12 reveals only one significant difference among the pairs of programs on the Support scale; EBCE had significantly higher support than the HOSTS program. In other words, the eXpected ease of imple- mentation was more of a reason to adopt the HOSTS, ODOT, and CAP programs than it was to adOpt the EBCE program. Support from 69 Table 11 Scheffé Procedure for Comparisons Among All Pairs of Means on Expected Smooth Implementation Scale HOSTS ECOS EBCE FOCUS ODOT CAP SCCPP MCPRC HOSTS N5 + NS NS NS NS NS ECOS NS NS NS NS NS NS EBCE NS — ~ NS NS FOCUS NS NS NS NS 000T NS NS NS CAP NS NS SCCPP . NS MCPRC NS = Non-significant + ll Row program significantly higher on scale than column program (p < .05). Row program significantly lower on scale than column program (p < .05). 70 Table 12 Scheffé Procedure for Comparisons Along All Pairs of Means on Support Scale HOSTS ECOS EBCE FOCUS ODOT CAP SCCPP MCPRC HOSTS NS - NS NS NS NS NS ECOS NS NS NS NS NS NS NS EBCE NS NS NS NS NS FOCUS NS NS NS NS ODOT NS NS NS CAP NS NS SCCPP NS MCPRC NS = Non-significant + I! Row program significantly higher on scale than column program (p < .05). Row program significantly lower on scale than column program (p < .05). 71 organizational members was more of a reason to adOpt the EBCE program than it was to adOpt the HOSTS program. Organizational Demographics and Environments The Organizational Profile asked respondents (both adepters and unaware non-adopters) about the number of clients (people pro- cessed by the organization), number of administrative staff, and number of front—line staff in both the organization and the super— ordinate organization. In addition, respondents were asked for each of the above, whether these client or staffing numbers had been in— creasing, decreasing, or stable at the time of the adoption decision. Only analyses for the increasing, decreasing, and stable information are presented here. This was done for two reasons. First, as discussed above, there were often conceptual problems in trying to find a district-type organization for crim- inal justice organizations. This led to adopters and unaware non—adepters of the same program having different super-ordinate organizations. Given this situation it is clear that the abso- lute sizes of the variables will differ as a function of the program. Second, it is conceptually clear that schools are dif- ferent from courts, juvenile service agencies, police depart- ments, and prisons (although some students may argue on the last institution). Therefore, differences in the size of these orga- nizations is of little interest. Instead, whether these client and staff numbers were changing, the direction of the change, or the stability of the variables is a much less program-specific issue and therefore of more interest in the present research. \t no These demographic variables reflect environmental pressures towards the types of changes that the eight programs bring. r" The first question of in erest concerning stability and change in client and staff numbers concerns whether the sample of adopting organizations is significantly different from the sample of unaware non-adopters. Table 13 gives the percentage breakdown by program and adoption status for the three variables. In order to ascertain whether the adopting organizations were different from the unaware non—adopting organizations, three analyses of variance were performed. For these analyses, increasing was coded 1, stable coded 0, decreasing coded -1. The analyses of variance treated program as one independent variable and adoption status as a second independent variable. Table 14 shows the an- alyses of variance of increasing, decreasing, and stable numbers of clients, administrative staff, and front-line staff respec— tivelv, as a function of program and adoption status. In all three analyses, these variables differed as a function of program and did not differ as a function of adoption status. In all three analyses there was no interaction between program and adop- tion status. However, in terms of the numbers of clients and front-line staff, the main effect for adoption status approached statistical significance (F (1,294) = 2.86, p<.09; F (1,294) - 3.23, p<.07, respectively). Planned contrast analyses of variance comparing education to criminal justice were conducted for each of the three variables. All three contrasts were significant: the number of clients is AN_V moo noes Ao_v Aoev Ago flops Aopo Ao_o Ao_v Ae_v A_No flops Ao_v Amos Ammo e.eeem see sea son so“ ooo son soo soo soo so“ so“ see soo one one son eem E E E E E E E E E E E E E E E E . r so so_ so so so a so so soo soo sow ep_ so one sow so ooemeeeeeo e>eee“wewmmflmu goo A_o Ame goo “so loo A_o loo goo A_o goo gov Ame Aeo goo gov sow oo_ o o eg_ ooo saw so o_m so so so_ s__ sme one a__ so_ ooemoeeeoo Aoo moo on goo Aoev Ase Ae_v Amo Ame on goo Am_v Ao_o goo Ago goes eeseem one som ems eoo ems see eoo e_o soo soo ewe New sow aom som see . em Ame loo Ame loo ANS lee loo loo lm_o ,m_o Ago loo loo lee Ago Ao_o -ees soc ems so, s__ s_e o__ no so soo ego ago sow ems soo soo e_o ooemeeeeeo emwoewmmuu 3 E E E E E E E E E E E E E E E E eases ./ soo soo eoe soo ago see s__ so_ so_ see s_o sow so soo soo soo . Aoo A_o goo on loo All moo Ago goo Aeo Ame goo Ame Aeo “so goes epoeem so_ so, new eon eoe se_ see see ems so, new soo sow as, saw s_o Amv flow Amy Amv A_V ANV Amv on “OF AeFV Ao_v AFFV Aopv AmFV Aorv Aopv so_ so so. so_ so now see so soo soo soo sPo now new new spa oeemeeeeeo eomwmmyhm Am_o Ago Ao_o Aoo fi_ev Aoo Aoo Aoo A_o Ago Ame on Ame Ago moo AN_V ooaeeboH son soo soo soo soo ego soe see so new soo soo soo soo soo soo . oz o <2 o oz a <2 o <2 a oz o (2 o <2 < oeeoz eeoom loo eooo moooe moom mooo memo: v coopm moopupcoem use .meeum o>euoepmwcwse< .mpchPQ so smasoz m_ooem use .mcememgumc .mcemoweucfi m>oz on: Aeuoaumwcwan< so sensoz so consoz so amass: .ocemomeome .mcwmomcocev emoum ocesupcoed use .eemum m>wumnum_cw5e< .mocmwpo so memosoz co wocmweo> so moms—oc< eo mapsmmm «H spams significantly more likely to be stable or decreasing in education compared to criminal justice (T (151) = 5.01, p<.0001), educa- tional organizations are more likely to exhibit decreasing or stable numbers of administrative staff than criminal justice or- ganizations (T (151) = 1.81, p<.07), and educational organiza- tions are more likely to show stable or decreasing patterns of the number of front—line staff compared to criminal justice orga— nizations (T (151) = 3.41, p<.001). The Scheffe post-hoe pair— wise comparison procedure was performed for all three variableS. There were no significant pairwise differences among programs on any of the three variables (p>.05). Participation in Decision Making The number of different levels of organizational and extra— ‘organizational actors participating both in the adoption decision (ad0pters) and in decision making in general (adopters and unaware non-adopters) were analyzed to see if program or adoption status differences existed. These questions were asked on a four—point scale ranging from not mentioned or no influence to a great deal of influence. The minimal, moderate, great deal of influence distinction was collapsed into one category. The scale then became a binary no influence/influence scale. In addition to collapsing across the amount of influence dimension, collapsing across the various levels of the actors was also carried out. Given the discussion above concerning the problems in finding and defining an appropriate super-ordinate organization and the fact that the measure was devised to tap the differentiation in educational organizations, it was decided to collapse the various levels of the actors irto three more global levels: influence from tne super-ordinate level, influence 3'“ C. L 0“: the organizational level, and influence from outside the organi— zation. Tables 15 and 16 show both the original and collapsed by level program means and standard deviations on both the number of levels participating in the adoption decision and the number 0 levels participating in decisions in general. Table 17 shows the analysis of variance for the number of levels involved in deci~ sion making in general by both the program and the adoption status. Again, there is a main effect for program, no main effect for adoption status, and no interaction between adoption status and program. Table 18 shows the analysis of variance of the number of levels involved in the adoption decision by prog- ram. Again, the main effect for program is significant. A planned-contrast analysis of variance was performed on each of the above variables to see if education significantly differed from criminal justice. Educational organizations have significantly more levels involved in both general decision mak— ing (r (152) = 4.67, p<.0001) and in the adoption decision (r (152) = 5.92, p<.OOOl) than criminal justice adopters. The Scheffe procedure for post-hoe pairwise comparisons revealed no significant differences among any pair of programs on these two variables (p>.05). A third variable was formed by subtracting the number of levels involved in decision making in general from the number of 01 Table 1 Means and Standard Deviations for Adopters on Number of Levels Participating in the Adoption Decision by Program Uncollapsed and Collapsed to Control for Possible Bias Original Collapsed Mean SD lean SD _ 1 HOSTS 3.4l l.39 2.25 .76 l 1 E608 3.63 l.44 2.37 .65 EBCE 4.2l l.8l 2.36 .83 FOCUS 3.48 l.23 2.04 .68 000T 2.56 l.09 l.75 .68 CAP 3.ll l.69 2.00 .7l SCCPP 3.ll l.18 2.l7 .79 MCPRC 1.63 .74 1.25 .47 Means and Standard Deviations for Adapters on Number of Levels Participating in Decisions Table 16 in General by Program Uncollapsed and Collapsed HOSTS ECOS EBCE FOCUS ODOT CAP SCCPP MCPRC Original Collapsed Mean -0 Mean SD 3.28 .40 2.16 .85 3.88 .36 2.50 .59 3.96 .64 2.25 .70 4.40 .22 2.32 .63 2.38 .09 1.56 .73 2.57 .98 l.44 .13 3.06 '.ll 2.00 .77 2.75 .04 l.75 .7l Table 17 Results of Analysis of Variance of Collapsed Number of Levels Involved in Decision Making in General by Program and Adoption Status for both Adapters and Non-adopters 29.9392 at. 55. L5. E s__ Program 7 32.20 4.60 9.92* .16 Adoption Status l .36 .36 .78 Program by Adoption Status 7 2 77 .40 85 Error 302 140.10 .46 2.34 Total 317 175.43 *p < .01 Table 18 Results of Analysis of Variance of Collapsed Number of Levels Involved in the Adoption Decision by Program for Adopters Source. 9.: i8. as f. .23: Program 7 12.23 1.75 3.31* .10 Error 152 80.01 .53 Total 159 92.24 *p < .01 CO 5.. J levels involved in the adoption decision. This variable repre- sents whether more or less levels were involved in the adoption decision than are involved in decision making in general. Table 19 shows the means and standard deviations on this variable by program. Table 20 shows the analysis of variance on this vari— "5 able by program. The e is no significant effect by program on this variable. Correlations for adopters between the four scales, number of levels involved in both the adoption decision and decision making in general, the difference in the number of levels involved in these two decisions, and the length of time the program has been in use at the organization, and the demographic variables are shown in Table 21. The ability of the demographic variables to eXplain programmatic differences on the scales would, from this analysis appear to be limited. Other Analyses to Egplicate Program Differences Two additional variables were eXplored as possible expla— nations for program differences on the scales. The first of these variables was the length of time that the program had been used. It was felt that the time frame in which the program was started might effect the reasons for adoption. Table 22 shows the analysis of variance on the length of time the programs had been used by the organization. Organizations did differ as a function of the length of time they had been using the program. A plannned con— trast analysis of variance revealed that educational programs had been in use significantly longer than criminal justice programs Table 19 Means and Standard Deviations for Adopters on Number of Levels Involved in the Adoption — Number of Levels Involved in Decisions in General Collapsed and Uncollapsed Original Collapsed Mean SD Mean SD HOSTS ~2.44 1.70 .09 .93 £608 -3.33 1.66 «.13 .74 EBCE -3.25 1.96 .11 .88 FOCUS ~3.76 1.51 ~.28 .74 000T ~1.63 1.45 .19 .66 CAP -.56 2.13 .56 1.13 SCCPP -2.22 1.26 .17 .99 MCPRC -2.38 .52 —.50 .53 83 Table 20 Results of Analysis of Variance of Collapsed Number of Levels Participating in the Adoption Decishni- Number of Levels \J I Participating in Decisions in General by Program forAdoption fair; 9:: is in 5 .3. Program 7 8.72 ..25 1.75 NS Error 152 108.22 .71 159 116.94 Total 84 Table 21 Correlations Among Four Scales, Number of Levels Participating in Both the Adoption Decision and Decisions in General (Collapsed), Number of Levels in Adoption Decision -- Number of Levels in General (Collapsed), Length of Time Program Has Been in Use in the Organization, and Demographic Variables for Adopters Changes Expense and in Roles Expected Financial and Role Smooth Support Relationships Implementation Support Length of Time Program Used In -.10 ~.16** ~.1l .18** Organization Number of Levels Involved In . Adoption .15 .08 .09 .24** Decision Number of Levels I"V°lved 1” -.14 .01 ~.11 .15 Decisions in General (Number of Levels In Adoption Decision -- .26** .06 .19** .08 Number of Levels In General) Number of C1ients* -.06 ¢.13 .09 -.11 Number of Adminis- trative Staff* ’03 '02 ’05 '02 Number °f Front‘ .02 .01 -.003 -.04 Line Staff* *Increasing coded 1, stable coded 0, decreasing coded -1. **p < .05 er Table Results of Analysis of Variance of How Long the Program Has Been Used by Program Source. if. as as. f. ; Program 7 88.27 l2.6l 6.76* .2l Error l52 283.6l l.87 Total l59 37l.88 *p < .Ol 5 shows the results of the i\3 (T (152) = 3.61, p<.OOl). Table Scheffe pairwise comparison procedure. It can be seen from the table that the E803 program has been, on the average, in use sig— nificantly longer than the HOSTS, ODOT, and MCPRC programs. Given that differences existed in the length of time that the programs had been in u-e the question of differences on the I 9 scales as a function of the length of time became important. To answer these questions, the medirn length of time that each orga- nization had been using the program was assessed for each prog~ ram. Each organization was assigned a score of 1 or 2 based on how long they had been using the program compared to other adop- ters of the same program. For example, an organization below the median of the HOSTS program that had adopted the HOSTS program, was assigned a score of 1. Length of time that the organization had been using the program thus became a two-factor variable; long or short period of time. Diffe-ences in the scales were then assessed as a function of both age and the program. Table 24 shows these analyses of variance. As can be seen from the tables, on none of the scales is there a main effect or inter- action involving age of the program. However, there are two sig- nificant correlations involving age of the programs and the scales. For the correlational analysis, age of the program was not Split by median age of the program. The second explanatory variable explored was the respondents job position following House at. al. (1972). It was thought that if the reasons for adOption were indeed reflecting the idiosyn- HOSTS ECOS EBCE FOCUS ODOT CAP SCCPP MCPRC Table 23 Scheffe Procedure for Comparisons Among All Pairs of Means on How Long the Program Has Been in Use HOSTS ECOS EBCE FOCUS ODOT CAP SCCPP MCPRC - NS NS NS NS NS NS NS NS + NS NS + NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS Non-significant + II Row program has been in use significantly longer than column program (p < .05). Row program has been in use significantly shorter than column program (p < .05). S. v as mm PP. mo. N_. no. qu coccm mm< am. so. RN._ N_. mm. mo. Na._ N_. a an Eaeooca oo._ PP. FF. po. mm. co. oo. oo. _ mm< mo. amo.m om. m_. «m_.o an. .F. aeo.m me. mo. a~m.m mm. N Emcmoca we a m: N3 a m: N3 a m: m3 m m: we monsom Mummmmw cowueucmEofimam mmwzmcowpmpmm pcoamsm _omo:mcwm :wooam oopooaxm mpom oco use mmcmaxw mm_om cw mmmcocu mm< use Eogmoca An w_wom ucoaasm uco .oFoom cowpogeoao_geH :oOOEm cooomaxm .mpmom mgwcmcowueFom mpom new mm_ox cw mmmcogu .m_mom pcooazm Fowocmcwm new mmcmaxm co mucowcm> mo mpmxpmc< mo mu_:mmm em mES 89 cratic perceptions and motives of the respondent, there might be }.__0 significant differences on the scales as a function of whetier . H the respondent was an administrator or a front-line staff memo (D 1‘ L I 0) It was thought that administrators might have different reason to adept than fronteline staff. Accordingly, respondents were collapsed into two rough categories; administrative and front— A line staff. As c-n be seen from Table 3, mos ri- \ of the respondents in this study were not front—line staff members. This collapsing must be regarded as fairly rough, since in some cases a Special— ized 0 Staff will actually be involved in service provision. Caveats such as this make it difficult to be sure of the scheme used for collapsing. All non front-line staff were collapsed into a single category. Table 25 shows the scale means, standard deviations, and the results of the multivariate analysis of vari- ance of the four scale means by the respondent's job position. There were no significant differences on the scales as a function of the respondent's job position. [‘0 Table 5 Scale Average Item Means, Variances, and Multivariate Analysis of Variance for Respondents‘ Job Positions* Administrative Front-Line Staff Staff . C. l Expense and Mean '4U37 «4910 1 Financial Support SD 2760 2741 Changes in Roles Mean .2l30 .2387 and Role Relationships 50 .3445 .4130 Expected Smooth Mean .2794 .l926 Implementat1on SD .3685 .2255 Mean .5896 .4775 Support SD .3406 .3l7l n=ll5 n=37 n=l52 Nilks F approximation (4,l47) = l.70 ns. *0 = not a factor in adoption decision, l = somewhat of a factor in the adoption decision, 2 = strong factor in adoption decision. DISCUSSION Four different reasons for adoption were formed into scales: eXpense and financial support, changes in roles and role rela- tionships, eXpected smooth implementation, and support from var— ious organizational members. Differences between the fields of criminal justice and education were found on the expected smooth implementation scale. Criminal justice organizations were more likely to cite expected smooth implementation of the program as a reason to adopt the program. This finding can be explained in a number of ways. First, since much of the research on innovations has been carried out in the educational area, and since "innova- tiveness" is a characteristic that we want organizations to pos- sess (Downs & Mohr, 1976), schools may be more likely to want to appear in a good light. That is, although ease of implementation may be a factor in education, respondents may have wanted to ap- pear in a good light and may not admit to considering the diffi- culty or ease of implementation. Second, changes in the opera- ting system in criminal justice organizations could be more dif- 91 ficult than changes in education. Upon reflection, differences between the two social policy areas in the difficulty in making organizational changes would seem to make a great deal of sense. Criminal justice organizations have an "Open house" quality about them. That is, public attention is much more focused on the failures and successes of criminal justice organizations than it is on the successes and failures of educational organizations. This could be due to the fact that as citezens, we are much more likely to come in contact with criminal justice organizations than educational organizations. In our contacts with the educa— tional system, we are much more likely to see a formal open house view of the activities of the system. Open houses in education, do not show us the regular interactions of the members of the system. On the other hand, interactions with the criminal jus- tice system are more likely to show us these organizations as they typically function. In view of both the amount of public attention to criminal justice organizations, and the difficulty these organizations have in putting on a different appearance from the reality, it is not surprising that the eXpected smooth- ness of the implementation would be more of a reason to ad0pt innovations for criminal justice organizations than educational organizations. For criminal justice organizations, making a change could be a much more difficult process involving numerous struggles involving organizational politics. For this reason, criminal justice organizations must take account of how easy the change will be prior to undertaking the change. Significant pairwise difference among programs were also (0 found on the EXpected Smooth Implementation scale. The HOSTS, ODOT, and CAP programs were all significantly higher on this scale than the EBCE program. All three of these programs would seem to be much easier to implement than the EBCE program whic' involves the student spending significant periods of time outside of the school. HOSTS mainly involves a thorough systemization of what might be an already existing Title I program and might not involve the changing of any organizational rules or procedures. ODOT involves improving on what the organization already does (process jurors) by spreading out the obligation more equitably across the pOpulation and keeping most jurors for a much shorter period of time. The major part of the CAP program is deciding that juvenile offenders who previously had been left alone, should have some sort of nearing concerning their offense. On content grounds then, EBCE would seem to be a harder program to implement. Conclusions such as the above should be understood with the following caution in mind. The present investigation was focused upon the adoption decision. At no point in the study was the content of the eight programs investigated systematically as a factor in the adoption decision. The present study has con- sciously taken an adoption perspective (Berman & McLaughlin, 1974), and has ignored implementation issues concerning the con- tent of the program. One significant pairwise difference was found on the Support scale. EBCE adOpters mentioned support as a reason for adoption more often than HOSTS adOpters. The comments above concerning ease of implementation could explain this difference. That is, EBCE as a program is harder to implement than HOSTS, therefore support from various quarters is much more of a necessity in implementing EBCE than in implementing HOSTS. In other words, it could be that support must be present prior to implementation for EBCE and not for the other programs. The analyses of the organizational demographics and environ- ments indicated that adopters and unaware non-adopters were not significantly different. These analyses would seem to indicate that the adopters as a whole were representative of the pOpula- tion of organizations in terms of these variables. The finding that there were differences on these variables as a function of the area that the program was in, should come as little surprise. Educational facilities and staffing came about as a reaction to the so-called baby boom. Given the well publicized decline in the birth rate, educational organizations face a much more uncer- tain environment than criminal justice organizations. If there is a trend, it would be toward decreasing the staff size at schools due to decreasing student enrollment. Criminal justice organizations face a much rosier future given increases in crime rate, litigation, and inmate pOpulation. The fact that client and staff sizes were stable or increasing in criminal justice should come as no surprise either. However, on the one hand, it is surprising that these variables did not correlate signifi- \fi) ()1 cantly with any of the reasons for adoption. Given the expected differences in the future for these two areas, one would eXpect the stability or instability of these variables to be related to the program changes. It would be instructive to compare a sample 0 f organizations that had heard of the programs but had explicit- y decided not to adopt the programs to the present sample on F4 (D these demographic type variables to s e if any differences exis- ted. On the other hand, perhaps it should come as no surprise that these variables did not correlate with any of the reasons for adoption. Reasons for adeption were thought to be organizational considerations taken into account during decision making or organizational justifications for decisions. Perhaps the expectation that the reasons for adoption would correlate with organizational demographics is committing the ecological fallacy of cross-level inference (Downs & Mohr, 1976). That is, the reasons for adoption are on a different level of analysis than the organizational demographics. The analyses of the data concerning participation in deci- sion making revealed that concerning decisions in general, adOp- ters were typical of the unaware non-adopters in the sample. The educational organizations had significantly more levels partici— pating in both the adoption decision and decisions in general than criminal justice organizations. Besides actual differences in decision-making practices between the two areas, this finding could indicate a lack of diffused responsibility in the criminal a justice area. Pincus' (1974) analysis indicates that educational organizations would strive for less clear lines of responsibility as a means to avoid accountability. A second explanation is methodological. Since the instrument used to measure partici- pation in decision making was designed with educational organiza- tions in mind, it is possible that the differentiatio: in crimi- nal justice organizations was not adequately captured by the measure. It is interesting to note that the adoption decision appears to have been a fairly typical decision for the various organizations. This was indicated by the small differences on the variable computed by subtracting the number of levels invol- ved in general decision making from the number of levels involved in the adoption decision. The significant correlations between the scaled reasons for adoption and this computed variable are interesting to consider. The larger the difference in terms of number of organizational levels participating in the adoption decision and the number of organizational levels participating in decision making in general, the more expense and the availability of financial support is considered. However, correlations do not indicate causality and it could be that decisions with costs (either high or low) necessitate the involvement of a certain number of levels. Cost may be causing more levels to be involved (high cost) or less levels to be involved (low cost). The correlation of the computed variable with the EXpected Smooth Implementation scale is also interesting. The more levels involved in the adoption decision when compared with decisions in general, the more ease of implementation issues are considered. The converse also holds; the more ease of implementation issues are considered as a factor, the more likely it is that more orga- nizational levels will be involved than in decision making in general. This correlation could also indicate an increasing reliance on the usual decision making apparatus of the organiza~ tion when a change is going to be difficult to implement. The correlation between the Support scale and the number of levels involved in the adoption decision probably reflects method bias. Interviewers were instructed to ask about the influence of various organizational levels on the decison to adopt immediately following the discussion of support as a reason to adopt (see Appendix c). The contiguity in the interview of these two simi- lar type questions/issues may have lead peeple to attribute sup- port or influence to actors or levels that they might not have ordinarily considered. The length of time that an organization had been using the program did vary as a function of program. Educational programs had been in use longer than criminal justice programs. The ECOS program had been in use significantly longer than the HOSTS, ODOT, and MCPRC programs. This probably reflects both the dif- ferent times when the eight programs were developed and approved, and the differential rate with which different programs diffuse (Rogers & Shoemaker, 1971). However, the analysis indicated no differences in the reasons to adopt as a function of being an 'earlier" or "later" adopter. 98 The length of time that the program had been in use did cor- relate significantly with two of the scales. The correlation between the Changes in Roles and Role Relationships scale, the Support scale, and the length of time in use can be eXplained by the fact that the saliency of these two reasons for adoption was effected by the length f time between the adoption decision and the interview. The further away in time that the adoption deci— sion was, the less likely that changes in roles was remembered as a factor. On the other hand, the further away in time the adop~ tion decision, the more likely that support was recalled as a factor. Perhaps, pe0ple do not recall wanting to make a basic change in roles a certain length of time after that change has been made, and instead focus on the support they have received in the time since the change. The current support is then remem- bered as being present at the time of the adoption decision. The lack of a significant multivariate difference on the four scales as a function of the job position of the respondent is also interesting. This finding would seem to indicate that administrators do not adopt a program for different reasons than a front-line staff person. However, front-line staff could have responded to the interviewer with the organizationally approved "story" of the reasons for adoption. The lack of significant differences as a function of job position would also tend to show that what were given as the reasons for adOption were, in fact, the organizational reasons for adoption rather than the indivi- dual respondent's motivation for starting the program. KC (4:) This study both partially supports and partially refutes the ideas f Downs and Mohr (1976). The fact that the reasons for C) adOption are significantly different depending on the program would seem to say that the reasons for adoption are not primary characteristics of innovations. The lack of significant inter- actions between length of time the innovation has been in use in the organization and the program on the reasons for adoption is dissappointing given the fairly large number of respondents involved in this study. Given the fairly large number of adop— ters (160) and non-adopters (158) interviewed, the power to de- tect interactions shoud have been higher than much of the past research in this area. An interesting question not answered fully by this study is why the reasons for adoption vary as a function of program. There may be primary characteristics of programs that are differ- entially salient to organizations depending on the situation and the people involved. In looking at this study and the results, one might be temp- ted to say "So what? Different organizations adopt different programs for different reasons". Given the lack of consistient findings in the literature, this would hardly seem to be a so what conclusion. It would have been difficult to say what the factors involved in the adOption decision are on the basis of the inconsistent findings of past research. This study has ration- ally and empirically derived what the different reasons are for different programs. The concept of reasons for adoption may still be of impor- tance. Due to resource constraints, it was not possible to ob- tain a sample of non-adopters who had seriously considered the various programs. The researcher's impression after interviewing respondents from a number of these organizations, was that there had been no organizational decision not to adopt the program. Rather, one member of the organization had heard of the program and decided that the program was not right for his/her organiza- tion. This organizational member was serving as a gatekeeper. It is very possible that these non~adopters would not have the same types of considerations as the adopters. That is, non—adopters may not be looking at what the adopters perceive to be the prog- ram's strong points. These strong points (reasons for adoption) would seem to be valuable for program deveIOpers to consider when they are trying to "sell" the program to potential adopters. In certain cases, these reasons for adoption are already used in this manner in the form of testimonials from program adOpters. The link between the reasons for adoption and fidelity, a variable of great concern to the prOponents of the RD&D model, must go unexplored at the present time. This does not mean that the reasons for adoption are not important in explaining how close the adoption resembles the original model program. The reasons for adOption could provide a measure of the adopting site's initial motivation to replicate the program. In other words, adopters with certain reasons for adOption might be more likely to implement and run a program that closely resembles the 1...; C) F4 original program. To the extent that this is found to be true, and to the extent that developers want high fidelity replica- tions, such empirically derived "reasons for adoption" could be used by developers to insure close adherence to the model. Limitations of the Study The present study has a number of limitations that should be noted. First, neither innovations nor adopting organizations were chosen randomly. Not choosing innovations randomly is probably not of great concern since the results of the study were not intended to generalize to all innovations. The limits of the study are defined by the reference and content domains of the innovations chosen for study (Bigoness & Perreault, 1981). In the present study, the reference and content domains are those organizations that have adOpted NDN programs or LEAA exemplary projects that are organizational in nature. The fact that organizations were not chosen randomly is also of little concern. In all the analyses discussed above, adopters and randomly chosen unaware non~ad0pters did not differ on variables of interest. Second, what the programs actually did was not of concern in the present study. Even though one of the four rationally and empirically derived scales concerned ease of implementation, the question of the implementation of what was not specifically addressed. In order to get at questions concerning the content of the program and what effects this might have on adoption, it is neccessary to first define what the program actually is and what it requires. This was never done in the present study. 102 Even given the limitations discussed above, this study has contributed to knowledge concerning shy organizations adopt programs. This study, although imperfect, has empirically demonstrated the differences that exist among programs. The investigation of these differences and the reasons why different programs are adOpted for different reasons would seem to be the logical next step for research in this area. APPENDIX A: Coding Form for the Adoption Decision Questionnaire f]? ltd ADOPTION DECISION QUESTIONNAIRE 1-2, Interviewer 1de at i’1cation number l~Cra1g 5-Save T 2-Dave R éedeff 3-Rand 7—lheresa 4—deana l lnnovat1cn identification runber l-HOSTS 6~GDCT 2~ECOS 7—CAP 3~EBCE (”i—CC?p d-FOCUS 9 MT P?” 5- O= 4-6 Organ1zat1on ‘dent1f1cat1on nlmber numbered conseCu-1vely (crronalogically) vitl‘" innovat.cn 7. 0rd n12a+1on ype- recorded longhand in Call Log - to be categorized and coded at a later date (eg. jr h1gh, c1ty or precinct police dept, d1strict ceart, etc ) 8 §Q_type- recorded in Call Log; also to be coded at later date 9. Respondent's job pos1t1on (_;) at time of Adopt1on Deciszou rec rded 1n Ca ll Log andc oded later Current program status: l— adopter (never implemented) 2- adopter (used less than two years) 3~ adopter (used greater rthan or equal to two years) fi- te. Hn ha c-r (1mplemented less than two years) 5- terminator (implemented greater than or equal to two years 6- aware non—3dc.; ter 7- unaware non— —adopt r l- creator (raver imple :ented) 11. 2— creator (less t.an two years) 3- creator (greater than or ecual to two years) 4- creator & termlnated (used less than two years) 5- creator & terminated (used great er than or equal to two years) 12. Val1d1ty check 0- 1f th1s specific organization is ppt_an inter spondent validity check l- 1f Q_is a randomly selected val1d1ty check and th1s 15 the primary R 2- if Q is a randomly selected val1d1ty check and tnlS 13 the . secondary R 13. Code 1 if this is a rel1abil1 y check l4. Coded l (record number for computer purposes) 15-17. Length of time 0 ha been using P first two Sdiglts refer to years implemented third d1git r fers to n1onths eg. f1ve yea rs and n1ne months would be coded 1059 in columns l3-l5 f1ve years and ten months would be coded 060 104 l8. Was 5 involved with E_betore adoption decision? 1- yes 2- no l9. Degree of 3’s 1r 1volveme.t in adoption decision (Ilse YOUR JUDG’l iENT; 0- no 1r vo olvement In'v olvement JOEC NOT l- lo ir Mvo vement equal decisionwmaking 2— medium involvement power) 3- hi involvement 20. Where in Q is(was) §_bein; used° C- in no Q_units l- in only one 9 unit 2— in some Q_units 3- i all Q_un1ts or a l 0 units that Coulc "uplement E 2% Ndere ‘n administrative ED .s(wasl 51b ng used? e1 0- in no a units ie_g. non adc enters or ”adopters but not yet implementing”) l» in only one 9 within the s: 2— in more tran one 9 but lOt all 0' s 3- in all 0 s or all Q‘s that could 1mp1eme. 3 BL ANK 1F NU SO '22. To what extent was Q_reguired_by_§g to adopt E} J- this was not a factor in this case l— low or moderate 2— high 23. Has 9_considering alternative Efs at the time of the adoption decision? 0- no alternative l- one alternative 2- more than one alternative 24. Did the evaluation (validation) of B_have an impact on the decision—making? l- yes 2- no IOS IMPORTANT LESS LESS IMPORTANT reason IMPORTANT not IMPORTANT reason NOT to adopt reas mentionedifli reason to adopt NOT to adopt unimportant to adopt l 2 3 4 5 AT THE TI ‘EC THE ADOPTION DR BION, THE FOLLONING NERE THOUGHT BY 0 MEMB ERS TO BE REASONS FOR, DR FACTORS INVOLVED IN, THE ADOPTION DECISION: PPOOr $1 CHARitTER STICS 2?. E was con51der red to be a flexible” orograw 10- E was consider ed to be not flexibIe In this item, flexibility 13 NOT synonymous with ”susceptibility to mod1f1cati on ’ (item e*27l i e., a program can be flexible, but adopted with hio h fidelzty, i.e. with no modifications from original site. In that case, it would be coded only here, and not in item 5 27, 27 .3 would be susceptible to modification (broadly interpreted; pre and expected post) 25 E_wou1d not be susceptibIe to modification Pre and expeCIed post. The issue is: would the program be suscept ibm to modification at the point of implementation once the decision was made, Ego ’cr woo d t he program be easily mod1fied as needs arose in tne iuture (foilowing implementation). Either 25 or 26 could be coded for reasons for the decision. 29. Then majo rug outcomes were desi rabIe in O 30. The EEL or Eb outcomes were not desirable 1n 0 The interviewer should be aware of the maior outcomes evaluated at the derons tration site. This item re. ers to those specir 1c outcomes onlv . not issues the 3_perceived as desirable. In this item, inherent within the notion of “desirable“ could be that the E_would be effecgiye in O. i e. A statement that ”the E was effective” is equivalent to ”the P was desirabie” for our purposes. NOTE: Outcomes must be relevant to gfs situation. 106 IMPORTANT LESS LESS IMPORTANT reason IAPORTANT not ”Pl? ANT reason NOT to adopt reason mentionedCNI reason to adopt NOT to adopt unimportant to adopt I 2 3 4 ' 5 AT THE TIME OF THE ADOPTION DECISION, THE FOLLOWING NE RE THOUGHT BY Q_HEN8ERS TO BE REESONS FOR, DR FACTORS INVOIVED IN, TH (.0 ....J (J) {‘0 J1» U) LOLA) 35. [TI .EDOPIION DECISION: 3 would Eot be effective in attaining {fjgf outcomes in Q. Again,1f t.e 5p nredicted he P wouId oe effective” in O tnat' s tjpicall/ red undant, and hard to d1 stirm uish from desirable (e.g. #27 .le'efore, code it there}. Th1 the 5 item VHOU d be coded if C as SUCH}, bUt the' th nR it would be he erztert this was a iactor in majo r outcomes were desira—bI 3 was not impIemented becaus Eff ective in their 9, Code tr 41,5 \o +- U (1’ (I) (I) - {IT (‘1‘ . J r) :1 C ‘8 CECISICH. The fact that the major outcomes were vaiidated was a factor in the adoption decision. Item refers specificaily to the fact that the E was evaluated and this influenced the decision. Other P outcomes were desirable in O Other P ou come , were not desirable in 0 Any outcomes not validated at devel oper/demonstration site wouId be included under other outcomes. This could include non— validated outcomes the R percei ived as maior, etc. q Remember: Outcomes mus t be re want to Q’s situation. P would not be effective in attaining other outcomes in O.J'If the R “perceived the P as likely to be effective in O, that' s redundant and hard to distinguish from desirable so it would be coded in #3l. If R saw the other outcomes as desirable but feIt the P was not likely to be effective in attaining these outcomes in §,~t he extent that this ian uenced the adoption decision is coded here. NOTE: it is possible that the R i.eIt severaI other outcomes were desirable but not likely tc 1e effective in this Q_but this still may not be influential in the decision (e.g. it's possibIe onIy the major outcomes were influential factors.) 107 IMPORT W LE SS LESS IMPORTANT reason IMPORTANT not IMPORTANT reason NOT to adopt reason mentionediflt reason to adopt NOT to adopt nr1nnnr+c to adopt l 2 3 4 5 AT THE TIME OF IN E ADOPT ION DECISION, THE FOLLO 39. 40. 42. -3“ c not involve a New P might b "Similar" mea have both rogi1ams in O at the same E REASONS FOR, OR FAN ms INVOLVED IN, THE ADO r or i3entical HE: by R C; 1n; 'e new maragement system ge change in practice, it ddi ion, expansion, E_would be likely to function smoothly in O i. e. ADMINISTRATIVE) E_would not be likely to function smoot ”Horkable” is a subset of ” with innovation adoption. {m h i ffic . in Q ier cv to E in O interview TION DECISION: . o might be coded in modification, e w‘ING WERE THOUGHT BY Q H 18E-S tc. (NORKABLE; O member-O, , irequently associated he Le.}nooa that P could be that IlKElIhGOd on the easily implemented and the influence 0. adoption decision is what is important for this well E_would work” in terms of ointeractions bet refers specifically itself. (e.g. interactions 'stration, etc.) outcc 1e F '1 ‘ S. Azso, between teachers and B would be likely to function smoothly in.g i.e. SERVICE/PROCESS 3 would not be likely to function smoothly (WORN in O ‘0- ABLE; Om ites tl' ween O membersa -J the membe n “0t hCW J If) _I + ’1 4 .. U D.) .J .5 r - CII§EL This item pair refers specifically to interactions between O members and the O’s clients. and students, correction .s officers and inmates, This m11‘t also be administrative. in processing of cases” (e.g. int e.g., eractions between teachers etc. for N08, H improvements “90 Tflii LESS LESS IMPCT TANT reascn I“‘C?TA11 not THAORTAKT reasor NOT to aoopt reason mentioned(fl2 reason to adopt NOT to adopt unimportant to aocpt T 2 3 4 5 AT THE TIME CF THE ACOPTTQN DECISION, THE FOLLOWIN GW WE‘E TH OUCHT BY 0 HEMSE‘q " 1 ‘- 1 i '3 13 (1‘ 5-4 :1 «c: Q r. <1 (*1 Cf) H :’ ~4 II: {71 3) C3 K) T) -—i »—a O L) {“1 (T) ‘v ( E '13] T) r—f .J O 2‘ O 'h (D -C\ .3 .J (D .3 (1 < processing c ients through u, proved, etc.) 1,1 ,+ A. E. '. WOU1U NC .. 1.9.14.9 8 :Ct T. if t me to execute wouid ta a it' 1 1 ’ue to execute t me’g requires to execute, or the 1o'n nt of time E_requires over and above cedures, i. e. incrementa] time. Code degree of f tni s is a REASON. NOTE: refers to existing 0 entia] additional staff. I'UI'U .13 '3 0‘1 0‘! (+ ab} e contract a1 type commitment cont ractuai type cormionent from Q financiai type comritnent oniy. Aiiows financial commitment, but aTso Tooking ua. type commitment (e.g. In NEH, schoois mmit a certain number members to a oots woutd have to imp ement a certain type as Ae in order to quaii s a program I“ r'l o‘i“ hm“. re~4lJI.‘t r2 ccnsi (‘1 I k E r (‘1 L-‘1 m 1.1., ‘ I 1- i To C -. r C) A") *< m "5 D) ...J. O ( f - -‘ (1.) J?) -1}. o: \J O O I i I /3 3 ‘I O S ' '51 \‘D :3 ”J- 3 (I: (D (1 (a (1 "h C) I’D "S ( l. C) “'5 (3 :3 r+ (‘1: .l r“? “i .4 . GL- 1 \ ”3 O «1- 170 :5 mm ~1 -.J. -J. . 3 (1'31 0) (3 U3 Q! "h L1 ““1 -'~ ' O H- C). H ‘3 r+ (D 1"ij C) J “h :3 -_| j: 1"» ~11 (7 H- A4 to - - item to code, because it‘s easy to think P' ‘suits 0 goals”. So: “GOALS” means the r "Purpose"of O, on a fairiy ABSTRACT TeveT, p. osophv is to favor individua.ized instruction here . use “GOALS” with "desired outcc -mes , which are Tess ings, Tike reading achievement. DJ —-' 1"? m 45 (D U') (I) O “:r' :1‘ '4. (1‘ Ln _' r... g—‘ g - - "f"! 1‘.) CJ ("Tim :7 c7 - U’ 3310 ft U“ (.71 p...) ‘\ .J 1.0 IMPORTANT reason 1 to adopt not MERtiCHEGL' unimportant 3 NOT AT THE TIME OF THE ADOPTIJN DECISI 3N, TEE F 70 BE REAS O\S FOR, CR FACTDDS "””‘“ED 1%, VUVU E C .9 :8 :1) L1) .111 ”(71 --.| f3 ‘0 O 0.) m — .J (15 (T) -4 . m Ln L") 3'3 '0 m fl° {D u C) .....1 C3 iri' Q "5 w ,o b (La (1* (D ('f (v +- C) )' . E wou1d not involve 1a - P wou1d require a lar 9 Hot c11ents e.g. HIT: 45L!) ('2 {.‘J w S 16 up...._..— (11 :3" :0 :3 .4. 'in ‘1 ) (D (D m (D ’ "hm m (D C) r+ us 1 (r. r... m 0 1'3 3 C)" (T) "S I ' ’ I P would 1nvo1ve a Iarge c: OLLD"T new EFs roles or role beha :65 are changed that' s re1evant fa GT f‘ I‘T' IMPQ {1. l (”12“: ‘ reason to acopt U’W WLNG Vanna: ~. A . .1; . lukL—‘s‘u'a ; A | o...) CT“ uAm 3' < .J- 0 U1. Q 0395 Cr 1" no? ‘3 C1) m o b. DJ <\ ‘1 C: (D 7.1 j 13 (D ' U1 11:) :5 (D '10,. U' (7.") r+ (t1 ( 2'3 m 01 . .. J 8 (11 K J :0 {1‘ T— r..—-‘ . —’o r (:5 J J ....1. ..‘_ -.J (”r ‘7 .f ‘3 L- D I ('1. if) (1 L f \I processed by th nts, eions, j uses the termc 1 THE wav p80? anmur'cy Arbt .J. e -1 u m \l 0’) (D m- - (‘1 C: acorn-x P. L363HD —.J -.J. ( ‘0- "fi ‘- *1 Q, 9 ("I- {T‘ --J. (.J D \/ 0 members ro7es or role behaviors 10? + \— O the tutors are st_ udents and coded in i8. potentia? new \ re1ationsh1ps (inter “ct fl 0?: between a y P would not involve a O actors Coding th1s refe1s to 6 r0,1 role benavior 15 thought b” g a We rg nge 1n the O W‘CE. ro1e re1ationsh1ps between an) —.J ' t. \ ~1 :3" -J. ..J. "1 UP) C." U.) o (D (D C). -_J C: (D 0 Cl. .‘3 ('1" P? .4. C) "T :3 0) LC) U) 1?) (T) U.) (D ~1- (“r :3 (I) (“*- 3. I‘D 3‘3 :3 fr —'- E 13 (.0 C.) ‘5 q. (+ ('3 CU ,3 {J ”i Q) I,” e.g. HIT: E ccu ro1es '0 c11en (O membe. , (teacher/tutor coded HERE 1n ,~‘ Us 9 / a!" a i/OF C- : F. v- t ~+ r r (f) n: -JI m m 0 I)“ (1‘) f1) (‘0 :3 IT (I) (‘f r".- 1") “3 m o --.l g” :3 fl "3(1) 5 *4 ( U U.) m P? 6+ (0 ”J- ("D (t— C) C) U‘ C) ”'5 change 1n role utee and/or tucor (f! \ fl {e CL \ IMPORTANT LESS LESS II‘PORTANT reason IM?CRTA%T not IMPOHITA 'T . reason N01 to adont reason mentionedCfiz reason to adopt NOT to adogt unimportant to ad0"t i 2 3 4 5 AT THE TIME CF THE ACCPTICN DECISION, THE FOLLOWING WERE THOUGHT BY 9 HERBERS TO BE RE SCJS FJR, OR F’CTCRS INVOLVED IN, THE ADOPTION DECISION: 3 (‘1- 1 m om.) V) V» C -4 ': practices/metho es ctices’metoOHs e ceded ins d chance in C c ract' s of ciients erzn ' ,4 T .g attorneys in a c p 01 ' '4' ~ 0 C (J “V; -‘ (I) f. I] C. -5 "i ([1 (1.1 .5 :3 {1| -4 r1+ - (D ‘<- ’ ' () -—’ (T) '0 (‘1 T") (I) -._I. J. (D _O 1') ——~' L) ' —- 11:) o, to O (T) 1 0 (1') DJ ‘< , asi prosecuti .J 1 :1 13" n1 (.11 + -J A P Q 31) ‘1) " - Jo " U (+ ._1. “I. C.) - a s tentacts with sonc e vine in the ibraiy re e 'sr cases in ccurt the c' i t .1. (1| .|‘ 'V "I L, I k/ L: .4 . (I) *4 ~-—J ‘3 (D '5 T} U“ (D . ‘S .3 O ——.l IT) ve a change in 0 pr ac _‘n s. On tne other hand, it the attornevs 2 6 prior to the adoption decision (as is L.oica: iv 1.1 jL ust focusing on prosecution or iust doing the researcn, e in role behaviors would occur as weii. Again, the ecision is whether these are REAS NS for the decision. ’1'“ .u—JI 1 been more \-—" "h < (l “" LE. 1 (/ C‘) .mmm 3m()(‘u OHOMU‘Ziflifl- ‘1 :1" m ‘C! {D O 1 -4 r1. 5!. (1 Cu 59 _E wou d improve the interoersonai reIa tionships in O 50 .3 wouid worse n the interpersonal reia ionships in O 7;... 11.5 item refers to aspa 01 in.terperson ti onshi bs; e.o. feeiings, f3“ ("1 ‘3 CU \\ ”J :3" (D :3 1 (1 (D (I) O) (.3 (A ‘ ('3 (D W r+ O Q) C. _“ H O 0) -J fl (1 DJ 0 Y} (t1 ..J «1‘: m 0 3 :3"; n} .3 f) .1 23 (D Q) ('1 «1+ C“ C U) L) (1.) g: (T rf C) '1 (r T.- f: (+<(31D L". D. a erper-onas reiationships be t and ciients, and/or member rs OE CHANGE ITEM 51—58 renme i- 58. . c and ndn Hie ber rs. iOT 61. P wouId be visibIe in C 62. P wouid not be very visibie in 0 Be sensitive to antagonism or poiitica] issues. Is E_observabie either inside or outside of 92 Has this a reason for the decision? IMFORTAAT LESS LESS IMPORTANT reason IMPORTANT not IMPORTANT reason NOT adcp reason mentionedifi? reason to adopt NOT to adopt unimportant to adopt I 2 3 4 5 AT THE TIME CD I THE ADOPTION DECISION, THE FOLLOWING 'ERE THOUGHT BY 0 MEMBERS IO BE REASONS FOR, 0R FACTORS I“’CI“? N ULis.‘ L.) 1’4 1': 1- r 1 ‘1‘) L) (- 3 51 P“. (3 .2 (.3 ’71 ('5 H (n H C) :1: 53 E_wcuId be very triaIabI e in 3 54 E wouId not be very triaIab‘e in Q Does not refer to ‘incnciai commitnent. One aspect of ”triaIatIe is ”reversabIe” {once an Q starts 5 they can stop). Or, it might be that the 3 does no: respir: fuII 1mhIementation. i.e. E couId be tried or a Iimite’ basis in a smv I part of O or S‘ for a short period of time, etc DOES NOT IVCLUCC Q_CRARACT v~ ISTICS PER SE, E.G. REE TAR: 1h 3. I 55, .Ew uId be easin co.1municabie to Q_members 55, .Ew wouId not be easin C0u1m nicabIe to 0 members This is a primary attribute of E that doesn't vary from Q_to g. This item means ”whether they can speak about E_to others easin” 57, E_wo uId be reIativer ine pensive for 0 55 P wouId be reIativer inex pensive for 7 due to grant support brought .- in by R , 69. P wouId be reIatiVer expensive for C ‘- ReIative expense refers to Q“: interpretation of the cost factor a: 1t pertained to the a:option decision. SURPCRT /A“TAGm USM NOTE: Remember, SUP RIRT does not equaI INFLLE'CE. These items are cIues for O? coding, bu t you must use “ianu me .ce probes“ to do an accurate job on the CR. ALSO: Antagonism ma y mean actuai co an 1ct OR just Iack 01 support!! y—J )—.—o P\’ IMPORTANT LESS LESS IMPQRTAHT reason IMPSRTFHT not IMPORTFMT reason NOT to adopt reason WGHtIOHECCfli reason to adopt NOT to adoot un1mgortant to adogt I 2 3 4 5 FT THE TIME OF THE ADOPTION DECISION, THE E1J LCxw’. NG WERE THOUW T81 g MEMBERS TO BE REASONS FOR, OR FACTORS I’“OLVE D IN, THE ADO PTICN DECISION: 70.k1rbe1s of poI1cy sg were support1ve ofP 71. Members of poI1cy S; were antagon1st1c toward P PoI1cy 3; sets the poI1C1es that 0 Eur: ‘oIIoz. En" Ea, orig cod sCERoI Eiifié gi_g§. as PoT1Ey 53‘s NOT other aor1sarJ counC1Is sash as ”T1t1e 1 Fd¢1sorv Co bC‘IS“, ‘Commun1ty CounC‘is , etc. For Cd, there 15 more var1CC1I1ty; coje whateve1‘ s reIevart. That m1ght ancIuce a commun1ty courc11, 11 1t m-ves poi1c; decxs1oos that Cxcr1 the operat1ons of Q, Coded orzv 1f suo ort/antagon1sm was reason for adopt1on. 72. Members of adm1n1strat1ve SO were support1ve of P 73. Members of adm1n1strat1ve §O were antagon1st1C toward P Adm1n1strat1ve SO coord1nates ope rat1ons 01 (usuaIIy) more than one 0; nires , wr1tes the budget, superv1ses operat1ons, etc. 74. 75. Adm1nistrators 1n Adm1n1strators 1n ChooI d1str1ct.1n court, etC., depend1ng In ED, Fd31n1strative so is Apryg thes Cd: it mI‘”t he a STate court, d1str1ct on an adopt1r.g un1t. de1n1strat1‘ie or execut1ve S 15 1nvoIved w1th day to day proc ceduraI act1v1t1es that are eIevant to the Q. Coded enIy 1f REFSOM for adopt1on dec1s1on. 9 were supbo1t1ve ofli g were antagon1st1C toward P Adm1n1strators 1ncIude pr1n C1pa15, VP's, off1ce managers, dean or stuc ents, etc. IMPORTANT LESS LE”: IMPORTANT reason IMPORTANT not INT“?TANT reascn NOT to adopt reason mentioned;fi% reason to adopt NOT to adopt unimportant to adopt l 2 ° ” A i J Q a AT THE TIME OF THE ADOPTECN DECISI N, THE FOLLOwING ME {THOUGHT BY 0 HEMEE.S TO BE REASCNS FOR, OR FACTORS IN‘JOL-VED IN, THE ADTPTIGN DECESIbN: 76. Specialized §Q_staff were supportive of E. (Might or might not oe directly involved). 77. Specialized §Q_sta aff were antagonistic toward E. e.g. , District reading personnel that only spend part-tine in Q. 78. Spec‘i alizedig staff directly involved with E_.mpie. ta,ion were sup :zortive of P. 79. Specialized 9.5m aff directlv involved with E_were antagonistic toward E e. 9. full time reLiding teacher in O that would oe directl involved with iniplementing P in the O. 80. Front line staff (potentially) directly involved with 3‘s implementation were supportive. 8l. Record Number 2. 82. Front ine staff (potentially) directly involved with Efs implementation were antagonistic. “Potentially d rec tly involved” distinguishes between sta hat world be ir volved with the actual implementation of the P on a rego lar basis and those that would not be directly involv ed; i.e., they had a vested interest, e.g., lawyers in ‘1— M08, corrections officers in DOT, teachers in HIT, etc. \ a 83. Front line staff ot directly irivolved with P were supportive. 84. Front line staff ngt dir ctl‘ involved with P were an eicr‘stic toward This it perta to the extent that ct her w‘rff influenced the adoption deCi sion e.g., correctional intaie i ° ‘T rn '13 since its a process E_. not an intake progra Q~in HIT (a reading etc. {‘0 U ....a y.) b IMPORTnNT LESS LESS IMPORTANT reascn IMFJRTANT not IMPORTANT reason N " to adopt reas.on umentionedtflz reason to adopt HGT tc adopt mmpcrtant to adopt 1 2 3 4 5 AT THE TIME OF THE ADOPTION DH WI N, THE FOLLO”INGk ’ERE THOUGHT BY 9 MEfiBERS 03 CD C 1 U1 87. 88. CO KC) 91. TORS INVOLVED IN) THE ADOPTION DECISION: Unions were supportive of P Uricns were antagor1istic toward P Unions 3 ogid be concei-Jed or troadiy as any profess1onai grouping that is reaevant. Could be an “association“ rather than 5 onion per se. 0 C1ients were supper tive of Clients were antagonistic toward P Race ‘3 ciients are processed by the 0 e. 9. students, felons, etc. Record the ex tent that they infiuenced the adoption decision. Ciients‘ parents, reiatives, and other " o~"on tv people were supportive O .C 81 C1ients‘ parents, relatives, and other “c nmon1tv peopie were a1tacorist1c toward E Kore parents of sto den nts supportive of educationai P, did they push for it? Note hat ciients' is possessive, their parents, not the ciients Other actors (not yet specified in ADQ) were supportive of 3. (”ho?) Other actors were antagonistic tcvard 3_ (Who?) [Record the Q role of other actors on Cal” Log and code the inf1uence here. ] . IMPORT"T LES: LESS INPCRTANT reason INRCNTANT not ENRORTANT reason NOT to adopt reason mentioneocfl reason to adopt NOT to adopt uniin oortant to adogt l 2 3 4 5 AT THE TIME OF THE ADOPTION DECISION. TNE FOLLDNINo NERE TH .“UGHT EY O MEMBERS TO sE REASONS FOR, OR F;"T“v2 INVOLVED IN, TLE AGED T '.ECISION: INN’ O v’ATlcN ”Li“"’ON (Ioszoe C‘ 93. Someore ‘n O or S3 was actively orooctfng P (evistence of a ct am: .on) 94. No one in Q_or §§_ras actively proecting R Someone blowing the truooet, saying “let‘s imolerent this prograw'. Tyoically this will be the respondent; therefore it might be difficult on some occasions to get info that suggeSts that it was in large part due to this respondent‘s actions that this P_mas ultimately adopted PRGEE do Some Goalit of the champion ENhat qualitv?) contributed significantly to adoption decision. [Record the quality co th: Call Log and code the inf luence.] Should be so me quality of the champion, e. g. persona.ity, chariswa, etc. NOTE could con tribute pCSl tivelv OR negatively; code where anorooria;e. If yo“ are interviewing the champion you can infer qualities from things they say. (e.g. ”i would have pushed this 3 no matter what“ ) CHANGE AGENT (outside 0} 96. State facilitator or someone (specific) within NDN, E A or HHSl (not HO TS) wa s promo ting E 97. State facilitator or someone (specific) .,:hin NDN, LEAA or HOST was not promoting 8 We‘re not looking for whether or not the ao encv wa s promoting the B. we are looking for some 5; ecific person that the L has had contact with from either ND N. LEAR, or HOST that has been DUShlfig P. }.—l ’——.\ Ch IMPORTANT LESS LESS IMPORTANT reason INPC RTANT not IMPORTANT reason NOT to adopt reason mentionedifll reason to adopt NOT to adopt unimportant to adopt l 2 3 4 5 AT THE TIME OF THE ADOPTION DE “I ON, THE F0 LONING WERE THOUGHT BY 9 MEMBERS TO 8E RCASCNS EOR, OR FACTORS INVOLVED IN, THE ACOR TI ON EEC ISI ON: 98. Other change agents were promo ting P 99. Other change agents were not prom .oting 3 Looking in or otier change agents NOT AFFILIATED NITH NDN LEAA or iGST that were or vere not or oroting P and HAS AN INFLUENCE on the adoption decision. Record the individual's rcle in the Ca‘l Log and code the influence on the five point scale. e.d. someone from NGSTS corporation; consultant from FCUS institute, etc This item could be a reason not to adopt if, for e>amole, sane significant change agent that the Q has hao contact with in the past, Is not prompting this 3, so B_”looks bad 100, Some quality of the change agent contributed significantly to the adoption decision. (What quality?) Record this source on the Call Log sheet and code the degree of influence. Could be reason for adoption or non-adoption. Note the champion is not a change agent Champion is someone hin the 0, (9-180 if S0 is involved in a opt on decision) he change agent is someone external to the Q, (or to the SQ f SQ is involved in adoption deciS‘i on) [5: .1. :‘I -4 2‘.) () T 2: SITE VISIT OR OTHER PRE-ADOP TION CO? 10]. Site visit(s) was useful 102, Site visit(s) was not useful Site Visit refers to a visit by some 0 or SO member to the site of the innovation prior to the adoption or implementation. It is crucial to note that the site visit was prior to the implementation and figured in the decision. In other words, it might have "simply cinched the decision” The notion in this item is that the irformation gathered from the site visit resulted in a positive influence on the ultimate decision even if it was just of a confirmatory nature because 9 was likely to a opt the P anyway. Similarly, if the 9 had already “decided to adopt” tne IMPORTANT LESS LESS IMPO TANT reason IMPCTTANT not EMPORTANT reason NOT to adopt reason mentionedCR reason to adopt NOT to adopt unincortant to adogt l 2 3 4 5 AT THE T ME OF THE ADO.TION DECISION, THE FOLLOWING HERE THOUGHT BY Q_HENBERS TO BER I03. IDA. lOS. loo. LEASOEIS FOR, DR FACT RS INVOLVED IN, THE ADD PTI ON DECISION. 0.: P THEN wert on a 'te visit had a negative experience and ch 4' l e" '7 ST ng their mino, that wcu lOl. It should also be noted ,4 u . that the visit does Eot have a - to the original development site. It could have be:n a s te Visit to a second generation adopter. Visit by consultant to Q was useful Visit by consultant to g was no: tse‘ul The consalLar: might be 1L8 cravge agent or the develooer or someone who was directly aiiiliateo with getting file progra. n started at other site. Consul ant = someoze who visits 0 to advise on adootion. ther contact(s) with P was (were) useful Other contact(s) with P was (were? not useful Refers to mail, iei‘QNST contacts, info gathered through other channels (e.g. word of noutn) Only code if any 0. these were a REASON tor adoption decision Includes initial contacts if they were nenticned. ed 118 EMPO; TANT LESS LESS IMPORTANT reason 'MTCRTANT not ifiPCR’ANT reason NOT to adopt reason ment1oneCC1R reason to adopt NOT to adopt unimportant to adopt l 2 P 4 5 AT 1:5 TIME OF 13E nSCPTION LECISIOH, THE FOLLCWZNG WERE THGU TO BE REASONS FOR, CR FACTORS INVOLVED IN, THE ADOPTION DECISION: m1ateria._§ lOT. Appropriate materials W B were avaiaatie be ore a option 108. Appropria e materials 1Cori were not ave11ab1e before adogtion e.g test .ater ria1s, texts, etc. were available, and this was a REASON fo r adoption decision 0?. Materials desired by 9 we ould be ot1a1ned 1f :_we< adotted llO. Materials des red by Q would no be oateioeo lj_§ was asoited (1Co1d no resuit from adoption) Note difference between tb1s and He preceeoing 1tew facilities lll. Appropriate 15:11 111es were a aila ble f01 P be to e adoption ll2 Appropr1ate fac1l1ties were not availabie 1Wr P be ‘3re adoption e 9. re ading or math lab, filing system, compute r, etc l131 Facilities desired by Q_would be obta1.ed if E was adopted ilé. Fac1lities desired by 0 would not be obm ned if P_1-zas aiop ted (would not Fesult fron1 adoption) Note difference between tnis and the preceeding item. 11; IMPORTANT LESS LESS IMPORTANT reason IMPORTANT not INPORTANT reason NOT to adopt reason n.entione dCNI reason to adopt NOT to adopt 1qn1rtant to adopt i 2 3 4 5 AT THE TIN OF THE ADOPTIO1l DECISION, THE FOLLONI NG HERE THO' GHT BY Q_HEMBERS TO BE REASONS FOR, OR FACTORS IlJOLVEO IN, THE ADOPTION DECISION: financial support (ESL NOTE: FOR THE FOLLOWING ZTENS SEALING WITH FINANCIAL SUPPOPT The appropriate Iere: 01 government is defined by the lQEEE of the specific ’“"ina allocatio. decisio.. e.g. CETA money starts in Nashingto.1, bat aITocation oecisions for CETA inzds are made by state and lnc;l a,encies. Also note that some 01: these items refer to avaiiabilipy of funds, while 0 hers refer to funds resuIting from adoption. e.g.. federai funding was availabie for starting the program (”seed mone'“) and O was expecting the local government to pick up is nding a1ter ederal money was terminated. 'NvaiTability” M O/o R ”resul ting from adoption“ might be REASONS for as doption A130, note that Q predictions concerning likeli Mo d of P fundi 1ng being picked up are always somewhat uncertain so don't expect them to be stated in certain terms. ALSO: Note the SLACK RESOURCES item (# l3l). llS. Federal financial support was availabie lI6. Federal financial support was not availabie ll7. Post-impiementation Federal ES wouid esult rom adoption ll8. Post~impiementation Federai ES w'oold not resalt 1rom adoption Refers toa another grant, NOT gr rar1t wnich is (or cosid be) supporting Pfs adoption. 119, State ESw was available 120, State FS was not available 10'“ 14. 111nn*in LESS LESS IMPORTANT reason IMPORTANT not IMPORTANT reason NOT to adopI reason mentionedfiflt reason to adopt NOT to adopt unimportant to adopt l 2 3 4 5 AT THE TIME Or THE ADOPTION DECISION, THE FOLLONING NERE THOUGHT BY 9 MEMBERS TO BE REASONS FOR, OP FACTORS INVOLVED IN, RE AOOPTIOI' O CISION l2l. Post-implementation State FS would reso 'lt from implementation 122. Prst~iwplemertation State FS would not result from impleinentation Aoeii. raters to O picki1g up support fro: ANOTHER grant, no: one nhich is, or could to supporting E_adoption l23. Local FS was available l24. Losal FS was net available l25. Post-1m‘lementation -ocal FS would result from adoption l26. Post-im lementati on Local FS would not result from adoption Again, not adoption grant. l27. Private FS was available l28. Pr11ato FS was not available 'Private“ FSr refers to any non-governmental support, at any level 9.9. United Way, local Anerican Legion Chapter, etc. lEJ Private F3 onld result from adoption l3 Prizat ES :ould not result from adoption Again, NOT adoption grant. 13l. 9 had money in its budget which it needed to spend This refers to slack resources w1hich actually motiyate the a1option decision e.g., “we have some funds in our NSF bo3get we‘re not spending, so we' re getting a computer terminal." 121 IMPORTANT LESS LESS IMPORTANT reascn IMPCRTAHT not 1MPOR.AHT reason NOT to ado; reason mentioned(x reason to adopt NOT to adoQt uniqgortant to adopt l 2 3 4 5 AT THE TIME OF THE ADCPTION DECIS ION, THE FOLLOWING HERE THOUGHT BY 0 KEfidERS J.- 3R FACTLRS INYCLYZC IN, THE ADOPTICN DECISID? SlAFF availabilitv NC 3: The followsrg 9f ob levels are differEi: than those used elsewnere in ADC a GP. i.e., distinction IS \CT made here between 9 and SO, and between ”involved" and ”not in mo olved” 132, Administrative staff required for 3 were available (These could be in SC_a:§ipr in O} l33. Administrative 5 aft required for P were not available Code here ONLY if given as REASON. Administration = coordination, etc. 134, Soecialized staif required for 3 were available 135, Specialized stiff required for P were not available Agcir, these couid be in Q.er in SO'.§ and, this lSST be given as VFJSOW to be coded. 'Sp ec ialis must have sore SuDStaHtl/Q function, rather than just coordination & administration 136, Front—line staf‘ req=ired byP P were available l37. Front-line staff requiredb c were not available He e we dor t nee d Etc distinguish between ”dire ct.l ly invo lved“ l" and others, since 'required by Ef' iiaplies "dire ctlv involved”. £3.) 0) .4...) (A30) 4401 T41; T44. T45. IMPo~TA“T LESS LESS IMPORTANT reason IMPORTANT not IMPORTANT reason NOT to adopt reason mentionedifi? reason to adopt NOT to adopt unimportant to adopt T 2 3 4 5 AT THE TEME OF THE ADCTTION DECISIC’L THE FOLLO”ING WERE THOUGHT BY Q_HEMBERS Support S Support staff reoui training Adequate Adequate E_was within the sk E_was not within t ed ”7 (N :U U f A) Th 7 .1 ( 3 ~ 4 p3 :13 W 1-4 k 3 T'— “1 .1 (J —-4 .V, THE ADOPTION DECISION: taff requ dire d for P were avaiTa bTe red for E were not avaiTabTe Exanoies ofs upoor t staff wc .Td be secretaries, cTerks, maintenan:e etc. Note t‘at “voiunteers“ mignt be front- Tine CR support depending on their primary funct‘on in P. e. g. a teacher‘s aide the t was tutoring pJoiTs woa.: d be "front- Tine, wniTe an aide whose primary roTe was xeroxin; materiais Vioqu be support. If totn functions are mentioned, code in both items. . d by O woqu be obtained if‘E was a aff des1red by O woqu not be obtained if E wa I O on ted 5 adopted and skiTT range tr a1 ning for s taffw was avaiTabTe trai ining for staff was not avaiTabTe i.e., wnether 5_feTt that the avaiTabiTity of adequate training was a REASON for adoption decision iTT range of the staff he ski TT range of the staff i.e. "5 eff co JTd run P w1 thout much additi ‘aT training“. OnTy co e here if give n as a REASON. e. 9. con 't make deductions such as for the NOS innovation, "MOB uses wyers, Q aTready has Tawyers, so M08 is within staff 5 skiT range, and ti1at' a reason for adoption". Refers to EXIST Ta T ITiG staff, not potentiaT additionaT staff. IMPORTANT LESS LESS IMPORTANT reason IMPO TANT not TMPORTAN T reason NOT to adopt reason mentioned£fia reason to adopt NOT to adop unimportant to adopt T 2 3 4 5 AT THE TIME OF TH ADOPTION DECISION, {E FflrinwiNG HERE THOUGHT BY 9 HEMEERS TO BE TEASCNS F0”, OR FACT:'RS INNTTLVLJ ii, 1%; ADOPTK ' MC SIGN: immediate environment T46. Socio—econm mic TeveT of £1” _ communitv was changing T47. Socio-ec onon zTC TeveT of §_'s Lo1.sn..y was rot changino er OR Tower income peopTe into n e.g. infusion cfa ddi tic.ai high 0‘ E iven as a R“ community might be SON for adoption decision; nan stagnancy might be a reason for c nging too. T48. There was a CLIENT need for P T49. There was No CLIENT need for 3 Must be cTear that it was a CLEENT nee d not just 0 need. Remember, ”cTients" are the peepTe crocessed by C that is, O' 5 targets. e.g.' 5: Students TI”dEd a remediai reading P, youths needed an arbitration P, jur01$ needed a smoother system, etc. T50. Other factors in 0's immediate environment { e 93 community) demanded or faciTitated program cfange reT eiant to P. (What factors?) T Record Number 3. 32. ther factors in gfs immediate envirozw 1t ( e g, conniunity) were obstacTes to program change or did not demand program change reTevant to P. (What factors?) " Record these in CaTT Log, ano then code. IMPO.TANT ESS LESS IMPORTANT reason INPSPTTNT not IMPORTANT reason NOT to adopt reason mentionedihi reason to adopt NOT to adept unimnortant to adopt T 2 3 4 5 AT THE TIME OF THE ADCPTION DECISION, TPE FOLLPNINa NEQE TerGrT BY O V5.8ERS PACT OPS ‘1! - 1"“ 1"" f‘xf‘”:'?".’“m IN”OL\"'T '. In; .3111!“ : ;=..‘-;v. ‘nv [5.5-1 BE REA.SONS FC P OP —4 o wider environment T53. State or federaT poT1c1e§ rec uiredr or fac1T tatec program c T54. State or federal 'CTiCies did not reCL:ire CT ange or made ch e.g., Policy supporting bi- Ting ua T education ini considerabie change in educ: tion. NOTE Th1s 1 refer to avaiTabiTitv of funds. State or federaT economy demanded or faciTitated difficuTt. Michigan‘s present depression is a g:;o: exanpie. remember that the item refers to Tr 1E TIME OF THE HOT 1HE PRESENT. Other factor '.} State or federaT economy was an obstacTe to program change, P1537196 anoe difif'CUTt tiated tem does NOT program change made change however. ADCPTION DECISION, swider environment demand ed or faciTiteted 3 change {Nhat factors?) T58. Oth.er factors in gfs wider environment were ch 3 tacTes to ;_char;e (what f cto rs?) Again. record these in CaTT Log and tnen code. e.g.‘s micht be. eading scores, SAT’s,r ising crime rate, trends towards comm unity- based interventions, etc. y_l {‘0 U”! PGiTfii. LESS LESS IMPORTANT reason IMPCRTANT not IMPORTANT reason NOT to adopt reason mentionedtfii reason to adopt NOT to adopt unimgortant to adopt 1 2 3 4 S (f) AT TfiE TIME OF THE ADOPTION DECISIOH, THE FOEEOW’Na HERE THO GHT BY 0 HEHB ER TO BE REASONS FOR, OR FACTORS INVOLVED IN, THE ADO? iDECISIOH: INCENTI”ES {O MECHRNESMS FOR EHCO'RSJING/OISCOURAGING, AOOPTIOR) individLaT aii ——.J (D i incentives for aoastin” CY ifiDTementing P were avai {what inc thives?: incentives for aniatina er impienenting‘F were no _—J (I! \O y.- 1‘3 (3 4 f. —J (1 £- (2) " ‘I; v avaisabie (‘1- ._I (.73 (I) H :3 r\ .1 < _.1 ‘3 Q ——‘ (What iiicen‘t‘1"e3?} E.g. re: ease tine tra've: to conferences, saiary iicreases or bonuses, recogm ition in newsietters, awards, etc. Again, record what the incentive was on the Ca]? Log and then code. If you get some of this info. from the OP, BE CERTAIN that it was an incentive that had infiuenc e on the adoption decision concerning THIS SPECIFIC 3. Note: Epov Jidoai incentives are organizationai properties that are o;e ti ve in terms of encouraging or discouraoing tr.e act ivi es of indi viduais (as they reiated to this specific 3). I r I t .i organizational 161. Organ'z ati onai incentives for adopting or impiementing were availabie (what incentive s?) 162. Organizational incentives for adopting or idoierenting vere not avaiTabTe (what incentiv““} 5.9. oubiic reiat ic sfor gt survivai tfi O was enabzed by adoption of 3, etc. Again, record what the incentive was on the contact sheet and then code the amount of infiuence it had on the decision. Note: orgar nizationa? incentives are incentives for the entire 9, i.e. i.centives that wiTT benefit 9 rather than just individuais within 0. a.“ N )\ IMPO TANT LESS LESS IMPORTANT reason IMPSRTA“' not T“"’RTAH: reason NOT to adapt reason mentioned(fl2 reason to adopt NOT to adopt unimportant to adopt T 2 3 4 5 AT THE TIME OF THE ADOPTION DECISION, THE FOLLOWING HER ET" OUCHT BY 0 HB‘B E3 TO BE REASONS FOR, OR FACTORS INVOLVED IN, THE ADOPTION OE ISION: HANAGEMEHT OF ADOPTION DECISION .63 locatirn was we}? pianned, or adopticn decision was weii-managed. (9.9 gocd discussions, meetiris, partici.ation, information, etc ) Likeiy to be used as a WP CTOR" involved in AD; may not come up as a ”REASON”. 164. Adaption was not weii cianned, or adoption decision was not weii manages ’e.g. insu'ficient discussion, meetings, participation, infers ation, etc.) Likely to come up as a reason for ncn-adopticn; e 9., “the meetings bogged down a Tot”, ”we didn't have e .ough background info. on Ef,‘ 'the right peopie weren't involved“ , etc. APPENDIX B: Coding Form for the Organization Profile _ BCCCIQVSJ m U l’ T] C ) two "'1 t-«o I F) l70. l7}. l72J *3 ‘NJ 0.) 7J3*1‘l5"i,i p; ::-g 1., . hate. ' . o g nFfl‘. - ’0‘ "__. _, ‘ PA Code tnis during tn ngu wnen v u Jet ie3~va*: in*u"- . " 4 la - .. - .u matton, , when R ois;Js ses SJooor.,Jn:agor so or;"uwt” '0“? :l””V 7 "too oTs nJML VL.H: R 5L INFJU:J.; INFLUinCE & :JvrC\.: ..,.... I, .. (.7. . “;r~':';‘_.e..i._f—“~ Us: RUE5E RJ.HMF VKUDJ. «.ruv-r— r; e n (‘ sgrgrr‘fi r‘rx. gem“ tainr'xv A ~ r'; EVIL \ K “m3 .1 JCQJJLL' =3“ JTQ 111.4.VlUHJr‘Lb‘ DI‘I llvi “ ' " -V -. :Yfi“~' f‘fifllgF .3~.',y‘§'\"‘ A ‘P"1\lr.‘,‘ WO'JIC‘ ~UU S") that eJ":/u 3’ 1b} I IAugngJ'VIhL. 1.\ -‘~-.~.-C\"LJJI ~ I he; a rinira‘ moderofe, 3% cre=t deal :‘ influerge on vh --_. _.....—~—. inc: Fail” i u-Is- ‘J‘.\r » r ~»- on u r~ *- ~nm~¢ °mwvv+~u" ~r~*"v#n :v o [4‘ lht. ‘w‘: ‘|i‘ F F, Z” I”. 1;.” 'V {'1 il‘L‘A 1‘ .pfiln 5. ‘-~. \' . ‘. "’. 1.. LJV‘ H , qr:- fi:CDAI‘ Van Ts: pt '\'-I “f! (N \4V‘u’s. th‘ \r 5" :n’ :‘l r" a L1~-L.-H fie inf] no' V nimai fiooe~ate e : iei ,(‘r‘ P“ (V‘ {1! f") "."’:‘ .C,nr~r\ ‘: ‘IV’:‘| ((20 “’- I\\) DJ- 1: :l“ in meme .: .55. or: \ : a4; J7: E‘O'EG {-1 q u l a 3 o o O ' 1' .i O l a J S (Po ingBudget) p‘ I u l t \ 0 l 2 J SO Administrative, Executive, -..J .—.J [‘0 b) (A) D) D) (2‘) (A) Union Parents, indiv.duals; Q_Front- Q_Front-line st eff {Potentially uirectly in F8“ 9 staff (Not directly ‘Ub Q_Administrator(s) (D (D ('1 A) h “h "s "I? O (“f (D .3 (1* ._J i) - .1 (f Cirectly involved} p involved} involved) n+1“ ”community“ people (as ial community es, code spec GROUPS below) Clients Special Community Groups by B '3 12 Other (W‘iO?) Total number of group/individual levels mentioned Cl, 2 = 02‘ .7 l FF ‘ - A ’ 1‘ ‘ . £§-.;.“ ‘ *, L; . {Vlr‘ v'~‘~ ‘\l :' 4-? r.~r% -0\ have on jéClSTCflS reéete: to stzttzng new pctfiramst :;::é ATr‘h’ ’; ‘ "ir‘ ’- V ‘ A." ‘\A aw 5 ~ , A\/ n‘- ‘\ '- I' ' n 0n LiJlCH~i qu, from Yl‘lS pd t:LJ.A? ulholo L QtH9” WQTGSx What QYCJDS . h :4 .J, H i “’N ha + -L or inc ividuals wooio ytJ saJ J, btdEnhL had he mos. e u'nn; : ‘; — m - 5577.. x J ' v N no ‘EHCFNJ influence on Jrcgrau cnange .n yOur 0 at tie time of t*e AJ: ”.- -—.._ fy‘: q .tp- - W . - f n F. O. - :c“ . Q-‘Q g“‘ K n: FGPUCtaKC :U annher bECddse OT d.i*€”ent CL fl? ,1 - ‘A'H- - ‘ s ‘ .L ‘ , o l‘. r. .0 ’- I I" ‘l . ‘ - .- ‘ ‘ - .n Jimensions, fird cue anion o.m:nSion,leve: characser‘ze: most programs whicn are started by 0, and fill oo: nutrix —-.—-——v-—v . ’ j .- accordingly.; -' v- - #4“. 1"-_" “7:57 7,6-rn1'9f‘ 9'” in‘.(\'(‘ p 1‘? 1 ,fi‘ .. /‘;‘ ,J" ifrerentiet; 2 I79. JEQ: JJ. E,~;;H.JnildJ JLJERSJJJJ mENJIONED bf P? ges\l} n. _ : ‘ -- ‘.‘ ”.H!fi" ur-q— "':'\- c menSions; Whnl nth; inor? 1v“ '."‘ TC”;7" '~ rrT- Y\l"l“,"' Pm. 1:2 R T7315 L‘thJ U Ve‘r\1,/‘L.IUXVLLJLJIA«LS’ Sfiilj. :. _, fl ,_, 4.; -. f'frqfnr r filly-raging h nr-u *"Vfiau"r‘.: -I"CL]‘J YOU the 36“. Ll‘ult L: 13-.) GRCH Lib/1M Utv'p'iil. 1* It...‘ .1‘-‘lL..,/' a" r’“ W “‘1 M fJ‘V‘" ’3 ‘ ny~'~‘+ "1 ' ‘; Y“:~Il4 , Eco a .10 kc:, '-:. ‘x k N ." r; i' Y‘}‘ \ "( 1.,4 \ w \s. u _ .l v v ‘ v 9‘ ‘ '0~, s. - ‘~- - ~— -‘ ‘ ‘4 H _a — 7"--. .___-__, 7.- —— —Q— —— -<—~ h.,-'~‘-\’. A-‘—- 'r‘ \ r. ' . a, ,. - a ' ' ‘ '2‘ ’ -r‘ t *q : . '- L "r‘ . ptfiem L ~ .. o .2 . - ’ - c - \ a" {3‘4— '—J l.- L“ t _4 -m- ____‘ __ .__ .. _ - (sf- 2 3 - Q ~ ~‘ L. LE1" al. _ vt‘ - a U .— ‘r r- ’ . ~ v A ~ - . ‘~”- “awe o? C lrr i: ‘r .a ’« .1.” _ .\ l: I l... L i-e. = \w. ' » New ..__ ___“.' “—4.1 _.._..__ __ Name of sg o I. A (“I “ f. | ‘. t- ‘ . a: r \G- A |9J~éOL. Number or (CLI;N*S) in C a, the ‘ ”0 Cf deCision t0 Use/Wet 33“ n _,. r) by Y "P'*" l '- .f“: ' T I . - _,- . ’,.‘ .m P A . . .A ‘,-- -u:-haw. Numoer Ci leclzn.:/ in :. 3f the time or d::ision to use/no Jr: is) -.., ”w.— -._,_.______ _ , ._.I ll ‘..A D n “3 (1) CJ U3 - J. :3 LC: P 3 ll ( 1 D ( ) ‘1 [Ti (.1) if} —I .5 L1 ( ll Ln ( f (.1 f ID .1 ' (1“ 7, 3) r_. f”. - l ‘C f1! (‘1 adwinistretors c“" ' --“ v ’3 L ”Aft-1:7") l pfi_ A ' .- _,’~‘rr ." 'A a“ ‘ I . . .I.‘ ‘ ‘=-~ H» number of administrttiv; st;r* :n l at tne time of deCisicn to {C ”in-lg” pErsGrr‘é. ot use 3 \H :3 Zlfl. For 208 AND Ell, at the time if decision to use/not use 3 -—-— ....—-..._._. .4..— l=1ncreasing 2=Decreasing 3=Stable (ASK ALL THREE!) “’ “fits l.’ (It’s L?CY|’\ . ~ >J ~" ‘~ ‘l . , P\ ‘ ~ / I§r.;)"‘\ Li), ‘ .. \‘Z‘f" Cp;;~ .. ‘IQ '(‘1 If‘, A z i" ., A ~ . _ . J ‘ I Ex. Ika.» u..u3 -' Lr'v\\' Ore! L , A, '\ *‘~ ~- v 0- i-\ - .1 . .~ Ov "F‘O, I. r5 ‘5 ‘r r—‘o - 9 .‘p, »‘ p-‘f f, ‘n 5L» ‘5. ‘:-l"’ ' Feta" b; “ w "J \—| - \- 56‘ : 2-x. “ .4 Us . Tap. ,r-p :: ‘ UJV “ L U:L AH.‘ ' "fi ‘1‘. ’ r9 4 5; :L. "- ‘J ”J? H ‘1‘! ' -V 7 ‘ I 1‘ )0) b .. g P r o . ‘ u. \ . i (Q 99 y ‘r f I'- ~L.» .‘J IL.3" P -.‘: “I"- I, ,L 1 ..... L4,, l ._ \ ‘2', __l "" ~(‘ n' r h L" La~€ v. 54 x . - w \ fr‘; - r ' A -_r- .’ —‘ § 0- 0,;- 7" , ~~ a- t. r- L-‘ \— - '4 L- .- 5.I—‘_ C» b e L] "v v LIE. ., _. v- 4») av- '5 ‘- fA LJ" 7‘" ”_" AV»: 7 r ’J- o L, 1:! v ",C. -. -Kau ‘« P‘s/e gug ‘H J‘V'SVI: ’V‘ L- 5 - ‘ ‘\ l A \ _ - \~ ~-,- -~ "‘1/"“ —/ ‘27 r ~ “our . adosiad ani ‘ther rocr*"s ‘r the ‘a" w‘r'7 L... 3C.» Id.) V P t. '...l L) p :- Gl II x. » Iv: _, V~C D. l :: V’ '3 z: p 'LS C :0 HEARD OF OUR 228-229. Have you ever heard of, or do you know if anyone else in your PROJECTS OTHER E‘s‘ organization, has ever heard of any of the following programs: Code l. Heln Ore Student to Succeed (HOS;S) h) 'I (‘ aces Dissenination Project (I) Experience Basgd Career Education ’7 (W (.2 e. a %S Curriculum Management System CD bx) Q0 L‘ ‘ .4 CA) ' \3 ( II) 9. 2 8 4 l4. l, 2 3, 4 l5. ‘.3.4 233. To what extent would you say your §_h’s T (EDUCA.IDN: The National Diffusion Network/ CRIMINAL JUSTICE: The Exemplary Projects Program of LEAA) ? oan contact l=minimal contact 2=moderate contact 32R great deal of contact terminaturs, set up the possibility of future phone and/or nail contact. \ site visi*;. Make sure you have a SUMMEI phone a 8 address! APPENDIX C: Interview Guide l) When 5 the la st irne you read the AP 2) What are the validated outcomes for this P2 INTRO. (YOUR M E). . . Center for Innovation Research . . We're interested in how new programs get started in organizations, And we're talking to quite a few organizations about a large number of new programs. One of the many programs we’re looking at is (NAiE OF 5). Do you have a few minutes to talk now? PROGRAM -Have you ever heard of E? STATUS ‘Are you currently using 2? ’IF v53; When did you start using P2 'REBE ’ IF NECESSARY). IF 0 IS UNAUARE RON-ADOPT ER, GO RIGHT TOu A -----u--—------~--_--— a”“------”‘-‘-----------—--—------~------‘-~—------——--~--'----- GENERAL INFO. MENTION THIS INFO. AT ANY POINT IN THE INTERVI EH IF IT IS NECESSARY, ABOUT US BUT DON'T B lTHER UNLESS R SEEMS .iERUOUS, IS CUP IiUS, ETC. REASSURE re. We are not evaluating your 0 or P; we re interested in how new ”E”ALUATION programs get started in organizations, and we re looking into ANXIETY” a large number of programs and organizations in this study. REASSURE re. Let me make it clear that all reports of our project will CONFI DENT.l .ALITY refer only to groups of organizations; the names of organizations and of respondents will NOT be used. CONVEY TO.E THAT THEIR RESPONSES ARE VERY IMPORTANT TO OUR PROJECT. FIRST CONTACT How did your Q first learn about P? 2.1RVDLVEMENT When did you first become involved with 32 To what extent were you involved with the original decisions involving 22 What was your job position at the time? Hhat is your present jOb POSlthH? TON LOCUS As far as the decision to USEA’POT USE P was concerned, what individual or group would you say had the final decision? (Use “Rubber Stamp“ probe if §Q_is mentioned here; i. e. “Do they pretty much approve all of the programs they review”) 131 REASONS FOR PROS S f?! has there anything about RELATIVE SO FOR THIS Q. IF YOU'RE NOT SURE, E.G.: y I‘m trying to get a picture of how your organization fits in with other organizations . Is there another organization that your organization is responsible to for decisions. an organization that's a ”level ah ve” your organization, like a district—level If it's not clear by now don't forget to ask the fo where in g ing being used? Where in SD What O (in SO) adopted_§ lst.? which was the most involved in AD? .------------------.¢----—-----~--------——--—‘-“—‘--~-‘-c---—--oo-’----— 1') Now, going hack to the time the decision to {USE/NOT USE E was made, I‘d like to know some of the reasons that influenced that decision. What were your reasons for [USING/NOT USING] P? AFTER FIRST RESPONSE: Were there any other reasons? MAxE SURE YOU UNDERSTAND REASONS SO THAT YOU CAN CODE THEM. USE 5;; PROBES WHICH B HAS NOT ALREADY MENTIONED: which had an impact in your decision making? 2. Characteristics of.g