140 800 HTHS *- ‘7'! « If. y_:‘4.‘5 A - -' .-. (i3. 9|firwa' A t v. .- F" t I.'\ j “VF ‘\ PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 2/05 m Mini-p.15 DISSEMINATION AND UTILIZATION OF KNOWLEDGE THROUGH EXPERIMENTAL SOCIAL INNOVATION by Jay Gardiner A Plan B Paper Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF URBAN PLANNING School of Urban Planning and Landscape Architecture 1983 CONTENTS Introduction PHASE ONE: EXPERIMENTAL INNOVATION I. Background for the Proposal II. Description of the "Traditional Model" III. Description of the Proposed Model IV. Hypothesis .V. Experimental Design and Outcome Measures VI. Participant Variables VII. Social Situational Variables @. VIII. The Research Team 3 IX. Administrative Agreements X. Instruments XI. Sampling XII. Experimental Constantcy XIII. Data Collection and Analysis XIV. Conclusion PHASE TWO: DISSEMINATION BY ESI I. A Plan for Dissemination II. The Research Team III. The Approaching Phase IV. The Persuading Phase The Activating Phase 10 12 12 13 14 20 21 22 23 25 25 25 27 35 36 37 38 VI. VII. VIII. IX. 1. The Diffusing Phase Assessment Devices Administrative Agreements Conclusion TABLES Dissemination and Utilization Perspectives PHASE ONE: EXPERIMENTAL INNOVATIONS 2. Experimental Design Outcome Measures Participant Variables Social Situational Variables . Experimental Plan Reliability and Validity of Instruments Instruments Used to Test the Hypothesis PHASE TWO: DISSEMINATION BY ESI 9. 10. 11. Experimental Plan Social Situation Variables Assessment Devices FIGURES Havelock's Need Reduction Cycle Dissemination Experiment by ESI 40 40 42 43 Page 7 12 13 14 15 19 24 24 29 31 41 34 46 INTRODUCTION The dissemination and utilization (D&U) of new and innovative knowledge plays an integral role in creating social, technological, and organizational change. As a practice, dissemination and utilization of knowledge is an age-old phenomenon, however informal or unplanned it may be. For example, the diffusion of technological innovations (e.g., gunpowder, horse-drawn plow, penicillin, etc...) from one culture to another was the saving grace for many early recipient cultures. Although in many of these cases, the D&U process was not a formalized one, its importance in creating change, whether good or bad was undeniably great. As a field of study, the dissemination and utili- zation of knowledge has remained relatively fragmented. Researchers in several fields have developed D&U models (e.g., Miles, education, 1964; Havelock and Benne, industry, 1967; and Gallagher, culture, 1964); however few have taken a macroscopic approach assessing the entire process from basic research to widespread adoption and subsequent evalu- ation. Instead, they seem to focus on isolated aspects of D&U theory, whether it be research, development, inter- vention, diffusion, or adoption. Havelock (1971), identifies three distinct per- spectives of knowledge dissemination and utilization: a (l) the Research, Development, and Diffusion (RD&D) per- spective; (2) the Social Interaction perspective; and (3) the Problem Solver perspective. The RD&D perspective places heavy emphasis on the research and development of knowledge irrespective of con- sumer demand- The D&U process begins with the research and development of knowledge and then proceeds to trickle down this newly found information to interested consumers. Havelock (1971), says of this process: "Although consumer needs may be implicit in this approach, they do not enter the pic- ture as primemmotivators for the generation of new knowledge. Research does not begin as a set of answers to specific human prob- lems. Rather, research starts as a set of facts and theories about the nature of the universe, knowledge which can only be made useful to men through an extensive process of development. In development, basic theories and data are used to generate ideas for use- ful products and services and these ideas are turned into prototypes which have to be tested and redesigned and retested before they represent anything that is truly useful to the bulk of humanity." The R&D model is typical of many scientific studies. Emphasis is often based on laboratory research, and the diffusion of information to potential users is largely passive. For example, biomedical and agricultural research appear to more of a knowledge seeking activity rather than a solution to a specific set of problems. ‘iv The second model, the Social Interaction approach, focuses on the diffusion aspect. Social Interaction theo— rists view society as "a network of roles and channels of communication with organizational and formal and informal associations forming barriers and overlapping connections." (Havelock, 1971). One of the most recent and complete works on the subject is that of Rogers and Shoemaker, 1971. The authors cover in detail the main elements in the diffusion of new ideas, with the major thrust being that communication is a necessary requirement in order to undergo social change. Proponents of the Problem Solver approach emphasize an active role of the outside change agent as an impetus for the desired social or organizational change. This model is social-psychological in nature and is highly character- istic of the human relations school of organization theory. Typically, in the human relations studies, an outside change agent acting in a collaborative manner with the client system, identifies a need and systematically searches for an optimal solution. Havelock (1971), likens this problem solving process to the psychological theory of need reduction. This type of theory is exemplified by Maslow's theory of human moti- vation. (Maslow, 1943). As Maslow based his theory of human motivation on the successive attainment of basic needs, Havelock identifies a "need reduction cycle" in which various needs may be systematically filled. This need re- duction cycle (Figure l) is a six step process beginning with the identification or "diagnosis" of a need and ending with an overall need reduction and subsequently a new tar- get condition. This process of constant reformulation of system needs or "objectives" so characteristic of the Problem Solver perspective is labeled "action research" by Thelen (1967). Thelen succinctly describes this process as the "application and correction of policy adopted to specific situations.“ His model, however, includes one aspect not addressed within the needs reduction cycle. Thelen's action research process begins with a determination of the crucial variables relative to change adaptiveness and analyzes their relationships within the total system before advocation of a specific change. This analysis is done through three related procedures: "(a) extensive 'life-space' interviewing of principals including change agents; (b) initial participant-observation experi- ences of the researcher; (c) assessment instruments and questionnaires which become standardized and used in a wide variety of organizational-change situations so that we can compare situ- ations and talk together about them." (Thelen, 1967). What has been lacking most of the RD&D, Social Interaction, and Problem Solver models is a scientific Figure l Havelock's Need Reduction Cycle 1. Felt Need 6. Need Reduction 5. Application of 2. Articulated as a solution problem 4. Choice of solutions 3. Search for solutions methodology for evaluation of alternative approaches. In order for D&U research to advance and increase its meaning- fulness to those involved in creating social change, there should first evolve some consistent methodology for dis- covering what approaches work and why. Fairweather (1967), developed a model to address such problems. His paradigm, labeled "Experimental Social Innovation" (ESI) was designed to "determine the degree to which selected variables influence innovation and dis— semination." This approach is quasi-experimental in nature and is applied to successive stages of the D&U process. At each of these stages, seemingly significant variables are either controlled or independent and the results of experimentation on this basis are evaluated for future study. A summary of the ESI approach, along with the three previously mentioned D&U perspectives is given in Table 1. In the remainder of this paper, I will take the reader through the ESI process as it could possibly be applied to the dissemination and utilization of an innovative ride- sharing model. a»: PHASE ONE: EXPERIMENTAL INNOVATION Jump “waonm noaumuom "Esau boa “udddmduflmduudIAMHdum 7 333%: .coflumowcsaaoo can cowuomumucfl Heaven HmEHom IGH 0cm HmEMOH mo maoccmso canons“ mommoonm cowumc IflEmmmHOIcoHumocfl pmfiwflommmca .mxaa on» cam .mamom mo :oHumH Isowuum .auHHmHo 0H0“ .mnuomfim .cowamoacsaeoo Hmummum m>onmefl ou mxuoz .mmuwm too can momma m.cowumuficmmuo co mmm500m .oocoflosm ummumu Umumonmucfl cm on cmumcfiEmmmfip ma Hmpoa 0:» usonm coflucEHOMcfl Hmcowumm .Hmcoa on» no coaumsHm>o HMEHOM :0 mammnmam cmuflefiq .GOADMNflcmmHo pomuou Ho cowumHsmom unease scum names manna; .pwuomcflmcm Ho comm Ixomm mw H0008 m .mumc “an no momuooo .uonvwmbnfidm “mounom .mxno3 paw .ucmaow>coc ma .Hmownuw ma pa mcw©fl>oum mumefluflmma ma may“: Inomu and .maamucma nanomxm commafioo mum HmQOE pmvmcwao> man «0 coaumoom mcflmnMMAO can mcaum>fluom .mcwpmsm lama .mcwnomoummm mo mmo3 m>flumcuouam .cowu ImcwEmmmflc ucosvmeSm mOCwEHmumc cowumsHm>m .ucofiwummxm Hocacsuflo Izod HmEuom m ca manna uoHflm o co coucmEOHmEH ma Hopoz .mcficcmam cw cowumNficmmuo ummumu can .xoonpoom .cowuma rsmom ucowao .noummm Imu mcwm: Emmy Hmm cowuwcflEmmmwn a :oflumoea Hone: HH mmmnm ucmEmoHo>oa a :oHumoHO Homo: pmwmwommmco u:mmo.mmcmnu nouomwmu Gamma mafia: an Umummuo ma Hmpoz H mmmzm cofluomumucH Hm>aom :oflm3mmwn can name coaum>occH Hmwoom Hafioom Emanoum Imon>oa .nouoommm Hmucmeflnmmxm mo>wuommwumm cowuoNflHfluD can newumcflsommfia H wanna 1m. . my I. BACKGROUND FOR THE PROPOSAL In this automobile oriented society parking needs have developed wherever pe0p1e congregate, especially in downtowns where parking demands have risen more rapidly than their daytime populations. For a community, an under- supply of parking in the Central Business District can affect the public's entire expectations of an auto trip downtown. In a national survey of public attitudes towards downtown parking, 60% of the respondents statedlthat they felt that downtown parking was relatively inconvenient (Alexander, 1974). Whether or not this assessment is true is insig- nificant in relation to the effect of public perceptions. In this sense the."parking problem" is more of a "percep- tion problem". The high demand for downtown parking puts these spaces at a premium. In 1975, commercial parking rates in the City of Los Angeles were reported at $35/month per space (Shoup et a1., 1982). These rates are typically paid by the business rather than the employee using the lot. Pratsch (1979), broke down the total trip costs for a downtown employee taking a ten mile trip. He found that approximately 36% of this cost would be paid by the employer in the form of parking expenditures. This compares to a 22% figure for the driver's out of pocket costs. It can be seen from this that the driver often does not pay his fair share of the driving costs. One means of reducing work related parking pres- sure and expense has been the practice of carpooling or ridesharing. Although successful in many isolated in- stances, it has not been implemented to a great extent in the form of urban-wide or regional programs. (Pratsch, 1979). Two of the major difficulties inherent in implementing a prearranged ridesharing program are: 1) "lack of know- ledge concerning ridesharing; and 2) lack of favorable atti- tudes towards ridesharing" (U.S. DOT, 1981). Lack of knowledge concerning ridesharing can be inferred to include public officials as well as the general public. Study of the reasons why people will or will not share a ride on a regular basis is necessary before any worthwhile implementation takes place. Traditionally, the use of priority parking spaces has been used as an incentive for ridesharing. This strategy, however, is not necessarily the only or best technique to induce carpool activity. Shoup and Pickrell (1982), designed a ridesharing strategy based on direct financial costs of parking as a user incentive. In this section I will formulate a research proposal based on an elaborated version of their model. II. DESCRIPTION OF THE "TRADITIONAL" MODEL In the traditional model of ridesharing, priority parking is offered as the inducement for participation. 10 These priority parking spaces are generally the closest and most convenient to the place of work with a specific space assigned to each registered carpool. No other in- centives are offered, and no stipulations of carpool size are made. III. DESCRIPTION OF THE PROPOSED MODEL The innovative ridesharing program is based on direct financial incentives related to the cost of parking in the Central Business District. Within this model, employers leasing parking spaces from the municipality will offer parking subsidies to employees who elect to carpool to work. This innovative method of ridesharing inducement will be ' comprised of a subsidy rate corresponding to the number of people commuting to work per automobile. If the organi- zation pays $50 for each employee parking spot per month, then a carpool of two persons could be expected to decrease the employerfismonthly expenditure by that same amount. Under this proposed model, the $50 saved would be funneled back to the carpool in their paychecks constituting a parking subsidy of $25 per carpooler. Likewise, a three- person carpool would createa.$100 parking subsidy to be split three ways ($33.33 each). It should thus be apparent that the greater the number of people belonging to a par- ticular carpool, the greater the parking subsidy to any 11 individual member of that carpool. In this manner, the proposed model is designed as an incentive to create more and larger carpOols. Those employees who choose not to join carpools, and drive to work alone, will incur the same parking costs as before, and are no worse off. To act as sort of a policing function to alleviate fraud of the system, carpools will have to be registered and will be issued a window sticker for a carpool of that particular size (e.g., a two person carpool would have a red sticker; a three person carpool would have a blue sticker; etc...). The parking attendant will issue to any carpool auto with the proper amount of people in it a vali- dated parking receipt. These parking receipts will then be turned in to the ridesharing office per pay period so that the parking subsidies may be calculated and paid to the carpoolers. A carpool without the proper number of people in it as specified by the carpool window sticker will for- feit that day's parking receipt. Employers will not experience any change in parking expenditures, however, they will be able to reap the bene- fits of increased employee satisfaction, additional parking for customers, and the greater social benefits of decreased congestion, pollution and energy consumption. IV. HYPOTHESIS 12 Employees exposed to the proposed ridesharing pro- gram based on direct financial incentives will be more likely to carpool than those exposed to the traditional priority parking program. V. EXPERIMENTAL DESIGN AND OUTCOME MEASURES The proposed ridesharing program will be compared to the traditional program in terms of population and sample demographics, employee attitudes, and parking and economic data. (Table 2). listed in Table 3. Traditional Program Financial Incentive Program TABLE 2 Experimental Design The outcome measures of this program are Pre-Implementation Demographic Data Parking Data Attitudinal Survey Demographic Data Parking Data Attitudinal Survey Post-Implementation Demographic Data Parking Data Attitudinal Survey Economic Data Demographic Data Parking Data Attitudinal Survey Economic Data 13 TABLE 3 Outcome Measures Social Change Outcome Criteria: 1. % of eligible employees using program; 2. % decrease in the number of parking spaces occupied by employees; 3. % of passengers who are from non-car owner households; 4. % change in carpooling. Performance Measures: 1. Cost/Revenue; 2. Cost/Passenger; 3. Passenger/Automobile. Attitudinal: l. Willingness of people to rideshare if this program were not available; 2. Satisfaction or dissatisfaction of employees with the program. VI . PARTIC IPANT VARIABLES Participant variables for the programs are listed in Table 4. All of the variables listed are independent as they cannot be controlled in a study of this nature. These variables will be measured in the pre-program and post-program demographic and attitudinal surveys. 14 VII. SOCIAL SITUATIONAL VARIABLES Social situational variables for the programs are listed in Table 5. These variables, both internal and external, can help to define the social parameters of the program as well as provide a basis for replication. TABLE 4 Participant Variables Independent Age These variables may effect a Sex person's willingness to car- Race pool with others on a regular Religion basis. Income The effect of income on desire to take part in program. Ability to drive Automobile availability for trip, based on: 1. automobile ownership; 2. condition of automobile owned; and 3. need to share car with other member of household. Alternatives The existence of alternative modes of transportation (e.g., bus, walk, train, etc...). 15 TABLE 5 Social Situational Variables Organizational Components Hierarchical structure Size Complexity Formality/Informality Group Dynamics Norms Both of the programs will be run by staff members under the general supervision of the project director. The staff members will be responsible for the for- mation of carpools and the daily monitoring of the program. The hierarchical structures of the two institutions will be nearly the same. Size of the organization will be approximately equal in each program. This is based on the number of employees and the per- centage of employees driving to work. The proposed ridesharing model will be somewhat more complex than the traditional model as it requires careful monitoring of parking receipts, and will involve the assistance of the parking attendant. Both conditions will have the same degree of formality in terms of marketing and carpool formu- lation. This should be nearly identical under both conditions; however, it is expected that a greater amount of peer pressure will be existent in the proposed ride- sharing model. These will be established inde- pendently by each group. It can be expected that norms will be nearly the same between programs. TABLE 5 (Continued) Leadership Composition Morale Reinforcement Fiscal Processes Income 16 As with norms, this is in both conditions a product of small group dynamics particular to each group. Participants in each program will be grouped by home residence. No attempt will be made to match riders on any other basis. Par- ticipants in each program are limited to the employees of that particular institution. This would be more of a fluctu- ating variable in the proposed ridership model due to the higher degree of peer pressure. It can be expected that morale will be the highest when the carpool was full and lower when the carpool subsidy is forfeited because of group member absence. In the traditional model morale will be more stable, however at a lower level than that exhibited by the proposed ridesharing model. In the traditional model this will be in the form of better parking spaces for those who elect to carpool. In the proposed ridesharing model, reinforcement will be in the form of parking subsidies to the carpoolers. Internal reinforcement or peer pressure will exist to alleviate carpooler absenteeism. In both conditions reduced gasoline ex- penditures, automobile servicing costs, reduced driving time, etc. will help reinforce carpooling activity. Funding for each of the programs will be obtained through grants from the Department of Transpor- tation. TABLE 5 (Continued) Cost Membership Voluntary/Involuntary Turnover External Variables Social Climate Specific Socioeconomic Indicators General Socioeconomic Indicators Geographical Location Folkways and Mores of Community Time Folkways and Mores of Institution Measurement Objectiveness 17 The costs of each condition will involve primarily the administra- tive costs of operation. In addition, for the proposed model, costs will be incurred through payment of parking subsidies. Carpools in each condition will bejknmmaion a voluntary basis only. Participant turnover will be an independent variable in each condition. Turnover rates are expected to be approximately equal. Measures will be taken to assure consistent staff over the experimental period. . The two conditions are being implemented in the same community over the same period of time so these variables Will be identical for each condition. The two conditions will be imple- mented in institutions of the same size and type, thus it can be expected that these variables will be approximately the same for each program. Regular monitoring of the pro- gram's progress by the project director without directly im- pinging on any of the processes will take place to insure that the program is properly measuring the desired variables. TABLE 5 (Continued) Publicity and Media Exposure Legal Constraints 18 This variable will be held con- stant. This is to be insured by administrative agreements with the institutions involved. This will be examined before implementing programs to assess their legality within the par- ticular municipality. TABLE 6 Experimental Plan Program Phase Time Schedule Planning minus 9 months minus months minus months minus months minus months minus month Program Initiation 0 months Program Operation plus 1 month plus 12 months plus 13 months 19 Activities Develop research proposal; Apply for funding. Initiate negotiation of administrative agreements; Develop instruments. Test instruments and modify where needed. Finalize administrative agreements; Interview possible staff members; Select acceptable insti- tutions. Train staff members; Set up office space. Gather archival data from each institution. Distribute pre-pro-. gram attitudinal and demographic surveys; Begin intra-institu- tion marketing programs; Begin intake of par- ticipants. Begin matching of par- ticipants by address. Registration of formed carpools. Conduct attitudinal and demographic surveys. Analysis of data. m. 20 TABLE 6 (Continued) Program Phase Time Schedule Activities Program Operation (continued) plus 15 months Analysis of per- formance measures. Program Termination plus 18 months Final Analysis. Decide on further implementation or research. VIII. THE RESEARCH TEAM The research team will consist of 1 project di- rector, 3 staff members (research assistants), 1 part- time statistician, and one economist. Project Director The project director is responsible for the over- all coordination of the programs including: securing fund- ing, selecting and training staff, establishing administra- tive agreements, determination of validity, etc... stag: Two staff members will be responsible for the daily operation of the financial incentive ridesharing program. This includes data collection and analysis, parking receipt collection, and subsidy determination. The third staff member will be assigned to the traditional ridesharing pro- gram. She/he will be responsible for data collection and 21 analysis. In both programs, staff members will perform an ongoing marketing campaign keyed to their specific market populations, to inform, encourage, and convince them to use the service. Statistician The statistician will be used on a part-time con- sultant basis only. She/he will aid the project director in evaluating the progress of the programs. Economist As with the statistician, the economist will be used on a part-time consultant basis only. The economist will aid the project director in determining future trends in Central Business District parking costs, as well as possible revenues from longitudinal dissemination of the proposed ridesharing model. It is possible that the statis- tician and the economist be the same person. IX. ADMINISTRATIVE AGREEMENTS The project director will have to negotiate several administrative agreements prior to the start up of the programs. These agreements will be with the municipality and with each of the two institutions involved in the study. With the municipality, agreements must be made to allow the cordoning off of a "priority zone" for carpools in a public lot; and for the use of municipal parking attend- ants to monitor carpool activity. 22 Formal agreements must be made with the two insti- tutions to insure that the research team can function un— hampered within the organization in the collection of data, and in the marketing of the program. Additionally, agreements must be made with both the muniCipality and the institutions to: allow the program for the full stated time period; provide access to pertinent records such as parking receipts and costs; demographic characteristics; etc. An agreement must also be made with the muniCipality and the institutions to maintain secrecy as to the research nature of the programs. X. INSTRUMENTS Pre-Program Measures Demographic and attitudinal questionnaires will be designed prior to program initiation. These question- naires will be administered by the staff members. The questionnaires, created to determine employee perceptions of ridesharing programs, will first be pretested on a small group of local citizens not possessing any special know- ledge of transportation. The study populations will be given the pre-program questionnaires as a paper and pencil measure distributed with the employees paychecks or newsletters. Reliability will be tested by comparing the collected data with demo- graphic records supplied by the employer. 23 Post-Program Measures Post-program measures will also involve the ad- ministration of pencil and paper attitudinal surveys. These questionnaires will be designed to measure employee satisfaction with the ridesharing program and to note any changes in transportation behavior since program initiation. Additionally, economic records will be used to de- termine such measures as cost/revenue, cost/passenger, and passenger/automobile. Staff members are responsible for the keeping of such records. Further discussion of instru- ments is given in the Appendix. XI. SAMPLING 'Participants for the study will be from two distinct yet similar populations. For comparison purposes they will be nearly identical in number and in type of employment. No attempt will be made to select or randomly assign employees nor will the groups be artificially limited in size. Par- ticipation in the programs is voluntary only. By allowing general sample populations inferences can be made which are applicable to real life situations. 24 TABLE 7 Reliability and Validity of Instruments Instrument Reliability Validity Demographic Data Inter-rater Correlation of survey reliability data with data ob- tained from other institutions Parking Data _ Inter-rater Correlation of the reliability institutionfis parking data with that of the municipality Attitudinal Test-retest Correlation of pre— Questionnaire Correlation and post-program Coefficient attitudinal data Economic Data Bookkeeping Periodic auditing of Procedures carpool passengers, revenues, costs, etc. TABLE 8 Instruments Used to Test the Hypothesis_ Hypothesis Measure Test of Significance Employees exposed to Attitudinal Fisher's Exact Proba- the proposed ride- Questionnaire bility Test sharing program based Chi-Squared Test on financial incen- tives, will be more likely to carpool than those employees exposed to the traditional priority parking model. vei' 25 XII. EXPERIMENTAL CONSTANTCY The project director will periodically monitor the progress of the two programs to insure that they are being implemented in a consistent manner. The staff members will be thoroughly trained in procedure, especially as it pertains to the keeping of economic records, and marketing. XIII. DATA COLLECTION AND ANALYSIS Data collection will be done primarily by the staff members. ArchiVal data will be gathered from both the municipality and the institutions. The staff members will distribute the attitudinal and demographic question- naires to the populations through the in-house mail system, with instructions as to completion and return. In analysis of this data, pre-program and post-program attitudinal correlations will be observed. XIV. CONCLUSION Whether or not the results from this study prove to be positive or negative, further study will be warranted. Restructuring of the program either by adding or subtracting singular elements, and then retesting is probably the best way to truly assess the nature of the incentive problem. Spurious relationships must be detected and direct positive 26 ones encouraged. If the results do prove to be positive, it would be easy for one to state that an immediate decrease in driving costs is the incentive for ridesharing, however this must be tested many times before proven. As more re- search is done, increasing the reliability, the model may be disseminated on a more longitudinal scale. PHASE TWO: DISSEMINATION BY ESI fy' 27 I. A PLAN FOR DISSEMINATION If the results of the innovation experiment prove to be positive, the next step in the research is to plan the dissemination of the innovative ridesharing model to other organizations and communities. The experimental plan for this dissemination research is outlined in Table 9. Funding will be granted for this experiment from the Michigan Department of Transportation. The intent of this research is to determine the most efficacious means of disseminating the innovation throughout society, and subsequently aid in the dissemination process. Without an advocate, the innovation may be doomed to obsolescence, unnoticed and unimplemented. Fairweather, et al. (1977), state: "many, if not most, social inno- vations will never reach a widespread degree of adoption, regardless of demonstrated success, and often disappear after their initial piloting and development." To avoid this problem, I will create and plan implementation of a dissemination experiment focusing on the various factors which impact dissemination. The most important aspects of this project will be the degree of accuracy of innovation replication and inter-organizational congruency with respect to this repli- cation. In this manner, "the change agent's role thus 28 becomes keeping the innovation intact so its results will be the same as those found in the prototype model." (Fair— weather, et a1., 1974). Table 10 shows the social situational variables which will be factors in this replication. The analysis of the dissemination process will be based on the four action phases of adoption as outlined by Fairweather, et a1. (1974). These four action phases are: (l) approaching, (2) persuading, (3) activating, and (4) diffusing. In addition, the experimental social inno- vation (ESI) method as developed by Fairweather will be used to evaluate this process. Using the ESI methodology, each of the four action phases will be tested experimentally to determine the factors which will impact each phase's relative degree of success. (Fairweather, et a1., 1974). Figure 2 details the ESI format to be used in this experi- ment. A ' ) ‘I- J ‘ ’7) Program Phase Planning L Program Initiation 1, Program Operation TABLE 9 Experimental Plan Time Schedule minus 15 months minus 12 months minus 9 months minus 6 months minus 5 months minus 3 months 0 months plus 1 month plus 3 months plus 4 months 29 Activities Develop research pro- posal. Apply for funding. Interview possible staff members. Train staff members; Set up office space. Contact prospective communities about the program. Develop administrative agreements and instru- ments. ' Approach acceptable organizations; Finalize administrative agree- ments with municipalities. Random assignment of organizations to per— suasion conditions; Distribute approach questionnaires. Organizations asked to adopt; Random assignment of cooper- ative organizations to the two activating conditions. Start activating phase; Initial con— sultant meeting for both conditions; Distribute persuasion questionnaires. (a) 30 TABLE 9 (Continued) Program Phase Time Schedule Activities Program Operation (continued) plus 5 months Action consultant meeting. plus 6 months Action consultant meeting. plus 7 months Action consultant meeting. plus 8 months Action consultant meeting. plus 9 months Action consultant meeting. plus 10 months Action consultant meeting. plus 11 months Action consultant meeting. plus 12 months Action consultant meeting; Diffusion interviews; Follow- ¢ up data analyzed. Program Termination plus 14 months Decide on further research. 31 TABLE 10 Social Situational Variables Organizational Components Hierarchical structure Size Complexity Formality/Informality Group Dynamics Norms The programs will be run by an organizational staff in con- sultation with the regional coordinator. The staff members will be responsible for the formation of carpools and the daily monitoring of the program. The hierarchical structures of the organization will vary from "vertical" to "horizontal," as this variable will not be con- trolled. Size of the organizations involved in the dissemination experiment will not be controlled; however, city size will be no less than 50,000 population. The dissemination program is quite complex, involving a large number of municipalities and organizations. The dissemination program will have a high degree of informality in terms of individual partici- pation. Formality and consultant communication will be much greater in the monitoring of program progress for adequate replication. A certain amount of individual peer pressure is expected to exist. This informal variable should also be affected by the type of organizational hierarchy. These will be established inde- pendently by each group. It can be expected that these norms will nearly be the same between pro- grams. TABLE 10 (Continued) Leadership Composition MOrale Reinforcement Fiscal Processes Income Cost 32 As with norms, this is a product of small group dynamics particular to each group. Participants in the ridesharing program will be grouped by home residence, unless they request to be grouped with specific individuals. Participants are limited to the employees of that particular institution. This should be somewhat of a fluctuating variable relative to the degree of peer pressure. It can be expected that morale will be the highest when the carpool subsidy is forfeited due to group member absence. Reinforcement will be in the form of parking subsidies to the car- poolers. Internal reinforcement or peer pressure will exist to alleviate carpooler absenteeism. Also, reduced gasoline expendi— ture, automobile servicing costs, reduced driving time, etc., will help reinforce carpooling activity. Funding for the dissemination experiment has been obtained through grants from the Michigan Department of Transportation. The costs of each condition will involve primarily the administra- tive costs of operation, including transportation costs, consul- tant salaries, and communications. In addition, costs will be in- curred through payment of parking subsidies. 1“) TABLE 10 (Continued) Voluntary/Involuntary Turnover 33 Carpools will be formed on a voluntary basis only. Participant turnover will be an independent variable. Mea- sures will be taken to insure consistent consultant staffing over the experimental period. 34 a: soaaom Hmscmz ucmuasmcoo cofiuod GmezmmwQ mcwumocfi mow Hmouc5H05 acoz .IIIJIWIII: OZ coflmwoo omucsao> Icoz OZ wusnoonm :onHoo Houomuwa mormxuos mow . . \uommcmz Hmnoqoo mcfipmsmnmm mcflnomoummd Hmm Na ucmEanmmxm :ofiuocflfimmmfln N HMDUHm ‘7'). 35 II. THE RESEARCH TEAM The research team will consist of 1 project di- rector, 20 action consultants, 2 statisticians, 2 office workers, and 1 lawyer. Project Director The project director will be responsible for the overall coordination of the dissemination experiment in- cluding: securing and managing funding; selecting and training the research team; establishing administrative agreements; determination of validity; etc. Action Consultants The action consultants will be available to the organizations in an advisory capacity respective to the administration of the innovative ridesharing program. The consultants will play an integral part in the persuading and activating phases of the experiment. Statisticians The statisticians will aid the project director in evaluating the progress of the programs. Office Staff The office staff will be responsible for filing of program progress information, typing letters, making appointments, etc. Lawyer The lawyer will assist in the drafting of the ad— ministrative agreements and the enforcement thereof. ‘7) 36 III. THE APPROACHING PHASE For the purpose of this experiment, only communities having a population greater than 50,000 will be studied. It is expected that it is these municipalities which will have a more developed and congested central business dis- trict, and thus would exhibit a higher propensity to adopt alternative parking strategies. Letters describing the nature of the ridesharing program will be sent to the city manager/mayor of all communities exceeding the 50,000 population level, and the first twenty positive responsdents will be selected for the program. At this point community commitment to the experimental program must be established through an administrative agreement to insure smooth imple- mentation of the later phases involving the individual or- ganizations (the primary unit of analysis). The next step in the approaching phase is the selection of eligible organizations, within the selected cities, for the adoption of the innovative ridesharing program. All employers within the established central business district will be approached for program adoption. These will be selected from a listing of community busi- nesses as supplied by the municipality. Randomly, one half of the businesses within the central business district in each community will be approached by telephone, and the rm I A- ‘W 37 other half by letter in order to assess the organization's receptiveness to the proposed ridesharing program. As it can be expected that organizational hierarchy may differ greatly from organization to organization, with no uni- formity of positions, the general manager/director of each organization will be contacted rather than varying this factor. Those organizations which positively respond to the approach phase will be started on the next phase of, the experiment, the persuading phase. IV. THE PERSUADING PHASE Of the organizations positively responding to the approach, one-half will be relegated to a workshop condition ’of persuasion and the other half will be in the brochure condition of persuasion, done by random assignment of matched pairs from the previous condition. In the workshop condition, a research consultant will give a pre-arranged presentation open to all employees of the organization. In this pre- sentation, the consultant will describe the proposed ride- sharing program; and outline its benefits to the employee, the employer, and the general welfare of the community. - This presentation will be drafted in advance by the project director and used uniformly across the workshop sample to insure consistency of the persuasion technique. 38 The brochure condition will consist of sending descriptive brochures to each of the organization's em- ployees. Again, these brochures will contain the same information as outlined in the presentation. The list of employees will be obtained from_the organization it- self. Those organizations positively responding to this phase of the experiment will then be moved on to the acti- vating phase. V. THE ACTIVATING PHASE Organizations which choose to go on to this phase will be divided into two more experimental conditions. Again, organizations will be randomly assigned to these conditions on the basis of matched pairs of conditions in the previous phase. -In both conditions, a consultant will meet with the organization's general manager/director to outline the procedures required for implementation of the proposed ridesharing program. In this meeting, the organization's resources relevant to the program will be discussed such as staffing, computer availability, parking waivers, the existence of similar work related transportation strategies, etc. At this point an administrative agreement must be signed by the general manager/director of the organi- zation to insure proper replication of the original, inno- vative ridesharing program. 39 In the first condition, the consultant will give the management of the organization a "how—to" manual de- tailing the exact procedures to be taken in implementation, including promotion, participant matching, and data col- lection. After this point, the organization will be on its own to follow the methodology specified in the manual on the establishment and operation of the innovative ride- sharing program. In the second condition, an "action consultant" approach will be taken. The consultant will make monthly visits to the organization to assist in the functioning of the program and answer any questions which have arisen. Action consultants will be available in both conditions by telephone to offer advice as major problems arise. The need for an action consultant is one of the more critical questions involved in this experiment. The need for some sort of adoption activating process was emphasized by Fairweather, et a1. (1974). They state: "The history of social change gives many examples of promises that never get trans- lated into tangible action, or action that destroys the nature of the innovation that is being put into effect. In short, when one is concerned with the adoption of a highly complex innovation, in an organiza- tional context, one cannot assume the adoption will automatically follow from a decision to adopt." 40 This phase of the experiment should test whether the action consultant is an appropriate or necessary method of adOption activation in an organizational frame- work. VI. THE DIFFUSING PHASE This phase involves all organizations originally contacted to determine the degree of movement towards adoption of the experiment. This phase will consist of telephonic interviews with the general manager/director of the organizations which did not adopt and on-site interviews with those that did adopt. In the former case, the interview will assess the likelihood of future adoption, and in the latter case, it will serve to measure the level of innovation replication. A cluster analysis will be performed to measure the relative influence of the social situational and demographic variables on the degree of innovation adoption. VII. ASSESSMENT PHASE Assessment devices will be necessary to measure the relative success of each of the methodologies employed in the four adoption phases. These devices will consist of questionnaires, demographic information, conference dis- cussions, and consultant research journal entries. These z 41 assessment devices are listed in Table 11. Further de- scription of instruments is given in the Appendix. TABLE 11 Assessment Devices Questionnaires: - About the initial contact of the research team with the organizations; - Workshop effectiveness; - Brochure effectiveness; - Biographical information; - Decision making processes; - Follow-up of both adopters and non-adopters. Demographic Information: - Characteristics of the organization; - Characteristics of the community in which the organization is located. Conference Discussions: - Interview about the persuading and activating processes; - Recording of action consultant meetings. Research Journal: - Record of general program progress, problems, successes, etc. 42 VIII. ADMINISTRATIVE AGREEMENTS Several administrative agreements must be made with the municipalities and the adopting organizations. Agreements must be made with the municipalities to: l) 2) 3) 4) Allow for the use of city parking attendants to monitor carpool activity; To clearly state the intended use of city lease parking as a financial incentive and insure cooperation with the participating organizations in terms of parking waiver dis- bursement for increased carpool activity, To specify time parameters of the experiment; and To provide access to pertinent records, such as parking receipts and costs; demographic characteristics, etc. Formal agreements must be made with the partici- pating organizations to: 1) Guarantee the implementation of the innovative ridesharing program; 2) Maintain monthly meetings with the consultant, keeping him abreast of program progress and/or difficulties; 3) Maintain records of carpool registration and employee participation; 43 4) Provide access to pertinent records such as employee demographics and employee partici- pation; and 5) To specify time parameters of the experiment. IX. CONCLUSION This dissemination experiment is intended to help assess the appropriateness of the innovative ridesharing model in an organizational context; and to aid in the acquire- ment of knowledge about the dissemination process itself. With the experimental social innovation methodology, this research can help determine the important impacting vari- ables on the implementation of the innovative ridesharing model. The best action technique for each of the four phases of adoption; approaching, persuading, activating, and dif- fusing; may thus be determined. Cluster analysis of the social situational, demographic, and organizational variables will establish the validity of replication, as well as indicate a direction for future research. APPENDIX: INSTRUMENTS FOR EXPERIMENTAL RESEARCH 44 INSTRUMENTS Research Journal Intended to record daily observations of the progress of the program. All daily procedures and critical events will be recorded. Historical Questionnaire Reliability and Validity - can be determined by checking against employer records and information received from parents, friends, relatives and acquaintances. (Fairweather et a1., 1977). Pool of Items - can be extracted from questionnaires from iother projects. These items will include such things as age, race, marital status, etc. Concepts to be measured - This questionnaire will be used to determine participant characteristics in order to perform future analysis of program results by demographics. Behavior Rating. To determine the degree of interaction among the organization's employees. Reliability and Validity - coefficient of concordance (W) may be used to determine inter-rater reliability. 45 Pool of Items — to be obtained from a literature review Scale - of related studies. Items will be selected which seem to fulfill experimental needs of determining employee interaction with peers, subordinates, and superiors. identifies particular types of behavior; and a subscale which identifies the relationship of interactants (i.e., peer, subordinate, superior). The scale will be similar to the Social Activity Scale as designed by Goldman (1964): l 2 3 4 5 6 7 8 9 Func. P.S. P.S. Path. Sleep Null NS Act Behav SG 2PG 3PG 1. Path - pathological 2. Sleep - sleeping 3. Null - unoccupied or apathetic 4. Func N.S. - functional nonsocial 5. P.S. Act. - parasocial activities 6. P.S. Behav. - parasocial behavior 7. SG - social games 8. 2PG - two person interaction 9. 3PG - three person interaction For identification of the relationship of inter- actants, raters must be fairly familiar with the study environment and participants. Definitions 46 of each of the three relationships are as fol- lows: 1. Peer - interaction between two or more persons of the same level within the organizational hierarchy; Subordinate - interaction with one or more persons of a lower level within the organi- zational hierarchy. Superior - interaction with one or more per- sons of a higher level within the organizational hierarchy. Instructions to the Raters - to insure an adequate repre- sentation of each of the various hierarchical levels, and increase the chance for interaction, ratings must be taken in a common area (e.g., cafeteria, lounge, etc.). Depending on the par- ticular study environment, raters should: 1. 2. select an unobtrusive observation spot; starting with the nearest person to the rater's left, count the first _____ people clockwise around the room for observation; identify these people and with co-rater record them on the score sheets; going clockwise around the room, observe each of these persons--one at a time, every five 47 minutes for a period of ten seconds simul- taneously with other rater (to be determined by signal). 5. record the observed behavior rating on the score sheet, based on the ratings established on the behavior rating scale. This should be two numbers, with the first number denoting behavior type and the second number denoting relationship of interactants (e.g., 7-1, 6-3, etc.). Attitudinal Scale An attitudinal scale may be used to determine the attitudinal ”climate" of the study population (the organi- zation's employees) towards the concepts of ridesharing and also as a measure of organization morale. Reliability and Validity - Odd-even split-half reliability can be used to control "for any systematic factors operating during the testing period that change the performance from early in the testing session to later periods." (Anastasi, 1976). Pretest- Post-test validity is an appropriate measure for this scale as it is expected that the implementa- tion of the innovative program will affect the employees' attitudes. 48 Pool of items - the pool of items can easily be extracted Scale - from a literature review of related transportation innovation and organization based implementation studies. Items which would likely to be represented in the literature review include personal space, travel time, organization cohesiveness, etc. two subscales will be used; one measuring ridesharing attitudes and the other, measuring morale. The scale will be of a five point Likert type format. The scale will be constructed by a combination of rational and empirical methods as described by Fairweather, et a1. (1977). They state: "After a pool of items has been obtained that logically represents the basic concept with which the scale is concerned a number of highly inter-related items are selected that express clearly the underlying rational dimension that one is attempting to mea- sure. Using a cluster analytic technique...these scale definers are preset in a cluster analysis so that they will define the cluster, which then becomes the scale. (In this way the experimenter (not the computer) determines the logical essence of the scale." Example Question: Ridesharing is a cost-efficient means of commuting to and from 49 work. SA A ? D SD___ (where SA means strongly agree, A means agree, ? means don't know, D means disagree, and SD means strongly disagree). Instructions to those completing the instrument - Below are a number of statements concerning ridesharing (def. - the commuting to and from work by a group of two or more people on a reguarly scheduled basis). Please check ( ) the phrase that most accurately describes your feelings about each particular question. Concepts to be measured - attitudes towards ridesharing and attitudes towards the organization (morale). Scoring of Items - the items will be scored on an interval basis, with scores ranging from one to five; assign- ing a score of five to the strongly agree choice and a score of one to the strongly disagree choice. Analyses needed for the scale - cluster analysis for the selection of items and the formation of the sub- scales. Split-half reliabilities will need to be computed. For validity, correlation analysis will be done to determine pretest-post-test validity. Field trial - will be done on a group of randomly selected citizens of the community to weed out unclear or ambiguous questions. 50 Expectancy Scale Reliability and Validity - Odd-even split-half reliability. Pool of Scale - Validity will be determined by the degree of inter— correlation among scale items; "a predictor will have higher validity the lower are the intercorre- 1ations among its components." (Anastasi, 1976). Items - same as attitudinal scale. One Likert-type scale will be used; no subscales. The scale is to be constructed by the rational/ empirical methodology described earlier. Instructions to those completing the instrument - Below Concept are _____ questions concerning the organization's new ridesharing program. Please indicate which of the five possible responses for each question most accurately reflects your guess about the future of the program. Example Question: I will like the ridesharing program. A great deal Quite a bit Somewhat Only slightly Not at all to be measured - expectations of ridesharing pro- gram participants. This measure is not particularly 51 essential for the experiment, however it can pro- vide a measure of employee support for the program at its outset. Analyses needed for the scale - cluster analysis for item selection. Split-half reliabilities will be com- puted. Correlation analysis will need to be done to determine validity. Field trial - to be pretested on a small group of the organization's administration before full dispersal. Open—Ended Questionnaire (taped interview) These will be used as a basis for item selection. Reliability and Validity - inter-rater reliability can be used to select items from the open-ended question- naire (transcript of taped interview). This will be done by rank order correlations. Pool of Items - to be obtained from the literature; basic concepts will be used from this literature review which seem to fulfill the experimental needs. Scale - One nominal scale identifying the occurrance of specific items on the tape. Instructions to interviewers - Ask the questions in a con- sistent order, allowing as much time as needed for the open-ended response. Instructions to the raters - from the typescript of the taped interviews, rate the items according to the following concepts: 52 1. attitude and knowledge towards ridesharing; 2. attitude towards the organization; and 3. attitude towards the innovative ridesharing program. Each item will be scored on the basis of frequency of occurrance, then ranked on the three concepts listed above. For this process, attitudes and knowledge towards ridesharing will refer to any personal experience or observation of ridesharing in general. Attitude towards the organization will refer to any reference to conditions within the employee's physical place of employment. Attitudes towards the innovative ridesharing program will refer to all items in reference to the experimental program. Analyses needed for the scale - rank order correlation of concepts and items. EssayiQuestionnaire This questionnaire may be used to determine post- implementation participant satisfaction with innovative program. Reliability and Validity - reliability may be determined by Q-sort rank order correlation as done with the open-ended'questionnaire. 53 Pool of Items - to be obtained from that already developed by the open-ended questionnaire. These will be identified in terms of the same three concepts (i.e., attitudes and knowledge of ridesharing, atti- tudes of the organization, and attitudes towards the innovative ridesharing program). Instructions to raters and those completing the instrument - same as with open-ended questionnaire. Scoring - Same as open-ended questionnaire. Analyses needed for the scale - rank order correlation of concepts listed above for inter-rater reliability. Analysis of participant satisfaction will be ration- ally based. Sociometric Choice Scale To determine the cohesiveness of participating carpools, administered at mid-point of program and at its conclusion. Reliability and Validity - Odd-even split half reliability and pretest-post-test validity. Pool of Items - to be obtained from literature review, selecting those rationally seeming most applicable to this situation. Scale - one scale measuring approximately five to ten cri- teria (selected areas including different aspects of possible association: work, play, visiting, etc.); 54 with the choice limited to those persons within the carpool. (Obviously for some choice to be made, this applies only to carpools of three or more persons). Concepts to be measured - (l) the degree to which the indi- viduals are accepted in the group; (2) relationships that exist among individuals; and (3) the structure of the group. Scoring - will be done by listing the other individuals in I the group which the participant would interact with for each particular criteria. Rating - the cumulative scores will be entered into socio- metric tables established for each group. The selections will be summed in order to make two specific measures: (1) the "social receptiveness“ score--the number of people choosing a particular participant; and (2) an "emotional expansion" score--the number of people chosen by a partici- pant. Analyses needed for the scale - correlations between odd- even items; constantcy of choice (actual preference on first test repeated later on). 55 REFERENCES Agyris, Chris; Intervention Theory and Method: A Behavioral Science View; Reading, Massachusetts: Addison- Wesley Puinshing Company, Inc., 1970. Anastasi, Anne; Psychological Testing; New York: MacMillan; 1976. Downtown Research and Development Center; How People Feel About Downtown: An Opinion Research Study; New York; 1974. Fairweather, George; et al; Creating Change in Mental Health Organizations; New York: Pergamon Press; Fairweather, George W. and Tornatzky, L.G.; Experimental Methods for Social Policy Research; New York: Pergamon Press; 1977. Gallagher, Art; "The Role of the Advocate and Directed Change"; Meirhenry, Wesley C. (ed.), Media and Educational Innovation; University of Nebraska ExtensiEn Division and University of Nebraska Press, 1964. Goldman, A.R.; "The Recreational Hour: An Evaluation of Behavior in an Unstructured Situation"; Fairweather, G.W. (ed.), Social Psychology in Treating Mental Illness: An Experimental Approach; New York: Wiley; 1964.‘ Havelock, Ronald G.; Planning for Innovation Through Dis- seminationgnd Utilizatiog of Knowledge; Ann Arbor Michigan: Institute for Social Research, The Uni- versity of Michigan; 1971. Havelock, R. and Benne, K.; "An Exploratory Study of Know- ledge Utilization"; Watson, G. (ed.) Concepts for Sgcial Change; Baltimore, MD: Moran Printing Services, Published by the N.T.L., NEA for COPED, March, 1967. Mann, Floyd C.; "Studying and Creating Change: A means to Understanding Social Organization"; Arensberg, Conrad M. et al. (ed.), Research in Industrial Relations: A Critical Appraisal; New York: Harper and Brothers Publishers; 1957. 56 Maslow, A.H. "A Theory of Human Motivation"; Patten, Thomas H. Jr. (ed.), Classics of Personnel Manage- ment; Oak Park, Illinois; Moore Publishing Company, Inc.; 1979. Miles, Matthew B. (ed.); Innovation in Education; New York: Teachers College, Columbia University; 1964. Pratsch, Lew; "Commuter Ridesharing"; Gray, George E. (ed.), Public Transportation: Planning, Operations and. Management; New York: Prentice; 1979. Rogers, Everett M. and Shoemaker, F. Floyd; Communication of Innovations: A Cross-Cultural Approach; New York: The Free Press; 1971. Shoup, D.C. and Pickrell, D.H.; A Solution to the Parking Pi ue; Los Angeles: University of California; 1982. Thelen, Herbert A.; "Concepts for Collaborative Action- Inquiry"; Watson, Goodwin (ed.) Concepts for Social Change; Washington, D.C.: National Education Association; 1969. U.S. Department of Transportation; Paratransit, Technology Sharing: State-of—the-Art Overview; Washington, D.C.; 1981. -. .9 ~_—. n~ I 31293 02645 9812