Content-Type: text/html

A Repertoire Approach to Environmental Information Channels





A Repertoire Approach to Environmental Information Channels

Garrett J. O'Keefe
Heather Ward
Robin Shepard

Department of Life Sciences Communication
440 Henry Mall
University of Wisconsin-Madison
Madison, Wisconsin  53706
(608) 262-1843
<a href="/cgi-bin/wa?LOGON=A3%3Dind0109B%26L%3DAEJMC%26E%3D8bit%26P%3D2364630%26B%3D--%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D_260024621%253D%253D_%26T%3Dtext%252Fplain%3B%2520charset%3Diso-8859-1" >[log in to unmask]</a>


Submitted to

Science Communication Interest Group
Association for Education in Journalism and Mass Communication
Annual Convention
Washington, D.C.
August 2001












A Repertoire Approach to Environmental Information Channels

A Repertoire Approach to Environmental Information Channels

Abstract
This study supports the hypothesis that given the multiple functions communication channels can serve, individuals use repertoires or groups of overlapping information channels for various purposes.  Landowners in three Wisconsin counties were segmented into urbanites, rural nonfarmers, and farmers.  We analyzed the frequencies with which these groups used different channels for information regarding conservation.  Channel use by the groups differed although the same repertoires were found for each.  Predictors of repertoires varied.


A Repertoire Approach to Environmental Information Channels

A Repertoire Approach to Environmental Information Channels


Little has been done to tie studies of public environmental information channels to the broader sweep of communication behavior theory.  Benefits would follow for literature on individuals' information preferences more generally (e.g., Salwen & Stacks, 1996; McQuail, 2000), on how people  process and use information (e.g., Petty and Cacioppo, 1990; Eagly and Chaiken, 1993) and how it may inform or influence them (e.g., Bryant and Zillmann, 1994, Perse, 2001).  Likewise, linking environmental information to the wider scope of communication could improve policy and strategies for public understanding of environmental issues, and likely science more generally.
We here hold to the view of audiences being active in information seeking and selection processes, rather than of audiences being passively acted upon by communication agents.  Continuing studies on the gratifications communication serve, for example, has supported a view of people choosing media and/or interpersonal channels that meet their expectations for their serving informational interests and needs (Rosengren et al., 1985; Rubin, 1994).
As applied to environmental issues, the underlying hypothesis is that the public uses communication channels, both mediated and interpersonal, as a function of their motivational need for particular information content, the availability of the channels, their expectations of the usefulness of the channels, their comprehension of and attention to the channels, their cost, and their credibility.  Demographic and socio-cultural factors may influence communication patterns as well.  We also hold that individuals, are more apt to use repertoires or groups of overlapping information channels for various purposes (O'Keefe, Boyd & Brown, 1998), rather than simplistically relying primarily on any one channel, as previous measures typically suggest (Sunae & Rodriguez, 1999).
Here we rely on survey research data from three broad publics with often differing environmental concerns: Urban (and near-city suburban) dwellers, rural landowners, and owners of agricultural land.  Using this breadth of groups, we attempt to build a more complete picture of environmental communication channels, including comparisons among the use of public (e.g., Extension and other governmental agencies),  private (e.g., commercial dealer) information channels (Wolf, 1998), as well as general mass media and interpersonal channels.  Findings eventually will be tied to communication models of information seeking, utility, satisfaction, and impact, particularly in the emerging realm of repertoires and complementary vs. convergent use of communication channels.

Background and Conceptual Justification
The development of information-seeking and processing strategies currently receives major emphasis in communication research and theory more generally (Salwen and Stacks, 1996; Bryant and Zillmann, 1995), and in science-related communication in particular (Friedman et al., 1999).  This is in part because of the increased interest in how complex scientific information can best be imparted to the public, whether involving environmental, public health and safety,  biotechnological, economic or related issues (e.g., Bracht, 1999; Friedman et al., 1999; Sexton et al., 1999; van den Ban and Hawkins, 1996; Dillard and Pfau, in press).  It is also a consequence of the technological sophistication of communication systems themselves (e.g., Nellist and Gilbert, 1999).
The uses-gratifications model noted above is having a resurgence for this reason, and here we have chosen to couple it with the repertoire concept (Heeter and Greenberg, 1985; Reagan, 1996).  O'Keefe et al. (1998) were successfully able to do so in examining how the public chose channels of information for health problems.  They found that people reported learning different amounts of preventive health information from different channels, and a mix in levels of learning across channels.  An exploratory factor analysis indicated three clear repertoire groups of particular mediated and interpersonal channels, and hierarchical regression analysis indicated demographic and psycho-socio predictors of use of those repertoires.  Repertoire distinctions were also found among non-farming agricultural opinion leaders, e.g. Extension agents, supply dealers, lenders, etc. (O'Keefe et al., 1997).
These repertoires are often individually tailored, and based on availability, awareness of options, access ease, and awareness of alternatives.  When audiences seek content on a specific issue, some seek a broader band of channels than do others (Heeter and Greenberg, 1985).  Greating interest or involvement has also been found to lead to more diversity of channels (Reagan, 1996).  In sum, we tend to choose a mix of channels to get at the information we think we need and some of these may converge with one another while others may complement one another.
O'Keefe et al. (1998) incorporated a previous concept of complementary vs. convergent information channels (Chaffee, 1986).  Convergent channels provide the same or overlapping messages, hence potential reinforcement or elaboration for the audience; complementary channels provide information in one channel that is not available in another.  Hence a "mixing and matching" of channels or channels provide another rationale for the communication repertories of individuals.  We also distinguished between information sources, or the initiator of the content (e.g., a scientist, change agent, etc.) and the information channel (medium, neighbor, etc.) by which the audience member received the information.  We argued that while there is clear overlap between these two concepts that needs further sorting out, sticking to one or the other -- channels in this case  -- considerably clarified the analytic framework.  The research reported here is a still-early step in developing a hybrid blending
 of these models and concepts into a more coherent framework for examining scientific information seeking and processing, particularly in the environmental realm.
This is especially an issue in public communication programs, notably where environmental issues are concerned, which are fraught with difficulties of conveying multiple, interacting fields of complex -- and often uncertain or contentious -- science, overlapping policy jurisdictions, risky social, political and economic consequences, and so forth (O'Keefe and Shepard, in press).
As for more practical application of the results, agricultural producers in particular are a distinct population subculture with communication patterns and information needs unlike those of others (e.g., O'Keefe and Rursch, 1997).  Previous work has understated and oversimplified this uniqueness.  The situation is further exacerbated in the case of the major issue of agriculture and environmental degradation.  Many consider agriculture and related large-scale land practices (ranching, timber) as the environmental problem of our times.  By comparing agricultural producers with other rural landowners and city dwellers, we hope to more clearly determine how each of these unique publics uses environmental information channels.

Hypothesis and Research Question
The following operational hypothesis is posed in this preliminary study:
H1. The environmental information channel preferences of each of the three publics identified above will form statistically significant clusters of channels, or repertoires, which will vary among the publics, and by demographic and communication characteristics within each public.
This hypothesis is empirically justified by previous evidence found among the general public on health issues, and among non-farming agricultural opinion leaders.  In each case, individual information channels were factor analyzed across respondent samples, and discreet clusters appeared lumping various mass media and personal information channels.  Individual use of these clusters varied significantly by certain demographics, salience of the topic to the respondents, and other factors particular to the samples involved.  The hypothesis is also supported by the uses-gratifications model, and by repertoire and convergent-complementary channel rationales, arguing that individuals choose groupings of information channels on the basis of availability, content needs, individual communication mode preferences, credibility, and the ability of the channels in each repertoire to reinforce and/or add to existing knowledge.
We also ask the research question:
RQ1. What convergent and/or complementary patterns can be teased out of the repertoires to help explain why the clusters are grouped as they are (what overlap or lack thereof in content, information characteristics, audience availability, appeal, etc. can be discerned)?

Methods
The Wisconsin Department of Natural Resources Priority Watershed Program funded in the 1990s, in cooperation with county governments, Extension, and other agencies, dozens information and education programs (among other assistance) in watersheds deemed at higher risk of degradation.  Several of the projects have also funded research efforts assessing citizen knowledge, information needs and channels, attitudes, and behaviors with respect to water quality.  The research was aimed primarily at both needs assessment to help in planning of the information and education programs, and as a pilot effort at greater involvement of watershed residents in the research, planning and conduct of the programs (O'Keefe, 1996; O'Keefe and Shepard, 1998; Shepard and O'Keefe, 1999).  The research included key informant interviews, focus groups, and telephone probability sample surveys.  Data from the most comprehensive of the survey studies carried out will be re-analyzed here focusing on informatio
n channel uses.
The population consisted of adult landowners of three Wisconsin counties; one primarily urban (Marathon), with the two others (Burnett and Polk) primarily rural.  The rural county respondents were further divided between farmers and nonfarmers.  In Burnett and Polk counties, a census survey of all landowners holding five acres or more was attempted, while in Marathon County, a probability-based digit dial sampling procedure insured as much representativeness as possible.  The University of Wisconsin's Wisconsin Survey Research Laboratory did the sampling, interviewing and basic data processing.  The response rate for the Burnett/Polk survey was 76% and the Marathon response rate was 63%.  The data from both of these surveys were combined into one data set.
The residential independent variable, actually a combination of residency and occupation, included three levels: urbanites, rural nonfarmers, and farmers.  (Urbanites were Marathon county residents who did not own farm land, nonfarmers were Burnett and/or Polk County residents who did not own farm land, farmers were Burnett and/or Polk County residents who owned farm land.  Burnett and Polk Counties are adjacent; some farmland and residences crossed the county border.)
The dependent variables were the frequencies of use of information channels for conservation and preventing water pollution.  Thirteen survey questions addressed the use of information channels with the following introduction: "Over the past 12 months, how often have you used any of the following sources of information about preventing water pollution and other conservation-related practices? Would you say not at all (1), rarely, sometimes, frequently, or a great deal (5)?" The information channels included were newspapers, local radio, local television, family and friends, magazines, talking with a commercial dealer, reading commercial dealer materials, talking with a county extension agent, reading county extension materials, talking with a county conservation agent, reading county conservation materials, talking with a Department of Natural Resources (DNR) agent, and reading DNR materials.  The term "sources" was used in the question since previous experience indicated that "ch
annels" caused confusion in some respondents; while, again, conceptually distinct for our purposes, the term "sources" appears equated with "channels" for the vast majority of respondents.

Analysis
A standardized analytic approach was used on the data set described above, following that validated by O'Keefe et al. (1998) for health information practices.  To operationalize the repertoire construct, the individual channel scores were subjected to an exploratory factor analysis using a principal components solution with varimax rotation (SPSS/PC V10), with eigenvalues greater than 1.0 as the standard for defining a factor.  Hierarchical regression models were used to examine the ability of other variables to predict respondent scoring on each of these repertoires.
First, one-way analyses of variance compared the differences among the means for information channel use and farmer, rural, and urban status.  Overall, the most-used channels were newspapers, magazines, and family and friends (Table 1).  All thirteen ANOVAs showed significant differences in channel use across farmer, nonfarmer, and urban groups except for newspapers, friends and family, and DNR materials channels.
Follow-up tests evaluated the pairwise differences among the means for those channels that had unequal variances (radio and magazines) as shown by Levene's Test of Equality of Error Variances.  Post hoc comparisons using the Dunnett's C test showed significant differences between urbanites and the two rural groups for radio use, with urbanites using the radio significantly more for environmental information than did either rural group.  The Dunnett's C test also showed significant differences between nonfarmer groups and farmers for magazine use, with farmers using magazines significantly more than either nonfarmer group.  This is consistent with previous media research on agricultural producers.  While pairwise differences were not found for other channels, the trend was for farmers to be the heaviest consumers of environmental information from both government agency and commercial dealers, followed by rural nonfarmers and lastly by urbanites.  This is not surprising, given that
such channels typically gear most of their communication efforts at farmers.  Urban dwellers, however, tended to turn more to television (and as noted, radio) for conservation information.  The pattern clearly suggests more specialized channels sought out by the more water quality and conservation involved - farmers followed by rural dwellers -- with mass media and family and friends either less sought out or about equally so with respect to the other populations.

Table 1 here


The dimensionality of the 13 information channel use items was explored using factor analysis for each of the independent variable groups.  Three factors were rotated for each group, using the varimax rotation (SPSS/PC V10.0).  The rotated solutions, as shown in Tables 2, 3, and 4, yielded three interpretable factors or repertoires for each group labeled as follows: agency channels, general channels, and commercial channels.  The agency channels included those information sources supported by government agencies, including county Extension agents and materials, county conservation agents and materials, and DNR agents and materials.  The general channels included newspapers, local radio, local television, friends and family, and magazines. The commercial channels included talking with commercial dealers and reading their materials.
While the same factors indicate the same three repertoires across groups, the amount of variance of the 13 variables accounted for by each factor differed across groups.  For example, agency channels accounted for the most variance among the 13 variables for farming, rural, and urban groups (31%, 32%, and 25%, respectively).  The loadings on the factors support the ANOVAs in that urban residents appear to rely less on talking with government agents than other groups do.  The general mass media-interpersonal repertoire accounted for the second most variance among the 13 variables for the farming, rural, and urban groups (18%, 18.5%, and 23.6%, respectively).  The urbanites in this case again show the most substantial difference in channel use.  The commercial channels accounted for the least amount (though still sizable) of variance among the 13 variables for the farming, rural, and urban groups (16.9%, 13.5%, and 14.7%, respectively).

Table 2 here



Table 3 here



Table 4 here


Nine multiple regression analyses were conducted to predict the three factor channel reliance by the farmer, nonfarmer, and urban resident groups (Table 5).  The sets of independent variables were ordered.  The first set included demographic variables; the sets that followed included pollution awareness and concern variables.
The demographic data included income, education, gender, and age.  For the rural nonfarmers, demographics as a set explained a significant amount of general channel use, R2=.107, F(4,131)=3.94, p=.005.  While demographics as a set did not significantly predict channel use for the other groups, the data for urbanites indicate that education level had a significant negative relationship with the agency channel; while income level had a significant positive relationship with commercial channel use.
In addition to demographic variables, current awareness, concern, and perceptions also concerning water pollution were used to predict channel use.  As a set, the awareness and concern variables accounted for a significant amount of urbanite and rural nonfarmer agency channel and general channel use.  The two significant predictor variables within this set were "need for information" and "comparative pollution knowledge." For the urbanites, a sense of a need for information about preventing water pollution significantly predicted general channel use (R2 change=.194, F(5,128)=6.329, p=.000); whereas, for rural nonfarmers, the need for information significantly predicted both agency channel and general channel use.  In other words, when urban residents think they need more information, they are most likely to turn to mass media and interpersonal channels.  Yet, rural nonfarmers, who may have more access to and more confidence in the agency channel than urbanites, are most likely to
turn to both agency channels (R2 change=.216, F(5,126)=7.46, p=.000) and general channels (R2 change=.090, F(5,126)=2.815, p=.019) when they need more information.  On the other hand, for farmers, who may have the same access and confidence in agency channels as rural nonfarmers, the need for information only significantly predicts general channel use.
Just as a need for information predicts channel use, a perception of having more knowledge than other county residents also predicts channel use.  For both urbanites (R2 change=.158, F(5,128)=5.064, p=.000 ) and rural nonfarmers, the sense of having a superior amount of knowledge about water pollution prevention significantly predicted agency channel use.

Table 5 here


Discussion
Environmental information channel uses of the three populations did differ from one another, and clusters of channels, or repertoires, were found for each of the three groups.  While the repertoires were the same for each, the populations varied in their use of them, and within each population different characteristics predicted such uses.  As might be expected, the more general, diverse and less conservation-attuned urban public were somewhat more attuned to mass media, while the rural population - especially farmers - relied more on specific and presumably more detailed channels of water quality and conservation information.  Importantly, agencies and dealers formed two distinct factors, and farmers were about as apt to turn to dealers for conservation-related information as they were to agency personnel and materials.  This adds more evidence of the impact that commercial channels may have on agricultural conservation decisions.  Rural dwellers overall indicating a greater need
 for such information were more likely to turn to general channels than agency ones or dealers.
The repertoires found here were on the one hand more distinct than those for health information among a general population, but demographics and attitudes within the three groups examined did not predict repertoires as well or as consistently (O'Keefe, Boyd, & Brown, 1998).  Health information was posed as a quite general content area in the preceding study, while here we deal with a specific water resource and other conservation content.  Persons at all involved or concerned with the topic may simply be a less varied population, producing less variance in their communication behaviors, save for differences among urban, rural and farm residents, each of which likely has a need for different kinds of information, and places different value on it.  Note that conservation concern within each population was not an effective predictor of repertoires.
In fact, demographics played a quite minor role here, and only self-perceived need for conservation information was a consistent predictor across all three populations.  Information need predicted greater use of more general channels such as mass media and family and friends, which one might expect if level of knowledge is low.  Indeed, the correlation between information need and knowledge was quite low across the groups, and knowledge was a notably significant predictor of  non-farmers getting information from governmentally related agencies, but not necessarily from commercial sources.
We have yet to examine such vital aspects of environmental information channels as their credibility (Sunae and Rodriguez, 1999; Williams, Vallei, Brown & Greenberg, 2000), confidence in them (Byrd, VanDerslice & Peterson, 1977), and cost versus utility (Trumbo, 1998).  However, we suggest that for most purposes, it is preferable to views each channel as part of a "combo" or repertoire, each channel perhaps playing a certain role for the individual.  Some may be cheaper, some may have more credibility, some may yield more information despite greater effort, and so forth.  But we choose combinations of channels based upon such attributes, not simply sticking to the "most credible" despite its cost, or the easiest to obtain despite its lack of utility.  Rather, the process is more "mix and match," likely based upon previous experiences with each medium, or agency, or individual, or specific publication.  As was the case with health information, more examination is needed of the conv
ergence and complementarity of these repertoires, both across the total population and within the types of subgroups studied here.  The immediate reasons underlying need for information, or interest in it, may lead to quite different information search strategies.
Nor have these data touched upon the highly likely refinements in environmental information channels brought about by the Internet and related channels.  Recent evidence suggests that Internet use can affect attitudes toward more conventional channels among farmers, and likely among other individuals (Gloy, Akridge & Whipker, 1999).
Continuing exploration of communication behavior using the repertoire approach appears valuable, and may grow more so as Internet content in particular enters the mainstream.  A true clustering of channels used by individuals in differing circumstances, seeking different gratifications, and with varying channels available to them, whether mass mediated, interpersonal, or electronically personal, would be very useful for policy planners, especially in the complex environmental realm, and in learning about how people make choices about the communication patterns they hope will be most effective.
 References
Bracht, N. (Ed.). (1999). Health promotion at the community level. Thousand Oaks, CA: Sage..
Bryant, J., & Zillmann, D. (Eds.). (1995). Media effects: Advances in theory and research. Hillsdale, NJ: Lawrence Erlbaum Associates.
Chaffee, S.H. (1986). Mass media and interpersonal channels. In G. Gumpert and R. Cathcart (Eds.), Intermedia (3rd Ed.) (pp. 62-80). New York: Oxford.
Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Fort Worth, TX: Harcourt, Brace, Jovanovich.
Dillard, J., & Pfau, M. P. (Eds.). (in press). The handbook of persuasion. Thousand Oaks, CA: Sage Publications Inc.
Friedman, S. M., Dunwoody, S., & Rogers, C. L. (Eds.). (1999). Communicating uncertainty: Media coverage of new and controversial science. Mahwah, NJ: Lawrence Erlbaum Associates.
Heeter, C., & Greenberg, B. (1985). Cable and program choice. In D. Zillmannn and J. Bryant (Eds.), Selective exposure to communication. Hillsdale, NJ: Lawrence Erlbaum Associates.
McQuail, D. (2000). McQuail's mass communication theory (4th Ed.). Thousand Oaks, CA: Sage Publications Inc.
Nellist, J. G., & Gilbert, E. M. (1999). Understanding modern telecommunications and the information superhighway. Norwood, MA: Artech House. 1999.
Nowak, P. J., O'Keefe, G.J., Anderson, S. S., Bennett, C. F., & Trumbo, C. (1997). Communication and adoption evaluation of USDA water quality demonstration projects. Washington, D.C.: U.S. Department of Agriculture.
O'Keefe, G.J. (1996, June). Using formative research strategies to involve landowners in watershed protection planning. Proceedings of Watershed '96: A national conference on watershed management (pp. 91-97). Baltimore.
O'Keefe, G.J., & Shepard, R. L. (in press). Overcoming the challenges of environmental public information and action programs. In J. Dillard and M. P. Pfau (Eds.), The handbook of persuasion. Thousand Oaks, CA: Sage Publications Inc.
O'Keefe, G.J., Boyd, H. H., & Brown, M. (1998). Who learns preventive health care information from where: Cross-channel and repertoire comparisons. Health Communication, 10 25-36.
O'Keefe, G.J., & Rursch, J. A. (1998). The media in rural america. In G. Goreham (Ed.), The encyclopedia of rural america. (pp. 457-461). New York: Garland Publishing.
O'Keefe, G.J., & Shepard, R. L. (1998, May). Assessing public attitudes prior to watershed planning: A Comparison of rural, agricultural and urban populations. Proceedings of American Water Works Conference: Watershed management: Moving from theory to implementation (pp. 331-338). Denver, CO.
O'Keefe, G.J., & Shepard, R.L. (1997, August). Standardized public assessments for watershed programs. Proceedings of American Water Works Association Conference on Water Resources Management: Preparing for the 21st Century. Seattle, WA.
Perse, E.M. (2001). Media effects and society. Mahwah, N.J.: Lawrence Erlbaum Associates.
Petty, R. E., & Cacioppo, J. T. (1990). Involvement and persuasions: Tradition vs. integration. Psychological Bulletin, 107, 367-374.
Reagan, J. (1995). The 'repertoire' of information sources. Journal of Broadcasting and Electronic Media, 40, 112-121.
Rosengren, K. E., Wenner, J. A., & Palmgreen, P. (Eds.). (1985). Media gratifications research: Current perspectives. Beverly Hills, CA: Sage Publications Inc.
Rubin, A. M. (1994). Media uses and effects: A uses-gratifications perspective. In J. Bryant & D. Zillmannn. (Eds.), Media effects: Advances in theory and research (pp. 417-436) Hillsdale, NJ: Lawrence Erlbaum Associates.
Salwen, M. B., & Stacks, D. W. (Eds.). (1996). An integrated approach to communication theory and research. Mahwah, NJ: Lawrence Erlbaum Associates.
Sexton, K, Marcus, A. A., Easter, K. W., & Burkhardt, T. D. (Eds.). (1999.) Better environmental decisions: Strategies for governments, businesses and communities. Washington, D.C. Island Press.
Shepard, R. L., & O'Keefe, G. J. (1999, May). Getting the job done at the ground level: Supporting local decision making. Proceedings of the National Watershed Coalition National Watershed Conference. Austin, TX.
van den Ban, A. W., & Hawkins, H. S. (1996). Agricultural extension (2nd. Ed.). Cambridge, MA: Blackwell.
Wolf, S. A. (Ed.). (1998). Privatization of information and agricultural industrialization. New York: CRC Press.

 Appendix
Survey questions
Income
What was your total household income for the previous year, before taxes? 1 = under 10,000; 2= 10,000 < 20,000, 3= 20,000 < 30,000, 4 = 30,000 < 40,000, 5 = 40,000 < 50,000, 6 = 50,000 < 60,000, 7 = 60,000 < 70,000, 8 = 70,000 < 80,000, 9 = 80,000 or >.
Education
What is the highest grade of school you've completed? 1 = 8th grade or less, 2 = some high school, 3 = high school graduate, 4 = some technical/vocational training, 5 = technical school graduate, 6 = some college, 7 = college graduate, 8 = post grad or professional degree, 00 = other.
Gender
1 = male, 2= female.
Age
In what year were you born? Year was translated to age in 1998, then ages were parsed into categories: 1 = <45, 2 = 46-54, 3 = 55-64, 4 = 65-74, 5 = 75>.
How Poll.
How polluted would you personally say the lakes and steams in your part of [your] county are? Would you say not at all polluted (1), slightly polluted (2), somewhat polluted (3), fairly polluted (4), or very polluted (5)?
Prevent Concern
How concerned would you say you personally are about preventing those lakes and streams from becoming any more polluted than they are now? Would you say not at all concerned (1), slightly concerned (2), somewhat concerned (3), moderately concerned (4), or very concerned (5)?
Need Info.
How much of a need do you think you have for more information about preventing water pollution? Would you say no need (1), a slight need (2), some need (3), moderate need (4), or a great need (5)?
Knowledge
Compared to most of the other people in you part of the county, how much do you think you know about the best ways to prevent water pollution? Would you say you know a lot less than most other people (1), somewhat less (2), almost as much (3), more than most other people (4), or much more than most other people in you part of the county (5)?
Concern Compare
Do you think most other people in your part of the county are about as concerned as you are about water pollution (2), less concerned (1), or more concerned than you are (3)?

 A Repertoire Approach to Environmental Information Channels


A Repertoire Approach to Environmental Information Channels

Table 1
Analyses of Variance of Channel Use for Urbanites, Rural Nonfarmers, and Farmers
Channel
Means
 F
Sig.
Urban
Rural Nonfarmers
Farmers
Newspapers
2.5
2.52
2.55
.070
.932
Local Radio
2.19a
1.81b
1.86b
6.624
.001*
Local Television
2.58
1.86
1.72
29.812
.000**
Family/Friends
2.20
2.32
2.29
.599
.550
Magazines
2.13a
2.16a
2.60b
6.233
.002*
Commercial Dealer Talking
1.56
1.64
2.05
9.433
.000**
Commercial Dealer Materials
1.64
1.82
2.20
10.794
.000**
County Extension Talking
1.26
1.51
1.98
23.257
.000**
County Extension Materials
1.61
1.87
2.22
13.107
.000**
County Conservation Talking
1.35
1.60
2.07
22.703
.000**
County Conservation Materials
1.65
1.86
2.24
12.591
.000**
DNR Talking
1.47
1.73
1.89
7.477
.001*
DNR Materials
1.75
1.86
1.98
1.922
.148
Note. Means in the same row that do not share subscripts differ at p < .05 in the Dunnett's C test of multiple comparisons.
*p < .01, **p < .001
 Table 2
Factor Analysis of Correlations between Agency, General, and Commercial Channels for Farmers
Factors
Channels
Agency
General
Commercial
Newspapers
.124
.625
.336
Local Radio
.001
.646
.354
Local Television
.161
.642
-.102
Family/Friends
.404
.638
.197
Magazines
.185
.516
.554
Commercial Dealer Talking
.397
.199
.775
Commercial Dealer Materials
.297
.125
.859
County Extension Talking
.807
.002
.312
County Extension Materials
.731
.257
.200
County Conservation Talking
.822
.001
.243
County Conservation Materials
.797
.238
.118
DNR Talking
.794
.228
.213
DNR Materials
.657
.498
.005

 Table 3
Factor Analysis of Correlations between Agency, General, & Commercial Channels for Rural NonFarmers
Factors
Channels
Agency
General
Commercial
Newspapers
.250
.580
.121
Local Radio
-.003
.807
.002
Local Television
.002
.801
.001
Family/Friends
.220
.589
.176
Magazines
.400
.514
.113
Commercial Dealer Talking
.200
.105
.890
Commercial Dealer Materials
.153
.153
.894
County Extension Talking
.740
.173
.150
County Extension Materials
.816
.196
.149
County Conservation Talking
.799
.116
.201
County Conservation Materials
.858
.188
.000
DNR Talking
.773
.008
.101
DNR Materials
.845
.006
.125

 Table 4
Factor Analysis of Correlations among Agency, General, and Commercial Channels for Urbanites
Factors
Channels
Agency
General
Commercial
Newspapers
.108
.802
.009
Local Radio
.004
.772
.182
Local Television
.141
.801
.133
Family/Friends
.149
.681
.184
Magazines
.296
.614
.004
Commercial Dealer Talking
.243
.263
.811
Commercial Dealer Materials
.009
.313
.825
County Extension Talking
.493
-.002
.539
County Extension Materials
.733
.174
.149
County Conservation Talking
.700
.002
.357
County Conservation Materials
.754
.233
.132
DNR Talking
.740
.008
.109
DNR Materials
.807
.293
.003

 Table 5
Learning about Pollution Prevention from Three Communication Channels
Farmers
Factor 1
Factor 2
Factor 3
R2
Adj R
.056
Beta
R2
Adj R
.052
Beta
R2
Adj R
-.030
Beta
Demographics
.038
-.011
.024
-.026
.022
-.028
     Income
-.009
-.072
.132
     Education
.013
-.066
-.125
     Gender
-.045
.243
.157
     Age
-.047
.037
.048
Awareness & Concern
.160
.156
.083
     How poll. (7)
.002
-.133
.042
     Prevent concern
.200
.064
-.102
     Need Info.
.209
.327**
.202
     Knowledge
.096
.059
-.168
     Concern compare
-.095
-.089
-.079
R2 change
.122
.132
.061

Rural Nonfarmers
Factor 1
Factor 2
Factor 3
R2
Adj R
.219
Beta
R2
Adj R
.140
Beta
R2
Adj R
.011
Beta
Demographics
.055
.026
.107 (s)
.080
.024
-.006
     Income
.066
-.188*
-.035
     Education
.023
.162
.082
     Gender
-.111
.083
-.068
     Age
-.008
-.183*
.107
Awareness & Concern
.271
.197
.077
     How poll. (7)
-.026
-.079
-.106
     Prevent concern
.011
.089
.045
     Need Info.
.181*
.254**
.112
     Knowledge
.373***
-.093
-.130
     Concern compare
-.117
-.010
-.180
R2 change
.216 (s)
.090 (s)
.053

Urbanites
Factor 1
Factor 2
Factor 3
R2
Adj R
.144
Beta
R2
Adj R
.158
Beta
R2
Adj R
.011
Beta
Demographics
.042
.014
.019
-.011
.035
.005
     Income
.090
-.043
.257*
     Education
-.251*
-.007
-.073
     Gender
-.029
-.081
.123
     Age
.075
.173
-.044
Awareness & Concern
.201
.213
.076
     How poll. (7)
-.065
-1.04
.069
     Prevent concern
-.031
.129
.051
     Need Info.
-.049
.339***
-.015
     Knowledge
.358***
.121
-.014
     Concern compare
-.139
-.085
.202*
R2 change
.158(s)
.194 (s)
.041
***=.001-.0000;**=.002-.01; *=.02-.05