:3 . . . , . . L. . . _ : , . . ; , . 1v... 9.... . y 3. . . _ A . p ,. V ‘ 1 . . . , .. . . . . . . . r . c v . . A , , . . o ._ . . . . . . . . r. . .. ,.‘ . . . ... , C.“ :3 I“ ‘ . t . . .. .. ‘ . . _ _ ‘ . f. . . 2.. o . . .1: . 21v .7. , . ! Date This is to certify that the thesis entitled ENERGY AWARENESS AND MEDIA CREDIBILITY: AN ANALYSIS presented by William W. Brownell has been accepted towards fulfillment of the requirements for M.A. degree in Telecommunication Major profess 631W 3am October 5, 1977 0-7 639 LIB RA R Y Michigan State University ENERGY AWARENESS AND MEDIA CREDIBILITY: AN ANALYSIS BY William W. Brownell A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Telecommunication 1977 Accepted by the faculty of the Department of Telecommunication, College of Communication Arts, Michigan State University, in partial fulfillment of the requirements for the Master of Arts degree. Gil/Ll", 8: éqtgt’fm (7 Robert Yadon Director of Thesis ii ABSTRACT ENERGY AWARENESS AND MEDIA CREDIBILITY: AN ANALYSIS By William W. Brownell Since there has been no research known to have linked energy awareness and media credibility in one study, this research concentrates on whether people perceive an energy problem, how much they know about energy use, how they re- ceive information about the energy problem, what degree of credibility they attach to the various means of communica— ting information, and the best media for reaching these pe0ple. The methodology involved the use of telephone interviews, and the sample consisted of young families drawn from the 1977-78 Grand Rapids directory. Data was treated with Pearson's correlation, chi-square, t—tests, and analyses of variance. In general, the results indicate that people do perceive an energy problem, perceive paying more for energy but use less, are not knowledgeable of energy consumption, do not .necessarily use the sources of information they believe to be the most credible, and contradict themselves in reference to media credibility and truthfulness. ACKNOWLEDGMENT I would like to thank Robert Yadon, my thesis advisor, and Professor John Abel for their eager contributions to this work. I would also like to thank my parents for their faith and support. Any errors are my own, of course. iii TABLE OF CONTENTS CHAPTER I. II. III. IV. INTRODUCTION . The Social Significance Purposes of this Study . Study Method . Sc0pe . . Definition of. Terms Thesis Organization FOOTNOTES BACKGROUND LITERATURE, HYPOTHESES, AND RATIONALE Energy Problem.Awareness . . Responsibility for the Energy Problem Energy Consumption Habits . . Energy Information Sources . . . Information Source Credibility . Research Hypotheses FOOTNOTES METHODOLOGY The Population . Selection of Respondents Sample Characteristics The Instrument The Questionnaire The Pretest . Conduct of the Interviews Treatment of Data Summary . FOOTNOTES RESULTS Energy Problem Awareness . . Responsibility for the Energy Problem Energy Consumption Habits Energy Knowledge iv PAGE H 00 \l(3‘£.J'IL;.>(J~>I\>l---'| TABLE OF CONTENTS (Cont'd.) CHAPTER Energy Information Sources Sex . Income Education and Income, Household Size and. Income, and Income and Sex . Race . . Information Sources Credibility Income . . Education . Race Sex . . Household Size Education and Income, Household Size and. Income, and Income and Sex . Believability and Truthfulness Compared : V. SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS Summary . Conclusions . . The Study as Designed Method of Study A Changing Society . Recommendations . APPENDICES A. Questionnaire B. Pearson Correlation Matrix . LIST OF REFERENCES PAGE 54 58 61 63 64 65 66 67 68 69 71 71 77 79 79 8O 83 91 92 TABLE C‘U'l-PUON \l 10. LIST OF TABLES Comparative Education Figures Comparative Income Figures ANOVA of Energy Problem Responsibility . ANOVA of Energy Consumption Habits T-test of Energy Information Sources Mean Scores of Income Energy Information Credibility . ANOVA of Energy Information Source Credibility . Mean Scores of Education Energy Information Credibility . Mean Scores of Race Energy Information Credibility . . . . . Mean Scores of Sex Energy Information Credibility . vi PAGE 30 30 47 50 62 65 65 66 67 68 CHAPTER I INTRODUCTION This study was conducted to determine the present level of energy awareness and what are perceived to be the most credible media. This chapter discusses the social signifi- cance of the subject under consideration, the purposes of this study, the scope of the study, and the organization. The Social Significance For the past few years, this country has been faced with an energy problem. The United States has approximately six percent of the world's population and land area but consumes nearly 40 percent of the energy the world uses. Additionally, the country's population has increased more than 33 percent since 1950, but energy consumption has increased 100 percent.1 In that time, the United States has become an importer rather than an exporter of energy, importing 15 percent of its total energy and 35 percent of all petroleum in 1975. While the use of gas and oil increased markedly, coal consumption decreased 36 percent between 1947 and 1962. Rising costs of labor and more strict mine safety requirements and air pollution standards have further retarded the use of coal. The progress of nuclear power has also been plagued with technical diffi- culties, economic efficiencies, public antipathy, and a growing concern for safety. When the United States backed Israel in the 1973 Middle East War, the Persian states shut off U. 8. oil imports. The emergency ended in March 1974, and domestic inventories were restored by that summer, but the crisis left serious doubt in the minds of many officials whéther similar incidents in the future could seriously criple everyday activities and industry.2 Purposes of this Study There have been numerous studies, articles, broadcasts, reports, books, and speeches about energy for the past three decades. Since it has only been considered a major problem since the Arab Oil Embargo of 1973, most of these studies appear to lack depth and a breadth of understanding of the energy problem and interrelated factors. They are surveys conducted by various organizations, such as the oil industry, to prove a point or substantiate a claim. A Michigan State University study attempted to connect media use with energy knowledge, but it did not attempt to ascertain what credibil- ity the public attaches to the various information sources.3 The purposes of this study were, then, to determine: 1) if the people perceive an energy problem; 2) the target audience's knowledge of energy use; 3) how the people receive information about the energy problem; 4) what degree of credibility people attach to various communication media (inter-personal, print, broadcasting, etc.); and 5) the best media for reaching these peOple. The results will hopefully help advertising experts to formulate a strategy and a media package. Study Method In brief, questionnaires were used to determine what the public knows about the energy problem, where it gets its information and how it feels about the credibility of vari- ous media. Telephone interviews were conducted by trained interviewers from a systematically—drawn sample out of the Grand Rapids Telephone Directory. A pretest of the question- naire was conducted in the Lansing, Michigan, area to deter- mine if adjustments were necessary. Analysis consisted of correlations, chi-square, difference—of—means tests, and analysis of variance. Scope Specific limitations were faced in the execution of this study. Since there was an economic funding problem in the conduct of this study, it was necessary to use assests avail— lable through Michigan State University. Special direct (WATS) telephone lines were used at the Department of Telecommunication. Thus, only three areas in Michigan outside of the Lansing area (which is constantly the subject of investigation) were usable in this study. Detroit is too metropolitan as is the Ann Arbor area. The latter is also under a constant barrage of studies originating at the University of Michigan. Grand Rapids is the only area in which there is both a metro- politan and agricultural cross-section. Mbst ethnic groups are represented as well. However, it must be remembered that this is merely one city not necessarily typical of the rest of the state. This application and inferences should not be used to predict the situation in other Michigan communities. This does not mean, however, that the design of this research cannot be applied effectively to other communities. Second, half of the findings of other studies are based on secondary analysis of other researchers and journal arti- cles which have been abridged, thus eliminating the raw data for personal, first-hand analysis. It is believed, however, that the previous studies tend to support one another over time. Third, age demographics were asked initially since only 18- to 40-year-old respondents were desired for reasons ex— plained later. This could have caused two negative results: 1) respondents may have refused to divulge their age for per- sonal reasons, and 2) they may have lied about age in order 'to avoid the interview, thus voiding that particular case as a refusal. Fourth, the use of telephone interviews did, of course, limit follow-up, in-depth questioning to ascertain the reason for discrepancies or contradictions in their answers. It is hoped that this study revealed what the population knows about energy and will give us enough information about how it perceives the different modes of communicating infor- mation, whether it be through friends, through print, and/or through broadcasting. It is also hoped that the results of this study will contribute to the overall energy effort. Definition of Terms Energy: that energy which is released from coal, nat- ural gas, or nuclear power. It is used to provide fuels and electricity.4 Energy Awareness: individual recognition of the energy situation, related problems, and knowledge of energy use. Energy Crisis: the culmination of a series of adverse events which occurred in the early—19703 climaxing in October 1973 with the complete Arab Oil Embargo. (NOTE: The crisis is not to be confused with the energy problem.) Energy Knowledge: the fact or condition of knowing (perceived directly) about energy with familiarity gained through experience or association; the range of a respondent's information or understanding of energy as demonstrated by 'successful answering of empirically verifiable energy use questions. Energy Problem: the unsettled, continuing energy short- age; this is not to be confused with the "energy crisis" which was a specific incident. (NOTE: Not one source in- vestigated has defined what is or is not an energy problem. Researchers and reporters dealing with energy use the terms " loosely and interchange- ”energy problem” and "energy crisis ably. All of the definitions above are based on the implied definition in the majority of cases and Webster's Seventh New Collegiate Dictionary.) Reliability: the quality of being worthy of belief. Truthfulness: the state of being the case; telling or disposed to tell the truth. Young Families (the population): 18- to 40-year-old single persons or couples who may or may not be married, may or may not have children, and make more than $8000 (house- hold) a year.5 Thesis Organization Aside from the social significance, purposes, method— ology, and scope already discussed, a history of previous studies and hypotheses will be outlined in Chapter II. Chapter III will describe the methodology in detail, Chapter IV the results, and Chapter V the conclusions and recommen- dations. FOOTNOTES 1William Brownell, "Ecology Kick," Fort Benning (Ga.) Bayonet, Sept. 8, 1972, p. 8. 2U. S. Council on Environmental Quality, Sixth Annual Report on the Council on Environmental Quality (1976), pp. 109-161. 3Institute for Family and Child Study, FamilyfiEnergy Pro'ect, no. 3152 (East Lansing, Mich.: Michigan State Un1versity, May 1976). 41bid. 5Defined by the Extension Staff Services, Michigan State University, staff meeting on Dec. 9, 1976. CHAPTER II BACKGROUND LITERATURE, HYPOTHESES, AND RATIONALE This study involved two major areas, media credibility and energy awareness. Since there is an almost endless amount of literature on the subject of energy and what to do about it, the information from these sources has been grouped so as to give some logical pattern for discussion: 1) awareness of an energy problem; 2) responsibility for the energy pro- blem; 3) energy consumption habits; 4) where people receive energy information; and 5) information source credibility. Energy Problem.Awareness Before a problem can be effectively solved, it must be recognized. Research indicates that energy usually takes a backseat to other issues although it is now beginning to achieve majority status. Attitudes about energy first arose in relation to environmental concerns in the early 19703. Prior to that, there was little interest.1 In a 1972 Gallup poll, 83 percent surveyed perceived an energy crisis; how- ever, its importance in relation to other issues was almost linsignificant.2 After one month, energy began appearing in C1 pr Gallup and CBS public opinion polls although the issue was at the bottom of the list (four percent of the people perceived it as a problem). In fact, it was not until the height of the energy crisis (1973-74) that energy was deemed a major problem by more than 10 percent of the Gallup samples.3 A study conducted early in 1973 reflected the need for an actual crisis in order to get people to believe that there is an energy problem.4 A study of both the news media and oil industry between 1971 and 1973 revealed that the oil industry had issued warn- ings before 1971. The news media carried little editorial content on energy problems, and those beginning in 1972 ques- tioned the believability of energy shortages.5 A nationwide survey taken in July 1975 revealed that providing enough elec- trical power for the future is the third most important problem communities face; however, opinions had changed from 24 per- cent in 1971 to 39 percent in 1975.6 A Harris survey conducted during the same time revealed that the majority of Americans perceive an energy problem, but leaders perceive the problem to be more acute than does the general public. This suggests that higher levels of knowledge about energy result in stronger attitudes.7 Another 1975 study resulted in respondents rank- ing energy shortages (26%) behind the economy and unemploy- ment,8 a drastic improvement. Although these studies indicate 9a trend toward crisis recognition, a 1976 study shows a de- cline. Roper reflected a decrease in energy as a national problem from 27 to 22 percent, ranking seventh among 11 10 problems; 52 percent believe the oil shortage was and continues to be a contrivance.9 Demographically, there are several points that are per- tinent to the issue of the energy problem. In a study con- ducted in 1974, females were more likely than males in almost all circumstances to believe that the 1974 situation was real and not a maneuver.10 The 1976 Roper study reported that when it comes to favoring either environmental or energy needs, sex makes little difference.11 The relationship be- tween income level and belief in the crisis also varied by sex; i.e., low-income males were not convinced of the reality while middle-income females were more skeptical than low- and high-income respondents.12 Roper also found that all income levels view the crisis with equal seriousness for their com- munity while those with high incomes were more personally bothered by the crisis and viewed it as more real.13 Another 1976 study of family values, household practices, and contex- tural values resulted in the conclusion that no difference among intrafamilial patterns of self-esteem, familism, and social responsiveness were found in the adoption of energy conservation practices; however, family economic consciousness was found to be a predictor in adopting energy conservation practices. This eco-consciousness appears to be related to higher levels of education and family income.14 Education- iwise, the more education a respondent has, the more likely the response that the crisis was real. A strong correlation also exists between education and the level of energy C< ll awareness.15 The Roper study found that college educated persons tend to favor environmental needs first. These were stronger among younger persons and liberals.16 Another pri- vate study by Farace in 1976 found that younger, more educated, higher income, urban cosmopolitan individuals provide most of the concerned population.l7 Responsibility for the Energy Problem Blame for the energy crisis, according to Roper, is generally given to the oil companies (57%) although there has been some shift toward other segments: Arab countries (37%), electric power companies (29%), the Administration (28%), Congress (26%), the consumer (18%), Israel (13%), and environmentalists (9%).18 In contrast, Warren concluded that the energy crisis was perceived as a failure of the American society and institutions rather than foreign country action. The crisis was experienced in terms of middle class phenomena.19 Farace tends to agree with the former analysis, saying utility and oil companies, governmental agencies, and mid-Eastern countries all share the blame equally for the current energy problem. Relatively few persons perceive that their own behavior is a part of the problem.20 Energy Consumption Habits It was discovered in 1973 that an information campaign designed to manipulate behavior did not influence electricity consumption. The energy crisis, likewise, did not influence 12 consumption; however, it did have a major impact on attitudes toward the desirability for conservation. Apparently, the lack of immediately perceived rewards (economical or social) is responsible for this discrepancy between behavior and atti- tudes.21 In contrast, General Publics Utilities of Pennsyl- vania and New Jersey have achieved some minor load shifting through time-of-day incentive rates to large users and "it pays to wait until eight" advertising campaigns. Some of the direct controlled load management systems have been operating for several years, and the impact of their operation can be readily observed; however, results of these programs using rate incentives will require several years of operating experience.22 A similar study was conducted in 1975 to investigate procedures to encourage reduction of energy use by residential consumers. One group that received an economic incentive in addition to a conservation manual averaged a lS-percent reduction for the first two weeks; however, the reduction was linked directly to weather conditions, not to the independent variable. It was also found that feedback to consumers may be a significant variable in energy conservation programs.23 A 1973-74 study reflected that households with a lower social status perceived the shortage in terms of its personal impacts and as a "real problem"; whereas, higher status house- holds became less supportive of severe restrictions on energy iuse as the situation progressed. This suggests that the experienced shortage did not match warnings or that people became disillusioned with economy of energy.24 13 The more persons in a household, the more frequently adjustments have been made. (The fact that single-member households seldom make adjustments is probably due to a high proportion of the elderly and younger households having a smaller inventory of appliances.)25 Almost 80 percent of owner-occupied households have made adjustments while only two-thirds of the renters have made adjustments. Renters making indirect payments for energy are not likely to be as aware of how much energy they are using or the cost.26 Income-wise, a 1974 study found that lower-class income groups had the least flexibility and the upper-income groups the most.27 Two years later, Morrison and Gladhart noted that income is the best predictor of energy consumption.28 These findings are supported by Smith who found that higher income households are more likely to have made adjustments and will continue to do so; however, they normally have more adjustment alternatives (more appliances and larger homes).29 Hogan found that individuals with higher incomes and higher educa- tional levels apparently are more aware of energy problems, but conservation practices are usually limited to in-home conservation, not transportation conservation.3O Smith supported Hogan, reporting that as the level of education increases, the more energy-reducing adjustments are 'made or might be made.31 Warren discovered results which con- flict with Smith's. In dealing with actual amounts of savings, he found that persons with the highest education levels reg St‘. 1e of GI 14 reported the greatest amount of energy consumption.32 A strong correlation also exists between education and the level of energy awareness. The greater the energy awareness of the respondent, the more likely he accepts specific energy policies, such as tax deductions and rationing. It is uncertain whether educational programs will transform public values and goals in the area of energy consumption, but increased environmental and energy efforts in the schools, continuing education programs, and information messages in the mass media could lead to a greater awareness of the issues and alternative responses.33 As the infusion of mass media information increases into a social system, those persons with higher socio-economic status tend to acquire this information at a faster rate than lower-status segments so that the "gap" in knowledge between these segments tends to increase rather than decrease.34 This knowledge gap widens with increasing levels of media input because of several factors. Formal education results in higher reading and comprehension abilities necessary to acquire knowledge. Those who are better informed are more prepared to understand new information in the mass media.35 Personal interest is also an important factor in learning. Readibility correlates highly with enjoyment of the article. Content does not affect enjoyment.36 The University of 'Michigan suggests that interest in science depends primarily on the direct impact the information has on a person's per- sonal life.37 Publics who are most likely to respond to mass Q) 11 be 31 tl CI re 1c 15 media information have a prior interest in the subject; how- ever, the media are relatively powerless in effecting changes in attitudes of consequence because of much resistence in being moved away from their comfortable indifference to many public issues.38 In dealing with age as a determining factor, Smith dis- covered that most adjustments are made by the lower-middle age group and least by the elderly,39 probably due to the latter's more inflexible, fixed income. Women generally have higher scores on the energy aware- ness scale and are more likely to believe in the reality of the energy crisis than men.40 Energy Information Sources A secondary analysis of national sample surveys between 1952 and 1964 revealed that newspapers account for 41 percent of primary science information, followed by magazines (28%), television (26%), and radio (2%). Respondents with at least some college and higher incomes use magazines more than lower status persons. Television is the primary source for those with a high school or less education.41 This finding was supported in 1970 by Tichenor, Donohue, and Olien who noted that most science and public affairs, with the exception of crisis events, are carried in print. Both are more heavily used by higher—status persons; science news lacks the constant repetition which facilitates learning and familiarity among low-status persons.42 A spring 1975 study revealed that the 16 science consumer of information is more likely to be male in areas of non-medical science and female in areas of medical news. The males tend to consume newspapers, magazines, radio, and television and likely to be under 50 years of age, were exposed to science in school, are probably suburbanites or middle-sized city-dwellers, have above-average income and education, and see science as beneficial personally and to the world in general. WOmen do not show a consistent set of characteristics.43 A 1974 study found that mass media is given credit as the major sources of information about energy. TV commenta- tors ranked first (83%), newspapers and TV specials second (79%), followed by magazines (67%), radio (65%), friends (non-neighbors) (45%), relatives (44%), friends (neighbors) (41%), co-workers (39%), unions (36%), utility companies (36%), and organizations to which the respondent belongs (30%). Un- employed persons turn to informal neighborhood networks of social support.44 A year later, it was noted that 56 percent of the respondents from the Chicago area make use of tips in newspapers and bill enclosures (NOTE: this survey was neither significant nor reliable, but its findings are similar to other studies).45 More current studies reflect that televi- sion has caught up to newspapers and are ranked about equal.46 A 1976 Michigan State University study was conducted to lascertain the ideal entertainment source. It was found that persons who do not subscribe to cablevision in the Lansing, Michigan, area do not perceive subscription cablevision as the id ca ne St ir. 16 tl it ir hi l m: tl re 61' re 0f 43 l7 ideal entertainment source; subscribers do relate paying for cable with the ideal source. Non-subscribers use the local newspaper as an information source for restaurants and movies. Subscribers do not link these together. As a source in learn- ing about cable, 42 percent of the subscribers learned about cable through a friend, 21 percent through mail flyers, and 16 percent through newspapers. Of interest, 75 percent of the non-subscribers were contacted by mail. These findings indicate the word-of-mouth (interpersonal) is a more powerful information source. Cable origination channels also enjoy high use: 72 percent of the subscribers favor local program- ming on cable; 60 percent watch the all-night movie channel (not to be confused with pay TV movie channels) either daily, weekly, or monthly; and, on a daily basis, 30 percent watch the weather, sports, and world news channels while 27 percent watch the Michigan news channel and eight percent the business news channel.47 Another study conducted in 1971 by Baldwin and Gluck on the adaptive behaviors of the new cable television subscriber resulted in 14 percent of the respondents indicating that they had changed, either reducing newspaper reading time or stopping altogether. Nine percent believed that cable television had changed (reduced or stopped) magazine reading habits. Sixty- eight percent read magazines, an average of three for each Vrespondent. Since subscribing to cable television, 46 percent of the respondents admitted more television viewing time while 43 percent less. In addition to the three networks, almost 18 two-thirds used the independent channel. Only 12 percent used the public channels regularly. Nearly one-quarter of the respondents had viewed one or more of the local origina- tion channels. Of interest, there seemed to be either an economic trade-off (substitution for the daily newspaper cost) or a displacement (taking time formerly devoted to magazines). It was also noted that over half of the respondents appreci- ated the greater "variety" (wider choice of the same old thing, not new types of programming).48 Information Source Credibility Mass media and activists organizations are perceived to be the most credible sources of energy information.49 This finding is supported by a 1976 Michigan State University study which noted that 43 percent of the respondents regard television to be more believable than newspapers in present- ing local news.50 Drawing less confidence is the oil and gas industry, decreasing from 25 to 13 percent (a decrease of 48 percent).51 There is also a rapid decline in the credibility of business and industry as a source of information.52 Part of the problem lies with the belief that the energy industries often place their own economic interests before that of the country's.53 Although universities may have greater credi- bility than the energy industries, potential problems in energy information credibility exist, regardless of the source.54 In public service circles, confidence in Congress has tl pe pe pat anc fOU ser1 19 decreased from 37 percent in 1975 to 20 percent in 1976. (Based on a poll taken of Maryland voters and high school students in 1976.)55 While a great majority of individuals appear to recognize the need or desirability for energy con- servation, a significant number appear to be skeptical about the credibility of government information on energy, at least that which is related to shortages.56 On the other extreme, environmentalists are believed to overstate the hazards of various energy technologies such that doubts are raised about the credibility of the group's efforts.57 Research Hypotheses H1: The public does not perceive there to be an energy problem (shortage of energy). H2: Females are more likely than males to believe that there is an energy problem. H3: Persons with higher incomes are more likely to perceive an energy problem than lower-income persons. H4: The higher the level of education, the more people perceive an energy problem. H5: Blame for the energy crisis lies with the oil com- panies followed by the Arab oil countries, utility companies, and government. H6: People perceive using more energy now than they did lfour years ago (before the energy crisis). H7: The larger the household, the more consumption con- servation adjustments have been made. 20 H8: High-income households are more likely to have made adjustments than low-income households. H9: As the level of education increases, the more energy-reducing adjustments are made. H10: Females know more about energy use than do males. H11: Persons with higher education know more about energy use than those with less education. H12: High-income households know more about energy conservation than low-income households. H13: Television is the main source of energy infor- mation. H14: Males' main source of information is newspapers. H15: Females' main source of information is television. H16: Lower-income households tend to use television as their main source of energy information more than upper- income households. H17: Persons with low education levels tend to use television as their main source for energy information more than upper education levels do. H18: Television is more credible than other information sources. H19: Information from the oil industry is less credible than information from consumer groups. H20: Persons with low income tend to perceive televi- sion as the most credible information source more than high- income respondents do. H21: Persons with high education levels tend to perceive 21 information from consumer groups as the most credible infor- mation source. HO: There is no difference in the way demographic var- iables perceive the energy problem, know about energy con- sumption, receive their information about energy, or perceive the credibility of information sources. FOOTNOTES 1Frank E. Armbruster, The Forgotten Americans: A Survey of Values, Beliefs, and Concerns of the Majority (New Rochelle, N. Y.: Arlington House, 1972). 2GallupPoll No. 99, ed. Jim Schriner (East Lansing, Mich.: Michigan State University, 1976). 3Jim.Schriner, "Communicating Energy Information" (A research report submitted to Department of Communication, Michigan State University, 1976). 4Thomas A. Heberlein, Conservation Information, the EnergngrisiSLgand Electricity COnsumption in an Apartment Complex (Madison, WiSc.: Department of Rural Sociology, University of Wisconsin, 1974). 5Linda A. Dangerfield, Hunter P. McCartney, and Ann T. Starcher, "How Did Mass Communication, as Sentry, Perform in the Gasoline 'Crunch'?" Journalism Quarterly, vol. 52, no. 2 (Summer 1975), pp. 316-25. 6Response Analysis, "The Electric Utility Industry Today," a privately printed report to the Public Information Research Program, Edison Electric Institute, Princeton, N. J., Dec. 1975. 7Louis Harris, "Poll Reflects Energy Concerns," Science Digest 78 (Nov. l975):16-l7. 8Daniel Yankelovich and Associates, "Consumer Attitudes Toward the Natural Gas Industry (Washington, D. C.: an un- published report to the American Gas Association, 1975). 9The Roper Organization, Inc., research conducted for a private oil company (1976). 10James L. Zuiches, "Acceptability of Energy Policies to Mid-Michigan Families," research report no. 298 from the Agricultural Experiment Station Project 3152, Michigan State University, March 1976. 11Roper. 22 23 12Zuiches. 13Roper. 14Mary J. Hogan, ”Energy Conservation: Family Values, Household Practices, and Contextural Values" (Ph.D. disser- tation, Michigan State University, 1976). 15Zuiches. 16Roper. 17Richard V. Farace, "Upgrading Energy Information: Preliminary Guidelines for Michigan," paper presented to the 1976 Summer Workshop on Energy Extension Services, Lawrence Laboratory, University of California, Berkeley, July 18-25, 1976. 18Roper. 19Donald I. Warren, "Individual and Community Effects on Response to the Energy Crisis of Winter 1974: an Analysis of Survey Findings from Eight Detroit Area Communities" (Ann Arbor, Mich.: Program in Community Effectiveness, Institute 0; zabor and Industrial Relations, University of MiChigan, l 7 . 20Farace. 21Heberlein. 22Edison Electric Institute, "Load Management: Its Impact on System Planning and Operation, Phase I," a report of the EEI System Planning Committee, Edison Electric Institute, New York, April 1976, p. IV-4. 23Richard A. Winett and Michael T. Nietzel, "Behavioral Ecology: Contingency Management of Consumer Use," University of Kentucky, 1975 (a paper to be published in American Journal of Community Psychology). 24Mary D. Stearns, The Social Impacts of the Energy Shortage: Behavioral andettitude Shifts (Springfield, Va.: NaEiOnaI’TeEhfiical Information ServiCe, 1975). 25Richard B. Smith, "Household Energy Adjustments," a speech made to the National Agricultural Outlook Conference, .Washington, D. C., Nov. 17, 1976. 26Ibid. 27Warren. 24 28Bonnie M. Morrison and Peter M. Gladhart, "Energy and Families: the Crisis and the Response," Journal of Home Economics (Jan. 1976), p. 16. 29$mith. 30Hogan. 31Smith. 32Warren. 33Zuiches. 34F. G. Tichenor, G.A. Donohue, and C. C. Olien, "Mass Media Flow and Differential Growth in Knowledge," Public Opinion Quarterly, vol. 34, no. 2 (Summer 1970), pp. 159-170. 35Ibid. 36G. Ray Funkhouser and Nathan Maccoby, "Communicating Specialized Science Information to a Lay Audience," The Jour- nal of Communications, vol. 21, no. 2 (Spring 1971), pp. 58- 71. 37Stuart Chase, American Credos (New York: Harper and Brothers, Publishers, 1962), pp. 138-151. 38Harold Mendelsohn, "Some Reasons Why Information Cam- paigns Can Succeed," Public QpiniongQuarterly, vol. 32, no. 1 (Spring 1973), pp. 50-61. 39 Smith. 4OZuiches. 41Serena Wade and Wilbur Schramm, "The Mass Media as Sources of Public Affairs, Science, and Health Knowledge," Public Opinion Quarterly, vol. 33, no. 2 (Summer 1969), pp. 197-209. 42Tichenor, Donohue, and Olien. 43National Association of Science writers, "The Public Impact of Science in the Mass Media," a special study con- ducted by the Institute for Social Research, Ann Arbor, 1958. 44Warren. 45Foote, Cone, and Belding, ”Energy Conservation Prac- tices by Chicago Consumers," an unpublished report to the People's Gas Company, Feb. 3, 1976. 25 46Farace. 47Martin Block, "Continental Cablevision,’ a special class study conducted at Michigan State University, 1976. 48Thomas F. Baldwin and Jon K. Gluck, "Adaptive Behav- iors of the New Cable Television Subscriber,” a special study conducted at Michigan State University, 1971. 49Yankelovich. 50John D. Abel and Michael O. Wirth, "Newspapers or Television: Which is Mbre Credible as a Source of Local News?" a special study conducted at Michigan State Univer— sity, 1976. 51Response Analysis. 52Yankelovich. 53Farace. 54Response Analysis. 55Senator J. Glenn Beall, Jr., "State-wide Poll on Public Issues," a special study conducted by the Republican party in Maryland, 1975-76. 56Federal Energy Commission, The Arab Embargo: Two Years Later (Washington, D. C.: U. S. Government Printing Office, 1975). 57Farace. CHAPTER III METHODOLOGY In general, the method010gy involved the use of ques- tionnaires to determine what the public knows about energy, where it gets its information, and how it perceives the credi- bility of various communications media. Telephone interviews were conducted between 6:30 and 9:45 p.m., March 7-30, 1977, by trained interviewers from a systematically drawn sample from the Grand Rapids Telephone Directory. The population, as defined in Chapter I, is young families. The independent variables consist of demographics: education levels, house— hold size, income levels, ethnic origin (race), and sex. The dependent variables consist of the scores/answers assigned/given by the respondents to each question. The Population The population, as defined in Chapter I, includes all young families. According to the Impact Committee, Extension Staff Services, Michigan State University, the reasons for this particular definition are: young families are starting to establish habits, are more likely to change their attitudes 26 27 and have the ability to make such adjustments but do not. Those households that make less than $8000 a year are forced to make such adjustments to conserve energy because of their poverty-level income (established by the federal government). Selection of Respondents Telephone questionnaires were used to ascertain responses. Once again, the sample of respondents was drawn from the 1977- 78 Grand Rapids Telephone Directory using a systematic design. Only private, non—commercial, non-government telephone list- ings were eligible. Since only 18- to 40-year-old persons are of interest, a funnel-type question was asked initially. Based on Grand Rapids age demographics, a particularly large sample (852) was drawn from a population of approximately 149,950 residential telephone numbers. It was expected that approximately 25 percent (213) of the sample would be non-responses (actually, their were fewer non-responses (183 total (21.5%) including 68 refusals, 59 disconnects, and 59 not-at-homes or no answer)) while approximately 50 percent of those remaining (319) would not qualify because of age (303 did not qualify in reality); therefore, it was expected that 320 responses would complete the questionnaire (366 were actually completed). The procedure used to select the respondents was a system- iatic sample. The number of blank pages (16 government and business) were subtracted from the total number of pages (584) from the population. At random, six columns were selected in 28 the directory from which the number of residential, non- commercial, non-government telephone listings were counted, added, then divided by six to ascertain the average number of listings per column. That figure was then multiplied by four (the number of columns per page) to compute the popula- tion (149,950). Since the sample to be drawn represents 0.21 percent of the pOpulation being studied, no finite correction factor was necessary. Since the frame is approximately 264 usable listings per page, it was determined that respondents would be selected by picking one on the first page, two on the second, one on the third, two on the fourth, etc., etc. Using a table of random numbers, the telephone listing positions were plotted on a clear acetate overlay; thus, the same random order (129 on the first page and 237 and 192 on the second page) would apply throughout. The procedure for dealing with blanks was to alternate going to the first private number below the systematic posi- tion then to the first one above the position. In order for a page to be considered usable, at least one-half of the page had to contain residential, non- commercial, non-government numbers. The random positions marked on the overlay were based on the number of printed lines on a full page. If a respondent gave more than one answer for any partic- ular question, a coin was flipped to determine which answer would be used. 29 Since special direct (WATS) lines were used at Michigan State University for the telephone interviews, a check was made to see if all numbers in the telephone book could be called; all exchanges did coincide. Since inferences are to be made about the entire city, it was important to reach a proportion- ate cross-section of all ethnic groups in the city. It was also desired to obtain data with both a cosmopolitan and agri- cultural flavor. Detroit is too cosmopolitan while Ann Arbor is already heavily used for surveying by a major university located there. Lansing, too, has the problem of being heavily surveyed. Sample Characteristics More than half of the usable cases (n=366) was female (52.6%). This is consistent with Grand Rapids demographics (53.1% female and 46.9% male).1 Race-wise, Grand Rapids characteristics fell short of the actual figures for that area; in that, only 5.7 percent of those surveyed were non— white as compared to 11.3 percent in listed Census figures.2 In reference to education, Department of Commerce figures for the Grand Rapids area differed from the sample (Table 1). The average size household according to the 1976 Survey of Buying Power is 3.13 persons; whereas, the average household size of the sample was 3.62 persons.3 The household income Afor those other than students, retirees, the unemployed, or the respondents who refused to divulge their income is at Table 2. 30 Table 1 COMPARATIVE EDUCATION FIGURES (%) (%) Sample Commerce4 Education Level Figures Figures- Less than high school 6.3 46.0 High school diploma 37.3 30.3 Some college 34.1 12.1 College degree 11.7—4 10.4 Graduate work 10.6——I (NOTE: the discrepancy at the less-than-high school diploma level may be attributed to Commerce figures reporting all current school children; whereas, only persons between 18 and 40 years of age were used in the sample. Additionally, Commerce figures are based on the 1970 Census which are now seven years old.) Table 2 COMPARATIVE INCOME FIGURES (%) (%) 5 Sample Buying Power Income Level Figures Figures Less than $8000 11.2 24.8 $8000-$15,000 37.7 33.4 $15,000-$25,000 30.9 32.2 Above $25,000 15.3 9.6 (NOTE: since only persons 18 to 40 years of age were used in the sample, this may have caused a discrepancy in the differ- ences of income for each bracket. There may also have been 31 those persons in the sample who lied about their income due to embarrassment (low figures).) Under the most conservative conditions, the sample size of 366 could be expected to contain approximately a 2.6 per— cent error with a confidence level of 95 percent. The Instrument The instrument used for this survey has been previously demonstrated to have validity. Systematic sampling differs from simple random sampling; in that, it does not give all possible samples of size "n" from the pOpulation of size ”N" an equal chance of selection. The selection of one sample member is dependent on the selection of a previous one. Sys- tematic sampling produces a more even spread of the sample over the population list than does simple random sampling. This usually leads to greater precision, except when the list is randomly ordered (not the case for this survey).6 Systematic sampling is often used in social surveys be- cause of its simplicity, especially when there is an ex- tremely long list (e.g., a telephone survey). It also insures that representatives of certain ethnic groups (e.g., O'Brien) are selected.7 32 The Questionnaire This questionnaire included an identification and status section followed by a brief introduction explaining the pur- pose of the survey and identifying who was conducting the interview. Telephone numbers of the persons in charge of the survey were also included in the event administrative pro- blems arose. All instructions were upper-cased so that the interviewer would not confuse those portions with that which had to be relayed to the respondent. The questionnaire is at Appendix A. Based on the findings of previous studies and the ob- jectives of this study, the following justification explains why certain questions were asked. After the funnel-type question had been asked on age to eliminate the non-population, the first question ascertained the respondent's perception of whether there is an energy problem or not. It was expected that much of the respon- dent's behavior and responses would probably be reflected by his perception of a problem. If the respondent does perceive a problem, then it is important to know who or what he or she believes is responsible for this problem. The question was open-ended to insure that the response was not biased. The next portion determined whether the respondent uses more or less energy now than he or she did four years ago before the energy crisis of 1973-74. This should indicate whether conservation practices have been inspired by the 33 crisis. Both cost and quantitative amounts told us if they are spending more but using less. This, too, told us if they have knowledge of energy use, especially when consump— tion questions are asked. Next, it was determined if the respondent rents or owns his or her household quarters. If the respondent rents, it was then determined whether he or she pays for the utilities separate from the rent payment. This told us if the respondent is less aware of the energy consumed (if an indirect payment is made) and also the motive for other types of questions. The respondent was then quizzed to determine his or her knowledge of energy consumption. The results qualified the answers to previous energy questions and upcoming informa- tion source questions and to compare respondents by demographics. The next question asks the respondent where he or she gets most of his or her information on the energy situation and lays the groundwork for media credibility questions. This, too, is an open-ended question to prevent bias. These credibility questions began by asking the respondent if he or she received conflicting reports about the energy situa- tion, which source would he or she believe most. It is a close-ended question. The respondent was then asked to rate the truthfulness of various information sources on a scale of zero to 100 percent. These results were used to rank these media and to see if the re3pondent's main source of informa- tion is perceived as his most credible source. 34 The next question asks the respondent if he or she has ever received a newsletter or pamphlet in the mail about the energy situation. This told us if printed literature is being distributed. A follow-up question told us how much of it is being read (if it is received) to determine its effectiveness. The next line of questioning involved cable television as a potential source of energy information. After deter- mining whether or not the respondent subscribes to cable television, those who do subscribe were quizzed about watching the cable origination channels (e.g., weather, sports, and news) and the all-night movie channel. A question was then asked to determine whether the respondent has learned anything about the energy situation at the library. This should indicate if energy information is being distributed at one of the best points for learning. The respondent was asked if he or she has learned any- thing from reading material while in a waiting room; if yes, he or she was asked if energy information is included in what he or she learned. Similar sets of questions are asked in reference to information posted on bulletin boards and on buses or trains. This told us if these sources are potential distribution points for energy information. Questions about other potential methods of distribution I(i.e., energy information sent home with children from school and through clubs and organizations) were asked of the respon- dent. 35 The final questions ascertained demographics. Since it was determined in several studies that income and educa- tion are key determinants of energy knowledge and consumption, levels of both were ascertained. The number of persons in the household is a determinant of energy consumption behavior. Sex and race were also ascertained to determine their signifi- cance in affecting energy attitudes and behavior. The Pretest A pretest was conducted in the Lansing, Michigan, area to check for flaws in the questionnaire. Approximately 30 respondents were selected from the Lansing Telephone Direc- tory using the same technique used in the Grand Rapids sur- vey. Trained interviewers were used in both instances and briefed with the same instructions used for the Grand Rapids study. They were debriefed for problem areas and recommen- dations. Conduct of the Interviews Interviewers were briefed both verbally and in writing about the procedures, objectives of the study, hypotheses, reporting times, dialing instructions, and interviewing in- structions. Last-minute instructions as the survey proceeded were given orally. The general instructions included the purpose of the survey, definitions, and objectives. The research hypotheses are those listed in Chapter II. Specific requirements 36 (i.e., where to report and expected work load) were followed by dialing instructions which specifically outlined the steps for using the WATS lines, the procedures for handling refusals, disconnects, no answers, not-at-homes, busy, and completed interviews. Interviewer instructions discussed courtesy, speech, the alternating of males and females to maintain a credible balance based on demographic characteristics of the Grand Rapids area, and the use of reinforcing comments. Treatment of Data The raw data obtained from the survey was analyzed using the Statistical Package for the Social Sciences (SPSS). Aside from standard frequencies, product-moment coefficients of correlation, chi-square, "t-tests" (difference of means), and analysis of variance (ANOVA) were used to determine the significance of the independent variables. The statistical procedures of correlation are one-way to assess the degree to which two or more variables show inter-relationships in a given p0pulation. Correlation, how- ever, will not give us the reason for a relationship, merely how much they vary together positively or negatively. This index of the degree of relationship is called a correlation coefficient (r). A t-test is used to check for significance. Given a value of ”t," it is interpreted for an associated 'probability level by consulting a sampling distribution of nt.u8 The calculation of ”r" reflects a ratio between the 37 maximal amount of variability that two measures could have in common. The coefficient squared (r2) is the actual pro- portion of variance that two measures have in common; it is useful when considering correlation with reference to predic- ting variables of one measurement from one or more other measurements. In contrast to r, the value of r2 can be thought of in terms of proportions or percentages of rela- tionship (when used for this purpose, r2 is called a coeffi- cient of determination).9 Product-moment coefficients of correlation were computed to describe the degree and direction of relationships be- tween: 1) knowledge of energy consumption and actual consump- tion; 2) sources of energy information and the credibility of those sources; 3) renters paying for their utilities separ- ately from rent payments and knowledge of energy consumption; 4) perception of an energy problem and actual consumption; and, 5) perception of an energy problem and knowledge of energy consumption. For purposes of this study, correlations are arbitrarily labeled: "negative" under : .20, "low” between i .20 and i .40 (definite but small relationship), "moderate" between i .40 and i .70 (substantial relationship), ”high" between i .70 and i .90 (marked relationship), and ”very high" between :_.90 and i 1.0 (very dependable).10 The level of rejection of the null hypothesis is at p<:.05. Chi-square is a nonparametric test which is used in nominal scaling when we are interested in comparing categor- ies among themselves or in contrasting how samples differ in 38 terms of assignment into the categories. It has widespread utility in communications research. In essence, chi-square is best thought of as a discrepancy statistic, meaning its calculation is based upon this discrepancy between the fre- quencies observed for a set of categories and some alternate theoretical set of frequencies.11 Chi-square, then, is one of the simplest and most useful statistical tests. Statis- tical tests are meant to compare obtained results with those to be expected on the basis of chance. Chi-square is a mea- sure of the departure of obtained frequencies from those ex- pected by chance; however, like other statistics that indicate statistical significance, it tells us nothing about the mag- nitude of the relation.12 It has a sampling distribution by which it will be possible to estimate the probability that a given value of chi-square will be expected under the terms of the null hypothesis. The level for rejection of the null hy- pothesis is p<:.05. The t-test is a statistical model used for testing the significance of difference between two pOpulation means, based upon the observed difference between two sample means and their distributions. It is a ratio between the sample mean difference and the standard error of that difference. It tests the null hypothesis against alternative research hy- potheses. Values of t have a sampling distribution used as a basis for estimating the probability that a particular value of t would be expected under the terms of the null hypothesis, or in other words, the sampling distribution tells us what 39 sampling error to expect in values of t. Given a calculated value of t, this value is interpreted for its probability of occurrence in testing a null hypothesis.13 Again, the level for significance is p<:.05. The ANOVA is a test similar to that of the t-test; i.e., it is actually an extension of the difference-of—means test. An ANOVA has the advantage of replacing several tests with a single test.14 This model was used to test more than two groups of answers for statistical significance. It uses var- iances to study the relationship between the independent var- iables and error variance. The level for significance is p<:.05. Single-factor ANOVA is a statistical model used for test- ing the significance of difference among two or more means when these means reflect the consequences of different levels of a single independent (factor) variable. The statistical logic behind ANOVA is incorporated in the "F ratio," a ratio of between-group variance to within-group variance. Given a calculated value of F, this value is interpreted in a sampling distribution for its probability under the terms of the null hypothesis.15 Multiple-factor ANOVA is similar to single-factor ANOVA except it can accommodate more than one independent (factor) variable, and each of these variables can have two or more 7 levels of their own. The multiple-factor model, then, pro- vides methods for testing whether different subgroups, or various combinations of subgroups, represent different 40 populations in terms of what is being measured as the inde- pendent variable; in other words, it is a statistical model for testing the consequences of manipulating two or more in- dependent variables in a single research design. Once again, the F ratio is the statistic used to conduct the appropriate hypothesis test. Significance tests among different levels of each factor are known as main effects. Whatever effects are due solely to the combination of factors are known as interaction effects.16 Summary This chapter discussed the population, selection of respondents, the instrument, the questionnaire, the pretest, the conduct of the interviews, and the treatment of data. The final two chapters will discuss the findings and con- clusions/recommendations. FOOTNOTES 1Michigan Statistical Abstract, ed. David I. Verway (East Lansing,TMicE}: Divisidn of Research, Graduate School of Business and Administration, Michigan State University, 1976), p. 50. 2Registered Community Audit (Atlanta, Ga.: Industrial Development Research Council, 1975). 3Sales and Marketing Management, 1976 Survey of Buyipg Power, p. C-104. 4U. S. Department of Commerce, "Social and Economic Statistics Administration, 1970 (Washington, D. C.: Govern- ment Printing Office, 1970), p. 9. 5Sales and Marketing Management. 6C. A. Moser and G. Kalton, Survey Methods in Social Investi ation (New York: Basic Books, Inc., Publishers, , p. . 7Hubert M. Blalock, Jr., Social Statistics (New York: McGraw-Hill Book Co., 1972), pp. 514-15. 8Fred N. Kerlinger, Foundations of Behavioral Research (New York: Holt, Rinehart, andTWinston, Inc., 1973), p. 149. 9Ibid. 10Frederick Williams, Reasoning with Statistics (New York: Holt, Rinehart, and Winston, Inc., 1968), pp. 127-40. 111bid., p. 112. 12Kerlinger, pp. 166-71. 13Williams, pp. 75-81. lQBlalock, pp. 328-29. 15Williams, pp. 83-93. 16Ibid., pp. 95-111. 41 CHAPTER IV RESULTS In general, this chapter discusses the results of this study in terms of the characteristics of the respondents. Specifically, energy problem awareness, responsibility for the energy problem, energy consumption habits, knowledge of energy consumption, energy information sources, credibility of those information sources, and finally a comparative look at source believability, and source truthfulness will be dis- cussed in terms of income, education, household size, sex, and race. In some instances, the demographic independent variables were analyzed in various combinations to determine their joint significance, if any. Energy Problem.Awareness Since it has been noted that problems cannot be solved until they are recognized, the degree of energy problem aware- ness was determined in terms of independent demographic var- iables, both collectively and individually, using a variety of significant tests. 42 43 Of the 366 cases in the sample, 84.7 percent perceive an energy problem, 9.3 percent do not perceive one, and 6 percent do not know or have no opinion. A standard error of .028 suggests that the public does perceive an energy problem, and therefore, the first hypothesis (H1) is rejected. A Pearson's correlation procedure was used to determine the degree of relationship between the perception of an energy problem.and actual consumption (using more quantities of energy now than four years ago) and between perception of the energy problem and knowledge of energy consumption. In the first situation, a coefficient of -.0268 was found to be insignificant (.305); however, in the second case a coeffi- cient of -.1135 was found to be significant (.015). Other correlations were also computed (Appendix B). Males (88.9%) tend to perceive an energy problem more than do females (80.6%). A t-test indicated this difference is significant at the .05 level. Thus, the second hypothesis (H2) is rejected. There is little difference in the perception of an energy problem based on income levels; in that, percentages range from 83.2 to 87.8. The chi-square of 3.963 is not significant at the .05 level; thus, the third hypothesis (H3) is rejected. Similarly, it was found through analyses of variance on 1) income, 2) education and income, and 3) income and sex that ithere is no significant difference at the .05 level; however, an ANOVA on household size and income (2.386) resulted in a significant difference (.017). 44 As the level of education increases, the more respon- dents perceive a problem. A chi-square test for significance resulted in a chi-square value of 15.587 significant at the .05 level; thus, the fourth hypothesis (H4) is accepted. While it was found that more whites (85.5%) perceive an energy problem than do non-whites (64.7%), a t-test of energy problem awareness analysis and race was conducted and no sig- nificant difference noted at the .05 level. Three times as many non-whites (23.5%) than whites (8.7%) do not believe there is a problem. It was discovered that knowledge of energy consumption is not a significant factor in perceiving an energy problem when an ANOVA was used to check for significance. No signi- ficant difference exists at the .05 level. Responsibility for the Energy Problem In order to identify what demographic characteristics are significant in determining who is responsible for the energy crisis, it was necessary to treat the data using a Pearson correlation procedure as well as an ANOVA and a chi-square test. Of the 310 cases who perceive an energy problem, 26.4 percent blame everyone, 25.1 percent blame consumers, 10.6 percent blame government, 7.7 percent blame the oil indus- V try, 3.5 percent attribute it to scarcity of fuels, and 2.9 percent blame the utility companies; 11.6 percent have no opinion or do not know; thus, the fifth hypothesis (H5) 45 specifying that the oil companies and the Arab Oil countries are primarily deemed responsible for the energy crisis, is rejected. When a Pearson's correlation procedure was applied, it was found that the degree of relationship between education and responsibility for the energy problem is -.1l42, signifi- cant at .014 (Appendix B). The amount of energy used now as compared to four years ago before the energy crisis of 1973— 74 is minimally tied to responsibility (-.1016) and is signi- ficant at .025 while sex is similarly correlated with respon- sibility (-.0921) significant at .039. Race is not signifi- cant at the .05 level (.06). As the level of education increases, less blame is placed on the oil industry while more is aimed at consumers. An ANOVA supports this finding, as education was found to be significantly different at .001 (Table 3). Except for grad- uate students/degree holders, there is an increase in the degree of blame attributed to government as the level of edu- cation increases. Persons with less than a high school di- ploma either have no idea of who is responsible or place the blame on everyone; 41 percent of the persons with a baccalaur- eate also attach the blame to everyone. An ANOVA was applied to income and responsibility, and it was found that the level of income is not significant at the 1.05 level. When income was analyzed with the size of household, there was a significant difference for income at .046. Simi- larly, when income and education were analyzed jointly, there 46 was a significant difference for education at .001. Such is not true with income and sex (Table 3). Race-wise, caucasians split the blame between everyone and consumers (26% each) followed by government (10.5%), the oil companies (7.8%), and big factories/businesses (5.4%). Blacks (n=ll) place equal blame on government and everyone (18.2% each) then on consumers, the utility companies, and the scarcity of fuels (9.1% each). A chi-square test was applied to the variable of race, finding it significant at the .05 level. Energy Consumption Habits In determining the degree of energy use now as compared with four years ago (before the energy crisis), respondents' perceived consumption habits were determined and any differ- ences checked for significance at the .05 level. Cost-wise, 67 percent of the respondents believe thay are spending more for energy now than they were four years ago while 6.3 percent report spending the same, 13.1 percent less, and 2.4 percent don't know. As expected, 46.4 percent of the respondents report using less amounts of energy now than they were four years ago; 16 percent report using the same amount, 25.2 percent more, and 1.2 percent don't know. A standard error of .047 suggests that the public does per— * ceive using less energy now, and therefore, the sixth hypoth- esis (H6) is accepted. Both the amount being spent on energy 47 Table 3 ANOVA OF ENERGY PROBLEM RESPONSIBILITY Independent Variab1e(s) __F__ Significance Energy knowledge (Quiz) .136 .938 Education 5.549 .001* Household size .509 .769 Income 1.378 .250 Income and sex: Main effects 1.031 .391 Income 1.356 .257 Sex .246 .621 2-way interactions 2.087 .102 Household size and income: Main effects 1.409 .192 Household size 1.157 .331 Income 2.697 .046* 2-way interactions .911 .552 Education and income: Main effects 3.449 .001* Education 4.970 .001* Income .458 .712 2-way interactions 1.560 .066 n=366 *Significant at .05 level. 48 and the quantities used now versus four years ago show a de- gree of relationship with consumption habits (pf§.001) (Appendix B). There is also a weak link (-.1016) between the amount of energy used and responsibility for the energy problem (.026). The size of the household does not affect the amount of money spent on energy. Except for one household that has nine members, all other sizes of households have at least two-thirds recalling spending more for energy. An ANOVA of household size was not significant at the .05 level. A simi- lar application was made to the quantities of energy used, and it was determined that the results lack significance (.306). A majority (52.2%) of the respondents report using less energy except for households of six persons (40%); 19 percent report using the same and 28.4 percent more. The above facts suggest that smaller households tend to use less energy; therefore, the seventh hypothesis (H7) is rejected. All income levels think they are paying more for energy now than they were before the 1973-74 energy crisis. As the level of income increases, the more persons recall spending and the more they report using. Of those who earn at least $8000 (n=37), 54 percent report using less energy and 28 per— cent more. An ANOVA of cost suggests that the level of income is significant (.032); however, the results applying to amounts 9 are not (.689). Thus, the eighth hypothesis (H8) is rejected (Table 4). Based on education, a majority of respondents (52.2%) 49 are using less quantities of energy now than before the energy crisis. An ANOVA procedure conducted on education level means revealed a lack of significance at the .05 level (Table 4) in reference to both cost and amounts of energy used, thus suggesting that the level of education does not have any effect on the perceived amount of energy consumed; therefore, the ninth hypothesis (H9) is rejected. When edu- cation and income were combined in a multiple-factor ANOVA, only income was found to be significant (.026), cost-wise; however, both were significant when dealing with quantities used on a two-way interactions (Table 4). Both white (75.9%) and non-white (58.8%) respondents think they are spending more for energy now. Whites (53.6%) report using less amounts of energy; whereas, non-whites (45%) report using more. A t-test resulted in a significant difference at the .05 level for quantities of energy used while a lack of significance when dealing with the cost of energy. In reference to sex, a t-test ended with no significant difference between males and females for either cost or quantities used. Two-thirds of the respondents own the quarters in which they now live while 21.6 percent rent. Of those who rent, 86.4 percent pay for their utilities separately from.the rent 1 payment. There is a high correlation (-.8799) between rent- ing or owning a home and paying for utilities (pf§.001). As the size of a household increases, the more the tendency to 50 Table 4 ANOVA OF ENERGY CONSUMPTION HABITS Independent Variable(s) F Significance Education: Cost .535 .710 Amounts .385 .819 Household size: Cost 1.893 .095 Amounts 1.207 .306 Income: * Cost 2.969 .032 Amounts .491 .689 Income and sex: Main effects (cost) 2.175 .072 Income 2.703 .046* Sex .300 .584 Main effects (amounts) .735 .569 Income .389 .761 Sex 1.709 .192 2-way interactions: Cost 1.397 .244* Amounts 3.720 .012 Household size and income: Main effects (cost) 1.870 .065 Household size 1.384 .230 Income 2.198 .089 Main effects (amounts) .880 .533 Household size 1.040 .394 Income .348 .790 2-way interactions: Cost .765 .716 Amounts .599 .875 Education and income: Main effects (cost) 1.689 .111 Education .672 .612 Income 3.123 .026* Main effects (amounts) .533 .810 Education .552 .697 Income .823 .482 2-way interactions: Cost 1.740 .058* Amounts 1.921 .032 *Significant at .05 level. n=366 51 own the quarters in which a person lives. Households of one or two persons tend to rent (56% and 50%, respectively). There is a low correlation (-.2758) between household size and making utility payments (pj§.001). An ANOVA.was con- ducted for significance of household size means. There was no significant difference at the .05 level. As the level of income increases, the more the tendency to own a home. Of those making more than $8000 a year, 78.5 percent own their own home while those making less than $8000 own 56.1 percent of the time. When analyses of variance were conducted on 1) household size and income, 2) income and sex, and 3) edu— cation and income, there were no significant differences at the .05 level. In reference to utility payments, a t-test was applied to both sex and race, and it was found that there are no sig- nificant differences at the .05 level. Energy Knowledge Depending on the scale of measurement, significance tests were used to determine the range of a respondent's information or understanding of energy consumption as determined by suc- cessfully answering five empirically verifiable energy use questions. Demographic independent variables were both indi- vidually and jointly analyzed. In general, respondents faired poorly with the energy knowledge quiz, missing three of five questions on the aver- age. When asked whether a refrigerator, color TV, air 52 conditioner, or gas heater uses the most energy, a plural- ity of 44 percent incorrectly responded air conditioner. The second most selected answer, the gas heater (26%), was actually the correct answer. When asked which type of trans- portation uses the least energy per person for long-distance trips, a plurality of respondents incorrectly said the train instead of the bus, the correct answer. Only three of 10 re5pondents correctly said that the ceiling is where most heat is lost; a plurality of 44 percent incorrectly said the windows. When asked which requires the most energy for most families, 69.9 percent of the respondents correctly said heating a house. A plurality also correctly answered the last question asking which type of housing of the same size and insulation requires least energy to heat; 49.6 percent said the apartment. Sexewise, both a plurality of males (35.3%) and females (38.4%) answered an average of two questions out of five cor- rectly. A t-test to determine significance was applied in this case, and it was found that there was a significant dif- ference only in the case of question two (.033) at the .05 level. A chi-square treatment was applied to the quiz collec- tively, and it too indicated that females do not necessarily know more about energy use than do males, and therefore, the tenth hypothesis (H10) is rejected. Similarly, a plurality of people with all levels of edu- cation averaged only two questions correct although there is a steady increase in the percentage of persons who answered 53 three questions correctly as the level of education increases. No one with less than a high school diploma got four or five questions right. There is a correlation between education and energy knowledge (.1485) significant at .002; however, an ANOVA treatment of all five questions collectively resulted in no significant difference at the .05 level, suggesting that persons with higher income do not necessarily know more about energy consumption; therefore, the eleventh hypothesis (H11) is rejected. An ANOVA treatment applied to income and energy know- ledge (as determined by empirical quiz) resulted in no signi- ficant difference at the .05 level. When education and in- come were jointly applied to energy knowledge, there was no significant difference at the .05 level (.08). This suggests that high-income households do not necessarily know more about energy consumption than low-income households, and there- fore, the twelfth hypothesis (le) is rejected. Households with one to six persons answered, in general, only two questions correctly. There were not enough cases in households with seven to nine persons to make any infer- ences. No household of one or two persons answered all of the questions correctly although two-person households cor- rectly answered four questions more than did any other size household by a two-to-one ratio. When a Pearson's correla- tion procedure was used on household size (.0787), it was found there is no significant difference at the .05 level (.066). Similarly, there is no significant difference at 54 the .05 level when a multiple-factor ANOVA of household size and income are combined in one treatment. In reference to race, a chi-square treatment resulted in no significant difference at the .05 level. The majority of respondents, regardless of the number of questions answered correctly, blame consumers and everyone in general for the energy problem. A chi-square test was conducted on energy knowledge and several pertinent independent variables (i.e., those who per- ceive an energy problem versus those who do not perceive such a problem, renters versus owners, high users of energy versus those who have made positive adjustments, sources of energy information, and those who pay for utilities separately from their rent payment versus those who have utilities included as a part of their rent), and in all cases, there is no sig- nificant difference at the .05 level. Energy Information Sources The main sources of energy information are the newspaper (39.3%), followed by television (30.7%) and remotely by mag- azines (8.9%), friends (4.4%), and radio (4.2%), suggesting that television is not perceived as the main source of energy information; therefore, the thirteenth hypothesis (H13) is rejected. Less than half (42.6%) of the people surveyed recall having seen a newsletter or pamphlet in the mail about the situation; 52.2 percent answered "no." Of those who had read a newsletter or pamphlet (n=65), 41.4 percent reported 55 having read all of it, 29.9 percent most of it, 22.9 percent some of it, and 5.7 percent none of it. Only 11.0 percent of the respondents subscribe to cable television in the Grand Rapids area, due mainly to the limited amount of construction completed thus far. Many areas still are not wired for cable service. Of those who do subscribe, 89.7 percent have watched the cable weather, news, or sports channels at least once in the past; whereas, only 27.5 per- cent of the cable subscribers have watched the all-night movie channel at least once. Since there are so few persons (n=40) involved in the cable sample, inferences must be viewed with caution. Eighty-six percent of the respondents have not learned anything about the energy situation at the library. A major- ity (56.1%) have, however, learned something (about any sub- ject) from reading materials distributed in waiting rooms (e.g., a physician's office). Of those who have learned some- thing, 84.5 percent (n=l74) recall having learned about the situation regarding the energy problem. The success of bul- 1etin boards is not as good, as only 40 percent of the respon- dents have learned something (any subject) from information posted on bulletin boards. Of those, 70.7 percent (n=104) recall having learned something about energy. Only 21 per— cent of the respondents have learned something (any subject) from information on signs on buses or trains. Of those that have, 57.7 percent (n=45) recall having learned something about the energy situation. 56 In reference to learning about energy from information sent home with children from school, 39.1 percent (n=93) of those with school-aged children have learned something about energy through their offspring. Approximately the same suc- cess (34.6%) has been achieved with information distributed through clubs and community organizations (n=127). Sex A plurality of both men (40.5%) and women (38.1%) choose the newspapers as their main source for energy information followed by television (26.8% and 34.4%, respectively) then magazines (13.1% and 5.3%, respectively). This suggests that males' main sources of information is the newspaper; there- fore, the fourteenth hypothesis (H14) is accepted. Likewise, the findings suggest that women also use the newspaper rather than the television as their main source of energy information; therefore, the fifteenth hypothesis (H15) is rejected. Additional t-tests were conducted to determine the differ- ence of means based on sex. A lack of significance was noted with all other sources of energy information (i.e., newslet- ters or pamphlets, cable subscription, watching cable news or all-night movie channels, library, waiting rooms, bulletin boards, signs on buses or trains, literature sent home with children from school, and clubs or community organizations). Less than a majority of both men and women recall having re- ceived a newsletter or pamphlet in the mail about the energy situation. Of those who do receive such items in the mail, 57 73.4 percent of the males and 68.4 percent of the women recall having read them all or most of the them. Almost 91 percent of the 22 female and 87.5 percent of the 16 male subscribers recall having watched the cable weather, news, or sports channels. Only 37.5 percent of the males and 18.2 percent of the females watch the all-night movie channel. Men and women both have learned about energy at the library at about the same rate (14.0% and 13.1%, respective- ly); however, this statistic was surprisingly low. The re- sults for learning about energy from printed materials dis- tributed in waiting rooms were more promising, as 59.1 per- cent of the men and 52.9 percent of the women recall having learned something (any subject). Of those who have learned something, 89 percent of the men and 79.4 percent of the women have learned something about energy from.that material. Sixty-four percent of the women and 56 percent of the men recall having learned something (any subject) from.bulletin boards. Of those who have learned something, 72.1 percent of the females and 70.7 percent of the males have learned about energy. The percentage is low (24% of the men and 18.3% of the women) for those who have learned anything at all from signs on buses or trains. Two-thirds of the females and 51 percent of the males have learned about energy from.these signs. Of those who have school-aged children, 42.9 percent of the females recall having learned something about energy from.information sent home with their children from school while 35.2 percent of the males have learned. Nearly 58 two-thirds of the females and 64.3 percent of the males have learned something about energy through clubs or community organizations. Income Those persons making more than $8000 a year select the newspaper (42.3%) as their main source of energy information followed by television (29.8%); those making less than $8000 rely more on television (41%) than newspapers (17.9%); there- fore, the sixteenth hypothesis (H16) is accepted. The number of persons who recall having received a news- letter or pamphlet in the mail is mudh less for those making under $8000 (29.3%) than those above $8000 (45%). There is no significant trend for levels above $8000 as determined by an ANOVA for both the main source of energy and the receipt of newsletters or pamphlets in the mail. As the level of income increases, there is a decrease in cable subscription rates although there is little differ- ence between those making less than $8000 (14.6%) and the $8000-$15,000 (15.2%) income levels. One hundred percent of all income levels watch the cable news, weather, and/or sports channels except the $8000-$15,000 level which has an 80-percent viewership. As the level of income increases, there is an increase in those who watch the all-night movie channel (n=10). The correlation coefficients determining the degree of relationship between cable subscription, cable news, 59 and the movie channel with income are significant at the .05 level. A Pearson correlation procedure resulted in no signifi- cant correlations between income and learning about energy at the library, in waiting rooms, from bulletin boards, from signs on buses or trains, from information sent home with children from school, or from.clubs or community organiza- tions (Appendix B). There are no trends that are significant based on income levels with regards to learning about energy at the library. Of the 23 respondents making under $8000 a year and who learn something in waiting rooms (56.1%), 87 per- cent recall learning about energy; of the 172 persons making more than $8000 who have learned something (any subject), 84.9 percent learn about energy. A higher percentage (86.7%) of those making less than $8000 recall learning about the energy situation from bulletin boards than those above $8000 (68%). Those above $25,000 recall learning about energy at 80 percent from these signs. There is no significant differ- ence for households bringing in less than $8000 a year versus those above $8000 when it comes to learning about energy through information on signs on buses or trains or information sent home with children from.school. The $8000-$25,000 income brackets have learned less about energy from clubs or community organizations than those in the below-$8000 and above-$25,000 brackets. An ANOVA treatment of education level and information source is not significant at the .05 level, suggesting that 60 persons with low education levels do not necessarily use television as their main source for energy information more than upper education levels; therefore, the seventeenth hy- pothesis (H17) is rejected. The correlations between receiving newsletters or pam- phlets and education levels and between reading these printed items and education are significant at the .05 level. There is no significant trend for the various education levels and watChing cable channels. Those persons with at least some college have learned more at the library than those with a high school diploma or less. There is a significant but low correlation (-.l457) between education levels and learning about energy at the library, but no such significance is attached to learning about energy from printed material while in waiting rooms. Only those persons having done graduate work have a majority which learn from bulletin boards. Here again, there is a def— inite but small relationship (.2402) significant at the .001 level. Such is not the case with signs on buses or trains; there is no significance with the relationship between learn- ing about energy from information sent home with children from school and education; the correlation coefficient is negative. When learning about energy from clubs or community organiza- tions, the level of education increases the more respondents report having learned something about energy. This has a def- inite but small relationship (-.2035) at the .05 level (.001). 61 Education and Income, Household Size and Income, and Income and Sex A multiple-factor ANOVA was applied to 1) education and income, 2) household size and income, and 3) income and sex to ascertain the significance of dual factors. In each case, there were no significant variable or combinations of varia- bles at the .05 level. Race A plurality of whites (40.5%) get most of its information from newspapers and secondly from television (30.8%); whereas, non-whites get most of their information from television (31.6%) then neWSpapers (15.8%). A chi-square treatment re- sulted in a chi-square value of 177.391, significant at the .05 level. A t-test was first administered to determine the differ- ence of means between white and non-whites who have received a newsletter or pamphlet in the mail then the degree to which it has been read. In both instances, there was no signifi- cant difference at the .05 level. There is a higher rate of cable subscription for whites (90.7%) than non-whites (40.9%); however, a majority of blacks (52.9%) do subscribe. A t-test was applied and found to be significant (.000) (Table 5). Another t-test, however, revealed no significant differences between whites and non- whites in viewing either cable news, sports, or weather chan— nels or the all-night movie channel. A t-test was used once 62 Hp>ma mo. up paeueweeweme mm. mam. em em.e- Ono. em Apeeeev Hpeemep peeoa unmeeuaee seep: mm. . «mm. em mm. Nam. em memeeeee papa enema seem: No.e eooo. men mm. ohm. mmm eoeeeepumeem meeeo mm.eu Noe. emu o3.eu ema. mme mppeeeaee\mppeepemspa came SN. mom. eem ee.e- emu. oem Heme pee pepeeeme\epppepemspz me. use. eem mm. I mam. oem meoeppueepweo eueepesou pp meeeo oe. eem. eem Ne.e 3N3. oem Hooepe seem empeeeee £uH3 mac: udmm coaumEHomcH eo.eu owe. me mm. mom. ms meeppe po pupae no memem we. see. see mm. . «em. See menace peepeaem no.3- mmo. mom oo.~- eemo. com peace meeeeez we.e oeo. eem em. . eoe. oem eempeee mm.~- moo. Nem me. ewe. wmm pepeom eoeppapomee ewppem mmwmwum epeeeeeeope .Iwmu peem>ue meeeeemeope we paepeee> ueeeemeee epeeepupeo epeemeupeo comm xmm mMUMDOm ZOHHe. 2132:. ~0. an.. a99m s.3:mn:cz 9...- co..- .9)». 999994399 00. 99.. .au.0\923.u-»ncu:u 9.90. we. c:.- 6990..zu .zum:9u.o=. :99». n0.: 00.: I=.=.w\uO::r-Axuscm sent. me. 90. 99.9.h.asna:-m:_z.9c< 99.2. .3.- 91 .330 5.9.9.1179.- ..25 9.9!... cm.. .o.-.-n¢ c999..:a-¢=.:99=« 9990. 99.- 99. 993:: 999.993-999999 c..s. .9.. 9c. usage 9993.13-99.:.>9< 9.19. .0. «0.. 9uusnua-xfiuucm noon. 59:0. 99. 99. .99 9.99: 999.9...9 9.9:: «0.: no. I—ucasgu .10: 9.930 599-: .0.: .0. cc9ua9uu-nam =o9a.>s.nsu ...: 90. ouaueaI-sxauuuus~uaoz vnsu 0~. n0.- .lnmxautuuo.r)vz nay—nus: no.- «9..- xu..9a9nu~0 .90 .911-sou .9. 99.- 9......99.9 .99 9.....9 99. 99.- 9......99.9 99-9nct 999 99. .9..- 999.99.9990 9991999999 99...:9vb.u 9099.939: 99.- .9.- 9......9o.9 >9 90.: no.- co99q19999. mc9uu...csv as. _=.- nau930m =99.nus=w u: 9.30 : 99n.: Io.aonm no. 999.999acoauau 9.9.: : auscaceac launch; 90.9cu -9u.m..-w mm o a n u I 3 I " d.M m I 14 U l 9:. c s a C. pl 3 d I q C J u n w v) I .9 u 91 LIST OF REFERENCES LIST OF REFERENCES Abel, John D. and Wirth, Michael 0. "Newspapers or Televi- sion: Which is MOre Credible as a Source of Local News?" A special study conducted at Michigan State University, 1976. (Mimeographed.) Armbruster, Frank E. The Forgotten Americans: A Survey of Values, Beliefs, and Concerns of the Majority. New ROcheIIe, NT’Y}: Arlington House, 19727 Baldwin, Thomas F. and Gluck, Jon K. "Adaptive Behaviors of New Gable Television Subscribers." A special study conducted at Michigan State University, 1971. (Mimeographed.) Beall, J. Glenn, Jr. (Sen.). "State-wide Poll on Public Issues." A special study conducted by the Republican Party in Maryland, 1975-76. Blalock, Hubert M., Jr. Social Statistics. New York: McGraw-Hill Book Co., I972. Block, Martin. "Continental Cablevision.‘ A special class study conducted at Michigan State University, 1976. (Typewritten.) Brownell, William. "Ecology Kick." Fort Benning(Ga.) Bayonet. Sept. 8, 1972. Chase, Stuart. American Credos. New York: Harper and Brothers, Publishers, 1962. Dangerfield, Linda A., McCartney, Hunter P., and Starcher, Ann T. "How Did Mass Communication, as Sentry, Perform in the Gasoline 'Crunch'?" Journalism Quarterly 52:2 (Summer 1975): 316-25. Daniel Yankelovich and Associates. "Consumer Attitudes To- ward the Natural Gas Industry." An unpublished report to the American Gas Association. Washington, D.C., 1975. 92 93 Edison Electric Institute. "Load Management: Its Impact on System Planning and Operation, Phase I." A report of the EEI System Planning Committee. New York, April 1976. Extension Staff Services staff meeting, Dec. 9, 1976. Farace, Richard V. "Upgrading Energy Information: Prelimin- ary Guidelines for Michigan." A paper presented to the 1976 Summer WorkshOp on Energy Extension Services, Lawrence Laboratory, University of California, Berkeley, July 18-25, 1976. Federal Energy Commission. The Arab Embargo: Two Years Later. Washington, D.C.: U.S. Government Printing ozrrae, 1975. Foote, Cone, and Beldin . "Energy Conservation Practices by Chicago Consumers.’ An unpublished report to the People's Gas Company, Feb. 3, 1976. Funkhouser, G. Ray and Maccoby, Nathan. "Communicating Specialized Science Information to a Lay Audience." The Journal of Communication 21:2 (Spring l97l):58-7l. Gallup Poll No. 99. Jim.Schriner, ed. Michigan State University, 1976. Harris, Louis. "Poll Reflects Energy Concerns." Science Digest 78 (Nov. l975):16-17. Heberlein, Thomas A. Conservation Information, the Energy Crisis, and Electricity Consumption in an Apartment Co lex. ‘Madison,’Wisc.: Department of Rural Socio- Iogy, University of Wisconsin, 1974. Hogan, Mary J. "Energy Conservation: Family Values, House- hold Practices, and Contextural Values." Ph.D. disser- tation, Michigan State University, 1976. Institute for Family and Child Study. Family Energy Project no. 3152. East Lansing, Mich.: Michigan State Univer- sity, May 1976. Kerlinger, Fred N. Foundations of Behavioral Research. New York: Holt, Rinehart, and Winston, Inc., 1973. Mendelsohn, Harold. "Some Reasons Why Information Campaigns Can Succeed.” Public Opinion Quarterly 32:1 (Spring l973):50-6l. 94 Michigan Statistical Abstract. David I. Verway, ed. East Lansing, Mich.: Division of Research, Graduate School of Business and Administration, Michigan State Univer- sity, 1976, p. 50. Morrison, Bonnie M. and Gladhart, Peter M. "Energy and Families: the Crisis and the Response." Journal of Home Economics (Jan. 1976):16. Moser, C.A. and Kalton, G. Survey Methods in Social Inves- tigation. New York: Basic Books, Inc., Publishers, National Association of Science writers. "The Public Impact of Science in the Mass Media." A special study con- ducted by the Institute for Social Research, Ann Arbor, Mich., 1958. Registered Community Audit. Atlanta, Ga.: Industrial ’Development Research Council, 1975.. Response Analysis. "The Electric Utility Industry Today. A privately printed report to the Public Information Research Program, Edison Electric Institute, Princeton, N.J., Dec. 1975. Sales and Marketing Management. 1976 Survey of Buying Power, p. C-lOO. Schriner, Jim. Communicating Energy Information. East Lan- sing, Mich.: Michigan State University, 1976. I Smith, Richard B. ”Household Energy Adjustments.’ A speech made to the National Agricultural Outlook Conference, Washington, D.C., Nov. 17, 1976. Stearns, Mary D. The Social Impacts of the Energy Shortage: Behavioral and Attitude Shifts. Springfield, Va.: National7TechnicaliInformation Service, 1975. The Roper Organization, Inc. Research conducted for a private oil company, 1976. Tichenor, P.G., Donohue, G.A., and Olien, C.C. "Mass Media Flow and Differential Growth in Knowledge." Public Opinion Quarterly 34:2 (Summer 1970):159-70. U.S. Council on Environmental Quality. Sixth Annual Report on the Council on Environmental Quality, 1976. U.S. Department of Commerce. "Social and Economic Statis- tics Administration, 1970. Washington, D.C.: U.S. Government Printing Office, 1970, p. 9. 95 Wade, Serena and Schramm, Wilbur. "The Mass Media as Sources of Public Affairs, Science, and Health Knowledge." Public Opinion Quarterly 33:2 (Summer 1969):l97-209. Warren, Donald I. "Individual and Community Effects on Response to the Energy Crisis of Winter 1974: an Anal— ysis of Survey Findings from Eight Detroit Area Commun- ities." Ann Arbor, Mich.: Program in Community Effec- tiveness, Institute of Labor and Industrial Relations, University of Michigan, 1974. Williams, Frederick. Reasoninggwith Statistics. New York: Holt, Rinehart, and Winston, Inc.,71968. Winett, Richard A. and Nietzel, Michael T. ”Behavioral Ecology: Contingency Management of Consumer Use." University of Kentucky, 1975. (Paper to be ublished in American Journal of Community Psychology.§ Zuiches, James L. "Acceptability of Energy Policies to Mid- Michigan Families." Research report no. 298 from.the Agricultural Experiment Station Project 3152, Michigan State University, March 1976.