WNWNW1NWWIMWWWWWI I-_'(-’-I Io: mcoco lllllllllll lllllllllllllllllllllllll 3 1293 01391 8200 This is to certify that the thesis entitled LOOK WHO'S LAUGHING: THE IMPACT OF EDITORIAL CARTOONS ON GROUPS WITH DIFFERENT LEVELS OF POLITICAL ACTIVITY presented by Andrew Halldorson has been accepted towards fulfillment of the requirements for Master's degree in Journalism C.- i ‘ Major professor Date I] i)a}q5— 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution . LIBRARY Michigan State University PLACE IN RETURN BOX to remove We checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE MSU leAnNflrmetive ActioNEquel Opporumny Intuition Wm: LOOK WHO’S LAUGHING: THE IMPACT OF EDITORIAL CARTOONS ON GROUPS WITH DIFFERENT LEVELS OF POLITICAL ACTIVITY By Andrew Halldorson A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF JOURNALISM School of Journalism 1995 ABSTRACT LOOK WHO'S LAUGHING: THE IMPACT OF EDITORIAL CARTOONS ON GROUPS WITH DIFFERENT LEVELS OF POLITICAL ACTIVITY By Andrew Halldorson Though some speak of the power of editorial cartoons, little is known about their impact. Existing studies focusing on general pOpulations suggest it is relatively small. This study began with the assumption that there are different impacts on groups with different levels of political activity. Other population variables were also considered. Surveys designed to measure the thought. emotional and behavioral impacts of editorial cartoons and people's perceived understanding of them were mailed to members of the Michigan House of Representatives. Results were compared to a phone survey of subscribers to The.Mbnroe Eyening was. Comparisons were also made within each survey group. The evidence showed that politically active. educated peOple were more likely than others to say they understand and are impacted by editorial cartoons. ACKNOWLEDGEMENTS Associate Prof. Frederick Fico. of the Michigan State School of Journalism. encouraged and guided this study from the beginning. My special thanks to him and everyone else who helped. iii TABLE OF CONTENTS List of Tables v List of Figures vi 1. Introduction 1 2. Theoretical Framework 4 A Model of Cartoons Effects 6 Influences on Cartoon Impact 8 Hypotheses Model 10 Statement of Hypotheses 11 3. Method 12 Concept Definitions 13 Choosing the Monroe Sample 15 Validity Testing 16 4. Results 18 Average Results for Individual Questions 18 Variable Construction 26 Hypotheses Examined 29 5. Discussion 35 Notes 39 Appendix A - Variable Measures 43 Appendix B - Validity Testing 47 Appendix C - Editorial Cartoon Sample 49 Sources Consulted 60 iv Table Table Table Table Table Table Table Table Table Table H (DQQODUIACDN LIST OF TABLES Exposure 19 Perceived Understanding 19 Behavioral Impact 20 Emotional Impact 21 Thought Impact 22 Political Activity for Monroe 23 Political Activity for Legislators 24 Demographics 25 Mean Scores 30 10 - Comparison of Correlations 31 LIST OF FIGURES A Model of Cartoon Effects 10 vi \\\\\ 3‘3 .. \ \ u ‘.‘ \“ . . ~ Boss Tweed by Thomas Nast INTRODUCTION Many cartoonists have touted the power of political cartoons to expose corruption or stupidity in government. Nineteenth Century cartoonist Thomas Nast. who has been widely credited with helping to bring about the downfall of Boss Tweed and his Ring in New York City. is the quintessential example.1 The best witness to the power of Nast's pen was Tweed himself. who sent threatening letters to Nast and. that failing, offered him $500,000 to stop his satirical attacks. As Tweed said. "I don’t care so much what the papers write about me. My constituents can't read; but damn it, they can see pictures!" 2 The world of American editorial cartoons has changed drastically since Nast’s time; perhaps more so than it had changed by the time Nast had come upon the scene. The oldest know American political cartoon dates back to a crude woodcut drawn by Benjamin Franklin in 1747. Though several interesting samples of early political cartoons have survived. their impact was severely limited by their rarity. One historian counted only 78 handbill caricatures made before 1828. and cartoons in newpapers were even rarer.3 Newsprint was extremely expensive, and it took great pains and long advance notice to engrave cartoons into copper or wood for printing. Lithography. a much cheaper printing 1 2 process. did not appear in America until 1829 and did not become common until the 18503. At that time. the biggest producer of lithographs was Currier & Ives. which had a neutral political platform. In addition to doing many other nonpolitical prints. Currier & Ives was hired by both Democrats and Republicans. Meanwhile, American political cartoons in general suffered through years of creative bankruptcy as artists rehashed old ideas or copied European poltical cartoons with only minor changes.‘ American cartooning did not come into its own until the rise of the great weekly humor and politics magazines such as Harper's Weekly. which Nast Joined in 1862. In the early 18708, when Harper's and Nast led their campaign against Tweed. the magazine’s readership tripled.5 Weekly cartoons were popular parts of such magazines. Harper’s was overtaken in prominence in later decades by such weekly magazines as Puck and The Judge. which also featured accomplished cartoonists. But by the turn of the century. the playing field again changed dramatically. Daily newspapers discovered the power of cartoons to attract readers and developed the technology to print them. Now, instead of being limited to a handful of New York weekly magazines. readers could pick from the hundreds of different cartoonists from newspapers all over the nation. The number of choices and the number of cartoon voices only increased as the decades wore on. More recently, social science research has brought the potency of political cartoons into question. Researchers have found satire less effective in changing people’s 3 opinions than more straightforward means of persuasion.8 Others have surveyed general populations and found most people may not understand political cartoons. and that political cartoons are less effective in changing opinions than written editorials.7 Although little research has been done on the impact of political cartoons. the results of existing research suggest cartoons are to be regarded more as harmless entertainment than as influential political commentary. Previous studies on the impact of political cartoons have focused on general populations and appear to assume that political power in a democracy comes from the masses. Under this view. any form of communication that is not widely embraced or understood cannot be considered influential. Because research has found political cartoons to be relatively ineffective among general populations. estimates of the impact of political cartoons have been lowered. This view is nearly the opposite of the older assumption that cartoons are influential because they have wide popular appeal among the electorate and are easier to "read" than words. This study starts with a contrary assumption: that most political power is in the hands of the few, not the many. It would logically follow - and the study hopes to show - that the impact of political cartoons is different among different politically active groups in a general population. The approach will not be "How do political cartoons affect everybody?" Rather, it will be: "What impact do political cartoons have on politically influential people in society?" THEORETICAL FRAMEWORK In 1968. Lawrence H. Streicher said there have been many commentaries on political cartoons and caricature. but cited a need for a body of theory on political caricature. As an example. he noted a study of Mexican political caricature where the author said caricature serves to organize mass hostilities and aggressions in a society. But how caricature can do such a thing was neglected by the author. and it has been neglected in other political cartoon studies. Streicher said.8 Martin J. Medhurst and Michael A. Desousa answered Streicher’s call in 1981 with the "initial step" of categorizing and naming the various means of graphic persuasion that appear in political cartoons.9 Their categories are patterned after those of classic oral rhetoric. This is a valuable breakdown and analysis of the various communication tools cartoonists use. but it makes no attempt to discover the effectiveness of each tool or the impact of political cartoons in general. Since the mid-19605, researchers have studied many aspects of political cartoons. but research on their impact has been scant. As a result. much is known about themes such as how cartoons are created. how topics are chosen and cartoonists’ 4 5 working conditions.10 What is not clear is why people should care about political cartoons. other than for aesthetics. There have been a few audience-analysis studies that address the impact of political cartoons. Two are frequently cited in research articles dealing with political cartoons. In 1968. Del Brinkman found that political cartoons are less likely to change readers' opinions than written editorials. He also found that a cartoon and written editorial combined will change more opinions than an editorial alone.11 Also in 1968. Leroy M. Carl tested whether people understood the meanings of political cartoons. as defined by the artists who drew them.12 Carl surveyed samples of people in two small towns and a university town. In the small towns. 70 percent of the people polled completely missed the artist-defined meanings of a random sample of 18 editorial cartoons. The other 30 percent either agreed with the artists’ intended meanings. or were in partial agreement with them. In Ithaca. N.Y.. home of Cornell University. 63 percent completely missed the meanings of a partially different sample of cartoons. Twenty-two percent fully understood them and 15 percent partially understood them. Together, Carl and Brinkman have cast doubt on the notion that editorial cartoons are easy to understand or are more influential than other means of persuasion. There is also further evidence in a related field that adds credence to Carl's and Brinkman’s findings. In 1965. Charles R. Gruner was unable to show that exposure to written satire will change attitudes. After further research. Gruner found 6 that exposure to written satire could make a significant difference if researchers explained the point of a satirical work before subjects read it.13 Gruner's research also cited studies that found satire less persuasive than direct prose. It is a reasonable assumption that visual satire shares the persuasive weaknesses Gruner found in written satire. There have been no follow up studies to Carl and Brinkman. Content research on political cartoons is going on year by year with only meager evidence on the persuasive impact of the cartoons.14 For example. a 1992 content study of political cartoons on the Cold War comes with this disclaimer:15 Note that we do not argue for the special influence of cartoons on newspaper readers. Indeed. if their audience were a cross section of the population. they could hardly have much influence since most peOple are unable to graSp the cartoonist's message (here. Carl's 1968 study is cited). Other times, studies seem to ignore the existing evidence. In a 1988 study of how black boxers were stereotyped in old newspaper cartoons. the author implies that the study is important because "newspaper cartoons wield considerable power." As evidence. he only cites a cartoonist and historians who say 30.18 A Model of Cartoon Effects The continuing interest in political-cartoon research shows the need for more information about their impact. yet the studies of Brinkman and Carl have only scratched the 7 surface. Many important questions are unanswered. If. for example. it is true that only a fraction of the population understands the artists’ "meaning" in cartoons. who are the minority who do understand them and what is their role in society? Do some cartoons have more than one valid meaning? And if a person totally misses an artist’s intended meaning. does that mean the cartoon has failed to influence that person? It may be true that political cartoons are less effective in changing opinions than written editorials. But that does not rule out the idea that cartoons have an important impact. It may be said that most cartoons are not written for the purpose of changing opinions, but to serve other functions such as humor. showing a complex idea in stark simplicity. or making a poignant statement that requires no specific reaction from the reader other than thoughtfulness. It is also important not to define "impact" too narrowly. Political cartoons should not be measured as if they were written editorials or solely verbal assertions. As a unique art form, political cartooning has its own unique powers. Oliver Harrington. who created Bootsie. one of black America's favorite cartoon characters in the 19308. said his drawings probably were never taken as seriously as written words. Yet Harrington also said he has been told his cartoons were the first thing many people turned to whenever they got their newspapers, and he said cartoons stick in people’s minds long after they have forgotten essays and 8 written arguments. The impact of his cartoons was vividly shown in a story he told about his days as a correspondent in Italy during World War II. During visits with black troops. he said he would see his clipped cartoons stuck onto the clay walls of their dugouts as one of the soldiers’ reminders of home.17 Influences on Cartoon Impact There are probably many factors that are useful in predicting exposure to political cartoons. The notion of Political Activity serves the interests of this study because some of the most powerful people in our society are politicians. Any findings suggesting that political cartoons have their greatest impact among these politically active people would provide new insight to existing research, which has lowered estimates of the impact of political cartoons. If the approach of this study is useful. future studies gauging the importance of political cartoons should take into account their varying impact on different populations.18 Specifically. it is assumed that: Political Activity is related positively to Exposure to political cartoons. Politically active people want information about politics. and political cartoons are such a source. People may also go to them for amusement or to find out what other people may be thinking about an issue dear to them. Research already has shown a strong correlation between political activity and media exposure. Gary Kebbel has shown that political activity is the best predictor of 9 newspaper usage. outweighing more traditionally cited factors such as age. income and education.19 Alexis S. Tan has found that political participation is the strongest predictor of using media to get public affairs information.2° Tan also found that interpersonal discussion of politics led to newspaper use.21 Since politically active people are reading newspapers. it is likely they also are looking at political cartoons. Similarly. Exposure to cartoons is logically related to Perceived Understanding of political cartoons. There has been little or no research of how political involvement. through exposure to politics and political cartoons. affects perceived understanding of political cartoons. Measuring this is important because of Carl’s findings that only a minority of general populations understand artist-defined meanings of political cartoons. The logic here is that the more people read political cartoons. the more they will think they understand them. In addition. those who frequently read cartoons will be more familiar with the methods and political symbols cartoonists use. Oliphant. for example. often drew Sen. Edward Kennedy with a fish in his pocket during his 1980 primary challenge.22 The fish. a reference to Chappaquiddick. was a symbol many people might not have understood. Perceived Understanding should also have a reciprocal. positive relationship to Exposure. The more people believe they understand political cartoons. the more likely they are to increase their exposure to them. The case for this is 10 best stated in the negative: people will not frequently read something they do not understand. This is why children usually read the simplistic Garfield comic strip and skip the politically charged Doonesbuny. Exposure should also be related positively to cartoon Impact. Frequent exposure to political cartoons must be a good predictor of the effect of political cartoons in terms of opinions. behavior or emotional response. Stated conversely. people who do not read cartoons are very unlikely to be influenced by them. Finally. Perceived Understanding should also be positively related to Impact. People who draw some meaning out of political cartoons are more likely to be influenced than those who are dumbfounded by them Just as literate people are more likely to use books than illiterate peOple. Hypotheses Model + Political;—s A; Exposure %* ; Impact Activity Perceived Understanding 11 Statement of Hypotheses The above assumptions may be restated as the following five hypotheses. which are illustrated in the Hypotheses Model above: Hypothesis 1: Political cartoons have a different impact on different population groups. therefore: Hypothesis 2: Political Activity (independent variable) is positively related to Exposure to political cartoons (dependent variable). Hypothesis 3: Exposure. in turn. is also an independent variable positively related to the dependent variable Perceived Understanding. Hypothesis 4: Exposure (independent variable) is positively related to Impact (dependent variable). Hypothesis 5: Perceived Understanding (independent variable) is positively related to Impact (dependent variable). METHOD A one-shot survey was given to two different populations chosen to insure a high contrast in political involvement. On the high end. the survey was mailed to 108 of Michigan’s 110 House representatives. Two members whose districts include Monroe County were left out of the survey because. as a political reporter for The.Mbnroe Eyening.News. the researcher feared questioning them may appear to be a conflict of interest. The other half of the Legislature. the 38-member Senate. was not surveyed. However. some members responded to test surveys by phone. A phone survey of a more general population was taken to contrast the results with the Legislative sample. 23 It was done by phone because of likelihood that many people would not return a mail survey. A random sample of Eyening was subscribers who may be reached from the City of Monroe by a local phone call was chosen. The.EVening was is a daily and Sunday paper with a circulation of about 23.000. About 16,000 of those subscribers may be reached by a local call. Metropolitan Monroe has about 50.000 people. Also in the local-call zone are rural areas and the small surrounding towns of Carleton. Maybee and Newport. 12 13 As mentioned above. studies have shown there is a high correlation between newspaper readership and political activity. So it is probable the subscribers are more politically active than the general population. yet obviously much less active than full-time state representatives. Surveying subscribers provided the advantage of having a population that at least had a frequent opportunity to look at political cartoons. In addition to contrasting the Monroe and Legislature samples. the survey was designed to test the predicted positive variable correlations within each group. The same survey questions were used to measure all variables for both groups except for Political Activity. That was because Legislators have high political activity ratings by definition and therefore could be contrasted as a group with the less active Monroe sample. However. questions to measure political activity were asked of both groups to find finer gradations within each of them. Each group was asked questions suitable for its activity level. Concept Definitions The following definitions guided data collection for this survey: Political cartoon: A cartoon or caricature. usually with a caption. that has political content and appears on the editorial pages of a newspaper or news magazine. l4 Exposure to political cartoons: A measure of how often and with what level of interest people expose themselves to cartoons. Perceived Understanding: The extent to which a person thinks he or she understands political cartoons. Political cartoon impact: A measurement of political cartoons’ influence on emotions. behavior and thought. Political Activity: The degree to which a person participates in the political process. These definitions were operationalized into survey questions designed to test the hypotheses. Neither sample was asked unnecessary questions. People who said they had no Exposure to political cartoons were not asked any of the Perceived Understanding or Impact questions. The Perceived Understanding questions simply did not apply to them. and the Impact questions were automatically scored zero. This was based on the logic that people cannot be affected by something they are not exposed to. The Monroe survey had a total of 28 questions compared to 19 for the Legislature. They had 17 identical questions. The Monroe survey’s nine extra questions were for age. income. education (a matter of record for Legislators) and Political Activity (Legislators were asked two Political Activity questions. and subscribers were asked eight). Legislators were also given the option to comment on political cartoons at the end of the survey. 15 The variable measures incorporated into the survey are included in Appendix A. Choosing the Monroe Sample The Eyening was supplied a computerized list of all of its subscribers along with their addresses and phone numbers. Subscribers were grouped by newspaper delivery routes. In all. the Evening was had 256 routes with some or all subscribers in the local-call area. In instances where local and long distance numbers were mixed in a single route. the long distance subscribers were crossed off and the local-call portion was used. Business subscribers were also crossed off.24 Because people with high levels of income and education tend to live in similar neighborhoods. care was taken not to overrepresent any type of neighborhood in the sample. Small routes were lumped together by neighborhood similarities. Subdivisions. mobile homes. nursing homes. and older residential areas were combined into units consisting of one or several routes. The large rural routes. which consisted of homes sharing miles of rural roads. were mostly left intact. Selection of the random sample was based on these neighborhood units.25 The random selection method yielded a total sample of 148. Respondents from each neighborhood were selected by consulting a table of random numbers from 1 to 100 plus the role of a die. If. for example. the random number was 76. 16 that meant the 76th. 176th. 276th. or 376th person on a subscriber list would be selected. When a neighborhood had more that 100 subscribers. a roll of a die determined which of the above numbers would be used. The Monroe phone survey was conducted by two callers from September 18 to October 5. 1994. A total of 140 surveys were completed. and 210 potential respondents rejected the survey.26 Meanwhile. the Legislature surveys were sent to the House members’ home addresses to lessen the possibility that aides would fill them out. A couple of surveys had to be sent to legislative offices in Lansing because home addresses were not available.27 Sixty-six surveys. or 61 percent. were returned. A few were not complete. but all completed information was used whenever possible. Validity Testing Validity testing was done on variables operationalized through multiple questions or criteria. The test was done to see whether a question failed to measure the variable it was intended to measure. Questions were dropped if two outcomes occurred. A question was dropped if there was no correlation (r a 0 to .1) between it and the other questions designed to measure the same variable. On the other hand. a perfect correlation (r a .9 to 1) would have indicated that two questions are duplicate and one may be discarded. This test was applied to all variables measured by more than one question -- Perceived Understanding. Exposure. Impact and 17 Political Activity. In the end. these questions were dropped: an Exposure question (How many times each week do you read a newspaper or news magazine?). a Perceived Understanding question (How often do you check with others about the meaning of an editorial page cartoon?) and party Rank. one of the five criteria for measuring Political Activity for House members (See Appendix B). RESULTS This section begins with a look at answers to individual questions before examining tabulated variables and correlations and discussing each hypothesis. Average Results for Individual Questions The Monroe sample had a total of 140 respondents. and 66 of the 108 state House members surveyed responded. The number of respondents reported in the tables below varies because some surveys were not completely filled out. Except where otherwise indicated. survey answers were based on the following scale: almost always frequently sometimes infrequently almost never AAAAA O H N (.0 I5 vvvvv Exposure to Cartoons The two questions measuring Exposure revealed that Legislators "frequently" read editorial cartoons while the Monroe sample read them "sometimes." Also. Legislators put them higher on their priority list. 18 19 Table 1 - Exposure When you read newspapers or news magazines, how Often do you look at the cartoons that appear on the editorial pages? Legislators: 3.17 n=66 Monroe: 2.01 n=l40 When you turn to the editorial pages, when do you look at the editorial cartoons? Are they: (4) the first thing you look at (3) the second thing (2) about middle (1) second to last (0) the last think you look at Legislators. 2.83 n=65 Monroe: 2.3] n=l35 Perceived Understanding of Cartoons The question designed to measure Perceived Understanding showed both sample groups felt they understood political cartoons most of the time. although Legislators were the more confident. Table 2 - Perceived Understanding About what percentage of the time do you feel you understand the meaning of editorial page cartoons? Legislators: 92.3 percent n=55 Monroe: 80.] percent n=101* *The number of respondents to this question was lower because it did not apply to those who said they "almost never" looked at editorial page cartoons. 20 Impact of Cartoons Legislators were also the more likely to have their behavior affected by exposure to editorial cartoons. Table 3 - Behavioral Impact About how many times in the last two years have you clipped out an editorial page cartoon for display? Legislators: 5.09 times n=64 Monroe: 1.74 times n=140 About how many times in the last two years have you used an editorial page cartoon to help illustrate a point? Legislators: 2.53 times n=63 -‘ Monroe: 1.37 times n=139 How often do you talk to someone about an editorial page cartoon you have read? Legislators: 1.51 n=65 Monroe: 1.01 n=l40 How often do editorial cartoons spark a discussion between you and others? Legislators: 1.43 n=65 'Monroe: 0.96 n=l40 21 The results for every Emotional Impact question showed that Legislators had stronger emotional reactions to editorial page cartoons. Overall. cartoons were more likely to amuse. satisfy or draw an indifferent reaction than anger. sadden or disgust. Table 4 - Emotional Impact How often do editorial page cartoons cause you to feel the following emotions: Anger? Legislators: 1. l7 n=64 Monroe: 0.7l n= 140 Amusement? Legislators: 2.40 n=65 Monroe: 199 n= 140 A lack of concern or interest?* Legislators: 2.26 n=65 Monroe: 186 n= 1 4O Disgust? Legislators: 1.09 n=65 Monroe: 0.92 n= l 40 Satisfaction? Legislators: 1.89 n=65 Monroe: 1.28 n= 140 Sadness? Legislators: 0.92 n=65 Monroe: 0.64 n= 1 40 *F or this question, the point scale was reversed so that an "almost always" answer scored zero points and an "almost never" scored four. So the Legislators' higher average score means they were less likely than the Monroe sample to respond to cartoons with a lack of concern or interest. 22 Legislators were more likely to have their thoughts affected by editorial cartoons. although both groups said they were very unlikely to have their thinking affected by a cartoon. Both were more likely to have their beliefs reinforced. Table 5 - Thought Impact How often do editorial page cartoons influence your thinking on an issue? Legislators: 0.57 n=65 Monroe: 068 n=l40 How often do editorial page cartoons reinforce your beliefs on an issue? Legislators: 1.35 n=65 Monroe: 1.03 n=l40 23 Political Activity of Survey Groups Monroe residents scored an average of 2.89 on a zero-to- eight scale measuring political activity. Respondents were give a point for each "yes" response to the questions in Table 6. Table 6 - Political Activity for Monroe 1. Did you happen to vote in the last presidential election? Yes: 83% n=140 2. Did you happen to vote in the last city, village, or township election or the last school district election? Yes: 69% n=140 3. Have you ever served on a local government committee, board or commission? Yes: 6% n=140 4. Have you ever been elected to a local govemment office? Yes: 2% n=l40 5. Have you worn a campaign button or displayed a campaign poster in the last two years? Yes: 41% n=l40 6. Have you written to a newspaper editor or to any govemment representative about a political issue in the last two years? Yes: 31% n=140 7. Have you contributed money to a political cause, candidate, or political party in the last two years? Yes: 29% n=l39 8. Have you ever done any work for a political party or candidate? Yes: 27% n=l39 24 The political activity measure for Legislators yielded the averages in Table 7. About how many times in the year have you given a speech outside the legislative chamber on a political issue? Average: 35.9 n=62 About how many times in the last year have you written an article for publication on a political issue, including letters to editors? Average: 14.07 n=62 Average number of bills introduced: 15.02 n=66 Average committee scores (2 points for each committee orjoint House-Senate committee chaired, and 1 point for each one vice chaired. The highest total for any representative was 4 point): 1.92 n=66 Other Variables In terms of demographics. Legislators were more likely to be male and were. on average. wealthier. better educated and younger than the Monroe sample. Although Legislators were not asked their income. but their $47,723 base salary alone put them above the average for the Monroe sample. Table 8 - Demographics Education (1) Grade School (2) Some High School (2) High School Graduate (4) Some College (5) College Graduate (6) Some Graduate Work (7) Advanced Degree Averages Legislators: 5.76 n=63 Monroe: 3.90 n=138 Income (1) $0 - $10,000 (2) $10,000 - $20,000 (3) $20,000 - $30,000 (4) $30,000 - $40,000 (5) $40,000 - $50,000 (6) $50,000 - $60,000 (7) more than $60,000 Legislators: Not available Monroe: 4.02 n=109 Sex Legislators: 53 males (80.4 percent) 13 females (19.6 percent) Monroe: 47 males (40.7 percent) 83 females (59.3 percent) Age Legislators' average: 48. 5 n=58 Monroe average: 51.7 n=132 26 Variable Construction The variable Perceived Understanding was measured by a single question. Obviously this was also true for the variables age. income. sex and education. For the other variables. the question results had to be combined to calculate overall variable scores. Respondents answered most questions on 0-to-4 interval scales. In some instances. however. raw data had to be broken down to a 0-to-4 scale. The variables were calculated as follows: Exposure: The two questions Exp2 (When you read newspapers or news magazines. how often do you look at the cartoons that appear on the editorial pages?) and Exp3 (When you turn to the editorial pages. when do you look at the editorial cartoons?) were both based on 0-to-4 interval scales. They were simply averaged. Those who reported on Exp2 that they "almost never" looked at political cartoons were automatically given a "0" on Exp 3. Impact: The total impact score was the average of three scores: behavioral. emotional and thought impacts. The six emotional impact questions measured the following possible reactions to cartoons: anger. amusement. a lack of concern or interest. disgust. satisfaction and sadness. Answers were based on 0— to—4 intervals and were simply averaged. The same was true 27 for the two thought impact questions (How often do editorial cartoons influence your thinking on an issue?. How often do editorial cartoons reinforce your beliefs on an issue?). The behavioral impact score was based on four questions. Two of them (How often do you talk to someone about an editorial page cartoon you have read?; How often do editorial cartoons spark a discussion between you and others) were answered on 0-to-4 interval scales. The other two behavioral impact questions were answered with raw scores. These raw- scores were translated to 0-to-4 point scales. and all four questions were averaged for the final score for behavioral impact. In all cases where raw scores needed to be translated to 0-to-4 point scales. the same principles were used to construct the scales. Namely. no points were given for no activity. one point for an average or slightly below average score. and an additional point up to 4 total points was awarded for each half standard deviation above the average. Accordingly. the raw answers to question Bi (About how many times in the last two years have you clipped out an editorial page cartoon for display?) were scored according to the following scale: 0 points = 0 cartoons clipped l-l-3 2:4-9 3=10~16 4:17 or more For question BZ (About how many times in the last two years have you used an editorial page cartoon to help illustrate a point?). the scale used was: 28 0 points = 0 times cartoons used to illustrate a point 1:1-2 2:3-6 3=7-10 4:11 or more Political Activity: Scoring Political Activity was simple for the Monroe sample. Respondents were simply given a point for each of the eight questions to which they answered "yes." Those eight questions are listed in Appendix A. The Political Activity measure for the Legislative sample was based on four criteria: bills. articles. committees and speeches. All were expressed in 0-to-4 measurement scales. The scales. made according to the method outlined above, were: Bills: The average representative introduced 15 bills. The scale used was: 0 points = 10 or fewer bills 1211-15 2:16-20 3:21-25 4:26 or more Speeches: The average was 35 speeches. The scale: 0 points - 18 or fewer speeches 1:19-35 2:36-52 3:53-71 4-72 or more Articles: The average was 14. 0 points = 4 or fewer articles 1=5-14 2:15-24 3:25-34 4:35 or more 29 The fourth Political Activity measure. committees. did not require the construction of a scale. Legislators were simply given two points for each standing committee or Joint House~Senate committee they chaired and one point for each one they vice-chaired. The highest total for any legislator was 4 points. In most years. this method would have awarded the most points to the party in power. since committee chairs traditionally go to the controlling party. However. the year this survey was taken the House Republicans and Democrats split the House 55-55 and had reached a power sharing agreement. Each standing committee had Democratic and Republican co-chairs. So this measure was considered a legitimate indicator of political activity. not membership in the maJority party. Hypotheses Examined The results supported all of the hypotheses to varying degrees. Each hypothesis and the relevant supporting data are discussed below. Hypothesis 1: Political cartoons have a different impact on different population groups. The average scores for each variable (see Table 9) clearly show that the Legislature and Monroe subscribers are not identical population groups when it comes to political cartoons. The Legislature sample had higher average scores for Exposure. Perceived Understanding and Impact. 30 Table 9 - Mean Scores Significance of Difference as measured by a Variable Legislature Monroe Two-Tailed T-Test Exposure 2.53 1.80 *3.55 Perceived Understanding 92. 8 80.1 *4. 53 Impact Total 1.28 .94 ”2.97 Behavior Imgict 1.29 .73 *428 Emotional Impact 1.62 1.23 * *3 .23 Thought Impact .96 .85 .77 * Significant at the .001 level (T=3.29 or higher) ”Significant at the .01 level (T=2.58 or higher) The null hypothesis that the two groups are the same can be reJected at the .001 level of confidence when comparing the average scores for Exposure and Perceived Understanding. It can be reJected at the .01 level concerning political cartoon Impact. The Legislators also scored higher on all three subcategories of total Impact -- behavioral. emotional and thought. However. there is no statistical certainty that there is any true difference between the two groups on thought impact. Hypothesis 2: Political Activity is positively related to Exposure to political cartoons. 31 This hypothesis is supported by the data in Table 9. which shows that the most politically active group. the Legislative sample. had a higher Exposure rating than Monroe subscribers. The average Legislator reported "frequently" looking at political cartoons. while the average Monroe subscriber "sometimes" read them. Within the two groups. the hypothesis is partly supported. A positive correlation exists within the Monroe sample (see Table 10). where Political Activity had an r-.41 correlation with Exposure. That this positive relationship exists is certain at the .01 level. Table 10 - Comparison of Correlations ~Legislature Monroe Political Activity/ Exposure .04 *.41 Education/Exposure . 16 *35 Sex/Exposure (favoring males) . I4 *22 Age/Exposure .02 -. 12 Income/Exposure Not . 13 Available Exposure/ Perceived Understanding .30 *.46 Exposure/Impact Total .72 *.77 Exposure/Behavioral Impact .66 *65 Exposure/Emotional Impact .67 * .76 Exposure/Thought Impact .55 *.59 Perceived Understanding/ Impact .00 *28 *Significant at the .01 level. ~State House census data. Significance tests were not rwufimuy 32 However. something else appears to be coming into play within the Legislature sample (Table 10). Here. there was virtually no correlation (r--.04) between Political Activity and Exposure. It would appear. then. the variable Political Activity reaches a sort of plateau once high levels of activity are reached. In other words. it is no longer useful in predicting political cartoon exposure within the confines of a group where all persons have extremely high levels of political activity. For the Monroe sample. Political Activity was a strong predictor of Exposure compared to the other variables measured (Table 10). It was stronger than Education (r=.35). Sex (r=.22. favoring males). Age (rs-.12) and Income (r-.13). Once again. however. things were different for the Legislature sample. where none of the other variables proved to be particularly strong predictors of Exposure (Income for Legislators was not measured). It should be noted. though. that Education had the biggest impact on Exposure for Legislators with a .16 positive correlation compared to .04 for Political Activity. It appears that Education is a useful predictor of Exposure. But. like Political Activity. the Legislative sample suggests the predictive power of the Education variable reaches a plateau and flattens out among groups with fairly uniform. high education standards -- as the state House members had. Legislators’ education scores averaged 5.76 (6=some graduate work) compared to the Monroe sample’s 3.90 (4asome college. but no degree). Both samples had a standard deviation on Education of about 1.3. 33 Hypothesis 3: Exposure. as an independent variable. is positively related to the dependent variable Perceived Understanding. Both samples supported this hypothesis. The correlation was .30 for the Legislative sample and .46 for Monroe subscribers (see Table 10). It can be certain that Exposure is acting as an independent variable simply because Perceived Understanding cannot exist without Exposure. Hypothesis 4: Exposure is positively related to Impact. This was supported by the evidence. There was an r-.72 correlation between Exposure and Impact in the Legislative sample and a r=.77 correlation for the Monroe sample. In other words. those who looked at political cartoons most frequently were those who most often clipped them. talked about them. and reacted to them with the strongest emotions. Hypothesis 5: Perceived Understanding is positively related to Impact. This was supported in the Monroe sample. but not in the Legislature sample. Although the correlation was not large in the Monroe sample (r=.28). there is certainty at the .01 level that the positive correlation exists. For the Legislative sample. however. there was no correlation between these variables (r=0). That may be attributable in part to the small variance in Perceived Understanding in the Legislature. The standard deviation from the Legislature’s 92.8 percent average was 8.3 percentage points. This gives a standard deviation range of between 84.5 and 100 percent. By comparison. the standard deviation from the Monroe sample’s 80.7 percent average was 19.8 percentage points. 34 yielding a much wider standard deviation range of 60 to 100 percent. The lack of a positive correlation between Perceived Understanding and Impact for the Legislature may be attributable to the plateau effect noted above. In other words. Perceived Understanding loses its predictive power among a group that has achieved a high. uniform level of Perceived Understanding. DISCUSSION The aim of this study was to shed more light on the Impact of political cartoons by seeing if an influential group was more affected by them than a less influential group. What the results suggest is that political cartoons have more impact on politically active. educated people than on the less active and less educated. This study shows that the more likely people are to do things such as voting or writing a letter to the editor. the more likely they are to look at political cartoons. And those who read them most often are more likely to say they are amused. angered or saddened by political cartoons. They are more likely to clip them. display them or talk about them. They are more likely to think cartoons reinforce their beliefs. It should be noted. however. that this study does little to resurrect any Thomas Nast-like notion that political cartoons are perhaps the most potent shapers of political opinion among the masses. But perhaps it does require an adJustment of Carl’s suggestion that "maybe editorial cartoons are not supposed to be understood by many people much of the time." 28 Perhaps that should be amended to "maybe editorial cartoons are not supposed to be understood by those who almost never read them or those who. 35 36 Judging from their political inactivity. have no interest in politics." Those who at least occasionally read political cartoons certainly think they understand them -- whether or not the meaning they derive would match what the cartoonist intended. Fifty-five of 65 state House respondents said they read political cartoons at least sometimes and that they understood political cartoons an average of 92 percent of the time. For the Monroe subscribers. 101 of 140 exposed themselves to cartoons. These said they understood them an average of 81 percent of the time. Clearly. people who expose themselves to political cartoons at any level think they are getting something out of them. The question remains, what are they getting? While readers are responding emotionally and behaviorally. it appears that, like other forms of satire. political cartoons hit people more in the belly than the brain. Asked how often political cartoons influence their thinking. the average state House member replied somewhere between "infrequently" and "almost never." Monroe subscribers scored nearly the same. although they were slightly more inclined to say "infrequently.” State House members said political cartoons reinforced their beliefs somewhere between "sometimes" and "infrequently" -- compared to a solid "infrequently" for the Monroe subscribers. On the other hand. those with the highest exposure to political cartoons were likely to say their thoughts were influenced the most. The correlation for Exposure and 37 thought Impact was r=.55 for the Legislative sample. and r-.59 for the Monroe sample. Nevertheless. even those with the highest exposure were not saying their thinking on an issue was often being changed or even reinforced by political cartoons. So. even if political cartoons are having their biggest impact among the most politically active and the educated. what difference do political cartoons really make? The answer to that should be pursued in studies that address political cartoons as the unique art form they are. As Brinkman found. editorial cartoons are outperformed by written editorials in forming opinions. But cartoons are not essays. and it may be that their impact has many facets that essays do not. Perhaps it could be shown that written editorials and other communication forms fail to do as well as political cartoons in accomplishing whatever it is political cartoons are meant to do. Political cartoonists themselves said their top obJectives are "to stimulate thought. to challenge and provoke." "to expose government wrongdoing. to criticize." "to present informed opinion" and "to entertain." 29 Perhaps another way to get at the impact of political cartoons is to study how well people who read them most think they meet such objectives. At any rate. this study strongly indicates that political cartoons are well read. especially among more politically active and educated groups. Further. the overwhelming maJority of these peOple feel they understand what they are reading. This only makes sense. If they didn’t "get" at least some meaning from editorial cartoons. why look 38 at them so often? Furthermore. editorial cartoons are having their biggest behavioral impact and are drawing the strongest emotional reactions from politically active. educated groups. What final difference political cartoons make may remain mysterious. Certainly. many people are ready to hazard a guess. such as one state House member who wrote. "cartoons degrade America and its representatives -- present our youth with the wrong impression of our American political system" Editorial cartoons also seemed to make a difference to another House member who simply wrote. "I like ’eml" Comments like these may not prove anything about editorial cartoons. but they do illustrate how many politically active tend to respond to editorial cartoons. The fact is that people who are actively making a difference in our society both read and react to editorial cartoons as if they were something important; it is for that reason. if for no other. that they are . NOTES 1. The Encyclopedia Americana. International ed.. s.v.. "Nast. Thomas." 2. John A. Kouwenhoven. Adventures of America 1857-1900: A Pictorial Record from Harper’s Whekly. (New York: Harper and Brothers Publishers. 1938). 138. 3. Stephen Hess and Milton Kaplan. The Uhgentlemanly Art: A History of American Political cartoons. (New York: The Macmillan Company. 1968). 58. 4. Ibid., 73-90. 5. Ibid.. 100. 6. Charles R. Gruner. "Ad Homonym Satire as a Persuader: an Experiment." Journalism Quarterly. 48 (1971): 128-131;. "A Further Experimental Study of Satire as Persuasion." Speech .Mbnographs. 33 (June 1966): 184-85:. "An Experimental Study of Satire as Persuasion." Speech.Mbnographs. 32 (June 1965): 149-153; "Editorial Satire as Persuasion: An Experiment." Journalism Quarterly. 44 (1967): 727-30. 7. Del Brinkman. "Do Editorial Cartoons and Editorials Change Opinions?" Journalism Quarterly. 44 (1968): 724-26: Leroy M. Carl. "Editorial Cartoons Fail to Reach Many Readers." Journalism Quarterly. 44 (1968): 533-35. and. also by Carl. "Political Cartoons: 'Ink Blots’ of the Editorial Page." Journal of.Popular culture. 4 (1970): 39—45. 8. Lawrence H. Streicher. "On a Theory of Political Caricature." comparative Studies in Society'and.Hlstory. 9(4) (July 1967): 427-45. 9. Martin J. Medhurst and Michael A. Desousa. "Political Cartoons as Rhetorical Form: A Taxonomy of Graphic Discourse." Cbmmunication.Mbnographs. 48 (1981):197-236. 10. David N. Ammons. John C. King. and Jerry L. Yeric. "Unapproved Imagemakers: Political Cartoonists’ Topic Selection. ObJectives and Perceived Restrictions." waspaper Research Journal. 9(3) (Spring 1988): 79-80; James J. Best. "Editorial Cartoonists: A Descriptive Survey." waspaper Research Journal. 72(2) (Winter 1986): 29-37: James R. Beniger. "Does Television Enhance the Shared Symbolic 39 40 Environment? Trends in Labeling of Editorial Cartoons." 1948- 1980." American Sociological.Review. 48 (February 1983)): 103-11: Thomas H. Bivins. "Format Preferences in Editorial Cartooning." JDurnalism Quarterly. 44 (1968): 724-26: Matthew C. Morrison. "The Role of the Political Cartoonist in Image Making." Cbntral States Speech JOurnal. 20 (1969): 252-60: Daniel Riffe. Donald Sneed. and Roger L. Van Ommeren. "Behind the Editorial Page Cartoon." JOurnalism Quarterly. 62 (2) (Summer 1985): 378-83. 450: Riffe and others. "Deciding the Limits of Taste in Editorial Cartooning." JOurnalism Quarterly. 64(3). (Summer-Autumn 1987): 607-10: Riffe and others. "Ethics in Editorial Cartooning: Cartoonist and Editor Views." waspaper.Research JOurnal. 9(3) (Spring 1988): 91-103. 11. Brinkman. "Do Editorial Cartoons and Editorials Change Opinions?" 724-726. 12. Carl. "Editorial Cartoons Fail to Reach Many Readers." 533-35. 13. Gruner. "An Experimental Study of Satire as Persuasion." 151: "Editorial Satire as Persuasion." 729-730. 14. In addition to the studies cited under note 7 above. there have been numerous political cartoon content studies or other studies focused on special interest areas. Some examples are: Roy E. Blackwood. "Ronbo and the Peanut Farmer in Canadian Editorial Cartoons." J0urnalism Quarterly. 62(2) (Summer 1989): 453-57: William A. Gameson and David Stuart. "Media Discourse as a Symbolic Contest: The Bomb in Political Cartoons." Sociological Fbrum. 7(1) (1992): 55ft; Ray Morris. "Cultural Analysis through somiotics: Len Norris’ cartoons on official bilingualism." The Cbnadian Review of Sociology and Anthropology. 28 (May 1991): 225-54: William H. Wiggins Jr.. "Boxing’s Sambo Twins: Racial Stereotypes in Jack Johnson and Joe Louis Newspaper Cartoons. 1908 to 1938." Journal of Sport History. 15 (3) (Winter 1988): 242 ff. 15. Gameson and Stuart. "Media Discourse." 62. 16. Wiggins. "Boxing’s Sambo Twins." 242-43. 17. Oliver Harrington. "The Sting of the Image." a lecture given along with the The Detroit.News political cartoonist Draper Hill at Michigan State University on March 17. 1994. 18. Carl. in "Editorial Cartoons Fail...." already has found that more people in a university town understand political cartoons than those in two average small towns. although a large maJority in both failed to guess what the artists' said their intended meanings were. 41 19. Gary Kebbel. "Strength of Political Activity in Predicting Newspaper Use," Newspaper.Research JOurnal. 6(2) (Winter 1985) 1-7. 20. Alexis 5. Tan. "Political Participation. Diffuse Support and Perceptions of Political Efficacy as Predictors of Mass Media Use." Cbmmunication.Mbnographs. 48 (June 1981): 133-146. 21. Alexis S. Tan. "Mass Media Use. Issue Knowledge and Political Improvement.".Public Opinion Quarterly. 44(2) (Summer 1980): 241-248. 22. Medhurst. "Rhetorical Form." 223-24. 23. The state House members and the Monroe residents are each exposed to different newspapers serving their local circulation areas. So the two groups are probably exposed to the work of different cartoonists. They have in common exposure to cartoons in national publications. such as Time or Newsweek magazines. and the state’s largest papers. the Detroit Free Press and The Detroit News. In addition to these. Legislators are likely to be exposed to The Lansing State Journal and various newspapers from their local districts. In Monroe. residents are likely to be exposed to political cartoons in The.Mbnroe Eyening News. The Evening News features political cartoons from the Newspaper Enterprise Association. which includes cartoonists Jim Berry of "Berry’s World" and Jeff Stahler. and Copley News Service. Copley’s roster of cartoonists includes Steve Kelley. Marshall Ramsey. Mike Thompson and others. In addition. the Evening News prints one or two editorial page cartoons by local artist Dave Riddle. He also does a weekly single-panel cartoon for the Living Page. 24. Although the average number of subscribers per route was 62. the size widely varied between newspaper walk routes and rural motor routes. The smallest routes had only a few subscribers and the largest had 490. 25. Neighborhoods with fewer than 53 subscribers were not surveyed. Only one neighborhood fell into this category. One respondent was selected from neighborhoods with 54 to 105 people; two from 106 to 160: three from 161 to 168: four from 269 to 376. and five from 377 or more. The proportions were based on a goal to draw a sample of about 150 subscribers. thought to be the number need to minimize sample error. A greater degree of sampling error on the Monroe population was allowable because the comparison group. being a census. should have no sampling error. 26. Each randomly selected phone number was written at the top of one of 148 survey blanks. The randomly selected number was followed by six alternate numbers. The six alternates were the numbers listed immediately after the 42 randomly selected number on The Monroe Evening News’ subscriber list. A roll of a die determined whether each survey was designated for the man or woman of each subscribing household. Callers asked for the man or the woman of a subscribing household. A respondent of the opposite sex was accepted when respondents said that there was no man or no woman of the house. Calling times alternated between daytime and evenings. weekdays and weekends. Callers made efforts to contact the first subscriber listed on each survey. including offering to call back at a more convenient time. Callers tried alternate phone numbers only after being refused. reaching a disconnected number or repeatedly failing to find someone home for a period of several days. A survey was considered "dead" if all seven of the alternative numbers failed. Of the 148 surveys. 140 were completed and 8 failed to get a response. The locations of the eight dead surveys were checked to make sure they did not skew the sample. It was found the eight did not represent any single neighborhood or type of neighborhood. In all. it took a total of 350 calls to get the 140 survey responses. On average. it took 2.5 calls to successfully complete a survey. 27. In August 1994. 108 of the 110 House members were sent an advance letter briefly stating the purpose of the Legislature survey. The surveys were mailed about a week later. About three weeks later. a second wave of surveys were sent to those who did not respond to the first. 28. Carl. "Political Cartoons: ’Ink Blots' of the Editorial Page," 44. 29. Ammons and others. "Cartoonists’ Topic Selection. ObJectives and Perceived Restrictions." 83-86. APPEND I CES Appendix A -- Variable Measures Political Activity: For the Legislature. this was based on the number of bills each legislator had introduced as of Aug. 1. 1994 (a few months before the end of the two-year session). committees chaired or vice-chaired. party rank. and the number of articles and speeches they said they had done in the last year. Bills. committees and rank are a matter of record. The other two ratings were based on the following survey questions: About how many times in the last year have you given a speech outside the legislative chamber on a political issue? About how many times in the last year have you written an article for publication on a legislative issue. including letters to editors? Political Activity for the Monroe sample was measured with eight yes-or-no questions. Respondents were given a point for every "yes" answer to these questions: 1. Did you happen to vote in the last presidential election? 2. Did you happen to vote in the last city. village. or township election or the last school district election? 3. Have you ever served on a local government committee. board or commission? 4. Have you ever been elected to a local government office? 5. Have you worn a campaign button or displayed a campaign poster in the last two years? 6. Have you written to a newspaper editor or to any government representative about a political issue in the last two years? 7. Have you contributed money to a political cause. candidate. or political party in the last two years? 8. Have you ever done any work for a political party or candidate? 43 44 Exposure: The survey had three exposure questions: Exp 1. How many times each week do you read a newspaper or a news magazine? Once a week or less 2 to 3 times a week 4 to 6 times a week once a day two times a day more than two times a day AAAAAA VVVVVV Exp 2. When you read newspapers or news magazines. how often do you look at the cartoons that appear on the editorial pages? ( ) almost always ( ) frequently ( ) sometimes ( ) infrequently ( ) almost never Exp 3. When you turn to the editorial pages. when do you look at the editorial cartoons? Are they: the first thing you look at the second thing about middle second to last the last thing you look at AAAAA VVVVV Perceived Understanding: Two questions were used to measure Perceived Understanding. PU2. however, was dropped as is explained in Appendix B. PU1. About what percentage of the time do you feel you understand the meaning of editorial page cartoons? PU2: How often do you check with others about the meaning of an editorial page cartoon? Impact: Impact was divided into three sections: behavioral impact. emotional impact and thought impact. While all the impact questions were to be combined into a single impact score. it was hoped the divisions might provide insight into what kind of impact political cartoons are most likely to have. 45 Impact - behavioral 81. About how many times in the last two years have you clipped out an editorial page cartoon for display? 82. About how many times in the last two years have you used an editorial page cartoon to help illustrate a point? B3. How often do you talk to someone about an editorial page cartoon you have read? ( ) almost always ( ) frequently ( ) sometimes ( ) infrequently ( ) almost never B4. How often do editorial cartoons spark a discussion between you and others? ( ) almost always ( ) frequently ( ) sometimes ( ) infrequently ( ) almost never Impact - emotional How often do editorial page cartoons cause you to feel the following emotions?: E1. Anger? E2. Amusement? E3. A lack of concern or interest? E4. Disgust? E5. Satisfaction? E6. Sadness? ( ) almost always ( ) frequently ( ) sometimes ( ) infrequently ( ) almost never Impact - thought T1. How often do editorial page cartoons influence your thinking on an issue? ( ) almost always ( ) frequently ( ) sometimes ( ) infrequently ( ) almost never T2. How often do editorial page cartoons reinforce your beliefs on an issue? ( ) almost always ( ) frequently ( ) sometimes ( ) infrequently ( ) almost never Other variables: Respondents were asked questions to determine their age. income. level of education and sex. Research has established that these variables have varying value as predictors of media usage. Measuring them may help reveal the relative 46 importance of political involvement in predicting Exposure to political cartoons. Monroe residents were asked their income and education according to the following scales: ( ) Grade school ( ) Some high school ( ) High school graduate ( ) Some college ( ) College graduate ( ) Some graduate work ( ) Advanced degree (1 ) $0-$10,000 (2 ) $10,000 to $20,000 (3 ) $20,000 to $30,000 (4 ) $30,000 to $40,000 (5 ) $40,000 to 550.000 (6 ) $50,000 to $60,000 ( ) more than $60,000 Callers also asked respondents their ages. and marked their sex. Legislators were not asked personal questions. Their age. sex, and education were retrieved from information on record. Their income was not requested because it was feared this would reduce the number of responses. It is clear, however, that state House members’ incomes are near the top of the tiers outlined above. Their state salaries plus expense perks put them over the $50,000 income range. Appendix B -- Validity Testing All the Impact questions passed the validity test described in the text. Refer to Appendix A for a complete list of survey questions. One of the two Perceived Understanding questions was dropped because the resulting scores had no correlation with one another. This was suspected as the survey was given to phone respondents. The least direct question. PU2 (How often do you check with other about the meaning of an editorial page cartoon?) was dropped. As a result, Perceived Understanding was measured solely by the more direct question, PUl (About what percentage of the time do you feel you understand the meaning of editorial page cartoons?). For Exposure. it was found that Exp1 (how many times each week do you read a newspaper or a news magazine?) was superfluous. In the state House results, correlation tests showed that Exp 1 had a .01 correlation with Exp3 (When you turn to the editorial pages, when do you look at the editorial cartoons?) and a weak correlation (r-.147) with Exp2 (When you read newspapers or news magazines. how often do you look at the cartoons that appear on the editorial pages?). For the local sample, it could not be ruled out at the .01 certainly level that Exp 1 had no correlation with either Exp2 or Exp3. It appeared from these tests that Exp1 was not helpful in measuring political cartoon exposure. Further testing showed that it made little difference if Exp1 results were used or not. Exposure calculated using all three questions had a near perfect correlation with Exposure calculated using only Exp2 and Exp3. The correlation for the House sample was r=.938. For the Monroe sample. it was r-.974. The third problem turned up by correlation testing was using party rank in measuring Political Activity for state House members. It turned out that Rank had a weak, negative correlation with all of the other four variables: -.18 with Bills: -.02 with committees: -.206 with Articles. and -.13 with Speeches. This was puzzling. and it was thought that the point system for party Rank might have been inaccurate. But there is little doubt as to the status of the Speaker and floor leaders. ReadJustments to the scoring for lower ranks were made. but the end results were again the same: Rank had little or no correlation with the other four indicators used. It was concluded that party Rank measures something other than political activity. and rank was dropped from the list of criteria. Once again. a test was run to see what difference deleting Rank made. It was found that it made virtually no difference. Exposure. using all five questions. 47 48 had a .95 correlation with Exposure measured by using all questions except Rank. Correlation tests also raised questions about two of the eight questions used to measure Political Activity for the Monroe Sample. Again. the questions were: 1. Did you happen to vote in the last presidential election? 2. Did you happen to vote in the last city, village. or township election or the last school district election? 3. Have you ever served on a local government committee, board or commission? 4. Have you ever been elected to a local government office? 5. Have you worn a campaign button or displayed a campaign poster in the last two years? 6. Have you written to a newspaper editor or to any government representative about a political issue in the last two years? 7. Have you contributed money to a political cause, candidate. or political party in the last two years? 8. Have you ever done any work for a political party or candidate? Correlation tests showed it was not certain at the .05 level that question 3 does not have a zero correlation with questions 1.2.5.7 and 8. The same was true of question 4 and its correlations with questions 1.2.7 and 8. At first. this evidence suggested that questions 3 and 4 do not measure political activity and should perhaps be thrown out. But it was concluded that questions 3 and 4 are valid. The Pearson’s R formula apparently was not useful in this instance because it was used on too small a sample. Only nine of 140 Monroe respondents answered "yes" to question three, meaning they had served on a government board or commission. Only three answered "yes" to question four, meaning they had been elected to a public office. Further investigation proved that these two questions were very reliable indicators of political activity. The Average score on a 1-to-8 scale for those who said "yes" to question 3 was 5.8 -- double the survey average score of 2.9. For the three office holders. the average political activity score was 7.78. Apparently, the problem was that so many people answered "no" to 3 and 4. While a "yes" answer proved to be a highly reliable indicator of political activity, the opposite was not true -- a "no" answer was no indication of low political activity. Appendix C -- Editorial Cartoon Sample A sample of the cartoons that appeared on the editorial pages of The Mbnroe Eyening News from Sept. 18 to Oct.5, 1994, when the Monroe phone survey was conducted. One or two cartoons appeared each day. Three were sometimes published on Sundays. 49 50 .‘l’l‘ a... (U. I I: O . 182.30 . Hmmm < mommzmz H29.52%me Sow: \ .mm_ :0 Eu; 20.530445 .m>szmwz_ oz< 02.502: 9 m: “zOF<>mmmmO (Em: "mm—(mu? zo_mmou. z. ,. sz cud-sz: z< . . 02:50 zo_._. zO . . 5.sz Imam mOmOmo .. ...__u_..o_>> pram 51 v .uoO $93923 \ ”WWW 546.2 .228 . . .336595 gamma 20P> 935m . .86 ES uncommon as S 22 So 0:26an 2 MES 63:6: mesa: E; .3me- 252, 9an 52 J ’ {'4 Maw/7' r/. /(/‘ ‘I / v . I 7’//7{/'/’/ . , , , ////" / ’/// W" / ., , WK /%0 ' '/ \\ h \\ ‘m . '2 19 Sept. 53 WRLD WPULATION CONFERENCE VENDORS... 20 Sept. ' C Q P . .o"\ e - ' ‘ . ',:. o- ‘ I, I, .'.' ..e.l ‘ . o .0 ' . '2'..'?e '. ‘ -- 1'. O 'u'..-‘ " ‘,"o‘ ~ 0' I .. I ,e ‘, . ... v. ."..':C f..‘ . e _.....-.'.~ e ' ’\' .0“ ’ '. . "o \ a "" ~C .. . ‘ ‘. N I. ' ' l . go... shamans. 89553 281 wmammm zmgszo 3.58? wwaE. ~52..sz z mazessoomm «85$ was, 3... as. SEE awn amiss 2 Eco; (H 9.4 ozmmaemgmz mm: magic 'l‘ 55 mm .uavm . .\\\. 32.3%.; hum 3.30.58 , . .\\\\\\ \\v\ gil‘!‘ .\\ v\\ \\\ §\\ \ a... .s NV #1 r .. “.85 mums as. was: as 23:. 302 it Marta _ 3 8m 1L 56 W Q. i] M [11:1 I Emil 57 In!!!” _ Z inc hzaozm 55. E mommzmo Mon mcz§mm mm o: .mEFHS m.>.5_oom no macaw amnazoousm 72m 20E 226496 5.5.. mm: “.6 2PM; r .55. mm c.2138 Ettozot E; v; . Ila Elimall /}/};w r. ' / r ; \NE 60 TOUR HOWARD... 11 SIN 1113 011i. MODE llNlE. GREAT ‘Nllllb‘i A1100 ~I o’.’ 0/ I Tl-llNKI Ck“... ITl-llNK 1 (All. as- ’ ' Iii?" ‘1?th .. .4— 317—“ [lg/MI! . ,l ‘ ‘1 [{I’Il'MII'I/Irlw 9!, (3:) /, WWW V l I! ll 1: .\ r 1 \ -'\ ‘.'\ Oct. 59 v . uOO , 2033on no Ham no. Eagwm me. 1111 ‘ .3 \ \ WQOAuquJuuozw Mg \. s. . 858024135 .. . zé I.” 93:13... . L . N. . 92mm? mm: ..... é .. rt 1 g dimozo 5914: 0.922 50 wzom . (V a HEE- 4w: K wins. / SOURCES CONSULTED Sources Consulted Ammons. David N.. John C. King. and Jerry L. Yeric. "Unapproved Imagemakers: Political Cartoonists’ Topic Selection, ObJectives and Perceived Restrictions." Amwspaper.Research Journal, 9(3) (Spring 1988): 79-80. Beniger. James R. "Does Television Enhance the Shared Symbolic Environment? Trends in Labeling of Editorial Cartoons. 1948-1980." American Sociological.Review, 48 (February 1983): 103-11. Best. James J., "Editorial Cartoonists: A Descriptive Survey." Newspaper.Research Journal, 72(2) (Winter 1986): 29-37. Bivins. Thomas H., "Format Preferences in Editorial Cartooning." Journalism Quarterly. 44 (1968): 724-26. Blackwood. Roy E., "Ronbo and the Peanut Farmer in Canadian Editorial Cartoons." Journalism Quarterly. 62(2) (Summer 1989): 453-57. Brinkman. Del. "Do Editorial Cartoons and Editorials Change Opinions?" JOurnalism Quarterly. 44 (1968): 724-26. Carl, Leroy M.. "Editorial Cartoons Fail to Reach Many Readers." Journalism Quarterly. 44 (1968): 533-35. "Political Cartoons: 'Ink Blots’ of the Editorial Page," Journal of Popular culture, 4 (1970): 39-45. Note: This is an essay by Carl rebutting the idea that cartoons are widely understood because pictures are easier to 'read’ than words. It discusses his previous research (cited above), but does not contain any new research. Gameson, William A.. and David Stuart, "Media Discourse as a Symbolic Contest: The Bomb in Political Cartoons," Sociological Fbrum. 7(1) (1992): 55 ff. Gruner, Charles H., "Ad Homonym Satire as a Persuader: an Experiment." Journalism Quarterly. 48 (1971): 128-131. 60 61 .. "A Further Experimental Study of Satire as Persuasion," Speech Monographs. 33 (June 1966): 184- 85. .. "An Experimental Study of Satire as Persuasion." Speech.Monographs, 32 (June 1965): 149- 153. ., "Editorial Satire as Persuasion: An Experiment," Jaurnalism Quarterly. 44 (1967): 727-30. Harrington. Oliver. "The Sting of the Image," lecture given along with The Detroit News political cartoonist Draper Hill at Michigan State University, March 17. 1994. Hess. Stephen and Milton Kaplan, The Uhgentlemanly Art: A .History of.Amer1can Political cartoons. New York: The Macmillan Company, 1968. Kebbel. Gary, "Strength of Political Activity in Predicting Newspaper Use." Nowspaper.Research Journal. 6(2) (Winter 1985): 1-7. Kouwenhoven, John A.. Adventures of.America 1857-1900: A P1ctorial.Record from Harper's woekly. New York: Harper and Brothers Publishers. 1938. Medhurst. Martin J., and Michael A. Desousa. "Political Cartoons as Rhetorical Form: A Taxonomy of Graphic Discourse." communication Monographs, 48 (1981):197- 236. Morris. Ray, "Cultural Analysis through Somiotics: Len Norris’ cartoons on official bilingualism," The conadian.Review of Sociology and Anthropolo&Y. 28 (May 1991): 225-54. Morrison. Matthew C.. "The Role of the Political Cartoonist in Image Making." central States Speech JOurnal, 20 (1969): 252-60. Riffe. Daniel, Donald Sneed. and Roger L. Van Ommeren. "Behind the Editorial Page Cartoon." Journalism Quarterly. 62(2) (Summer 1985): 378-83. 450. Riffe. Daniel, Donald Sneed. and Roger L. Van Ommeren. "Deciding the Limits of Taste in Editorial Cartooning." Journalism Quarterly. 64(3), (Summer-Autumn 1987): 607- 10. Riffe. Daniel, Donald Sneed. and Roger L. Van Ommeren. "Ethics in Editorial Cartooning: Cartoonist and Editor Views." Newspaper.Research JOurnal. 9(3) (Spring 1988): 91-103. 62 Streicher, Lawrence H., "On a Theory of Political Tan, Tan. Caricature," comparative Studies in Society and .History, 9(4) (July 1967): 427-45. Alexis S.. "Political Participation. Diffuse Support and Perceptions of Political Efficacy as Predictors of Mass Media Use," communication.Monographs. 48 (June 1981):133-145. Alexis S.. "Mass Media Use, Issue Knowledge and Political Improvement," Public Opinion Quarterly. 44(2) (Summer 1980):241-248. Wiggins, Jr.. William H., "Boxing's Sambo Twins: Racial Stereotypes in Jack Johnson and Joe Louis Newspaper Cartoons. 1908 to 1938," Journal of Sport.History, 15 (3) (Winter 1988): 242 ff. In... :23,an 7x .13.. , , ‘ V , i . L V . xive .. I .. 9.2 :I l '41,: :0! A .\i .1 ( . ntl Ill ‘ , n y , I {a}: Jinnai 3. :4: .561 n I 1‘. u If; .¢»§.(.!.L '6. I‘ll 2:9» , .21 e , . V. «fix... . uy. 1.5.3:? _ 24.2.... E y ‘ linsiiuwu , a: :1 1.1.3 V V 1}! K39- ». . V V . V . 2......nw z xmlfixin... . .‘ii- "‘XK.’¢§YA¢I" t'-‘\ i 0‘! . . . Eaétasii .53... V V V :5... iii 1‘! ... .. 9S v. 22. 3 ii .3. to... J... ..\ L o {£91, 51.. _ A .Iy‘iv‘o,flu ' x 1...... 1%.? flax-V. Yb". «swim... 3mm“? igwfihma. ._. «highs: .. . .l L 4 ”not. .nkuraun‘tifl saififirifxy, . . (“Fiat finnmn. \‘(Z‘ .i. 34" l .31.- Wie '4‘: It .. m. ‘< . i ‘ . . 1F. V . n V n . , , V , V V lwquhncllfi V . r .1 . . V . V .9 . , VFW. J , .LV. .. nu. «Hum... . .4. z .. . . 33!: @waflv , 5v; > .Vv «uni: a .