Development of Media Psychology Theory Early history Mass culture • Individuals unprotected from propaganda • Propaganda able to move masses, leads to radical action • Media powerful and broad-based, leading to a number of negative social outcomes The new propaganda war • The Nazi Party in Germany, in its attempt to gain power and to spread its ideology, applied propaganda techniques from WWI as well as a new set of techniques • Seemed to have spectacular success in the latter 1930s, capturing the government and fomenting widespread hatred toward a number of groups--especially Jews Characteristics of the mass • Great numbers • Aggregates of men widely dispersed over the face of the earth • Heterogeneous members • Anonymous individuals • No common customs or traditions – Open to suggestions – Behavior is “capricious and unpredictable” • Consists of unattached individuals – Do not play roles in a group Other research and theory that supported powerful effects • Robert Merton (a major figure in American sociology) studied a radio program with Kate Smith that was very successful in getting people to buy war bonds – Concluded that Smith’s celebrity status and image as patriotic were crucial to success. Then, in 1938 . . . • An event that seemed to prove all the worst fears people had about the startling power of the media. War of the Worlds • An accidental occurrence seemed to validate some of the worst fears about mass society and the power of the newest mass medium, the radio • Orson Welles’ Mercury Theater of the Air’s broadcast of an adaptation of H.G. Wells’ War of the Worlds on Halloween eve, 1938 Research on the broadcast • The invasion from Mars: A study in the psychology of panic (1940: Princeton University) – Cantril, Gaudet, Herzog The invasion from Mars: Radio panics America • Hadley Cantril (1940) • Personal interviews, scientific surveys, analysis of newspaper accounts and mail • Attempts were made to answer: • 1. What was the extent of the panic? • 2. What about this specific broadcast was so frightening? • 3. why did it frighten some people and not others? Two questions (p. 67) • Note: “Even this broadcast did not affect more than a small minority of listeners.” • Why did this broadcast frighten some people when other fantastic broadcasts do not? • Why did this broadcast frighten some people but not others? Why were people frightened? • Realism of the program – “sheer dramatic excellence” of the program – Use of ‘experts’ – Use of real places – early parts of the broadcast “fell within the existing standards of judgment of the listeners” – “If a stimulus fits into the area of interpretation covered by a standard of judgment and does not contradict it, then it is likely to be believed.” Standards • “Radio as accepted vehicle for important announcements.” • “Prestige of speakers.” – astronomers – military men – Secretary of the Interior • “Specific incidents understood.” – colloquial English/bureaucratese – real places, buildings, highways Standards • “Everybody baffled.” – on-air personalities claim bewilderment • “The total experience.” – projected environment – “experienced as a unit” • individual features of broadcast not adequate to explain reaction Tuning in late • CBS survey found that 42% of audience tuned in late – after intro – strong relationship to belief that broadcast was news report rather than play • “Contagion the excitement created” – Someone suggested respondent tune in after show had begun [21% in AIPO survey; 19% in CBS survey] Effect of late tune-in 80 70 60 50 40 News Play 30 20 10 0 From the beginning (n=269) After beginning (n=191) Personal characteristics Not paying attention to first announcements • “widespread habit” – station IDs and advertising • accounted for about 10% of those who misinterpreted the broadcast – actual n quite small (10% of 54 people who thought it was news reports) Personal characteristics of listeners • Possession of ‘critical ability’ • Strong religious beliefs • Personality factors such as emotional insecurity, phobic personality, lack of self confidence, fatalism Not paying attention to first announcements • most were casually scanning for something to listen to--not looking for Mercury Theater • Competing with Charlie McCarthy – vastly more popular than Mercury Theater • (34.7 to 3.6) – 18% of McCarthy listeners said they heard WOW, of which 62% said they changed after the first act (“Eddie Cantor effect”) Classifying listeners • Researchers chose following scheme from among a wide array of possible classifications: • Those who analyzed the internal evidence of the program and knew it could not be true. • Those who checked up successfully to learn that it was a play. • Those who checked up unsuccessfully and continued to believe it was a news broadcast. • Those who made no attempt to check the authenticity of the broadcast. Checked internal evidence • Did not remain frightened • 1. “Specific information they possessed and were able to project into the situation” – knew it was Mercury Theater – recognized Orson Welles – knew that time changes were too fast – knew that there weren’t three regiments of infantry in area – just recognized the events as too fantastic Successfully checked the broadcast against other information • Checked other stations (most common) – “I turned to WOR to see if they had the same thing on and they didn’t so I knew it must be a fake.” • Looked up program in newspaper – “I tuned in and heard that the meteor had fallen. Then when they talked about monsters, I thought something was the matter. So I looked in the newspaper to see what program was supposed to be on and discovered it was only a play. • Asked friend, looked out window – (1 respondent each) Unsuccessfully checked broadcast against other information • “Difficult to determine from interviews why these people wanted to check anyway” – seemed to be checking whether they were in personal danger yet rather than whether reports were authentic • Type of checking behavior was “singularly ineffective and unreliable” Unsuccessful checks • Methods – Look out window or go outdoors (employed by 2/3 of this group) – Called friends or ran to consult neighbors – Telephoned police or newspapers – “Only one turned his radio dial. Only one consulted a newspaper.” Reasons checks were unsuccessful • New information only verified their interpretation – “I looked out of the window and everything looked the same as usual so I thought it hadn’t reached our section yet.” – “I went outside to look at the stars. I saw a clear sky but somehow was not reassured.” Reasons checks were unsuccessful • Observed data were interpreted as additional evidence that the broadcast was true – “We looked out of the window and Wyoming Avenue was black with cars. People were rushing away, I figured.” – “We tuned in to another station and heard some church music. I was sure a lot of people were worshiping God while waiting for their death.” • “I looked out of my window and saw a greenish-eerie light which I was sure came from a monster. Later on it proved to be the lights in the maid’s car.” • Others felt unable to trust their own observation, believing others knew more about the situation than they did. – Trusted the announcer on the radio as a source – “My son came home during the excitement and I sent him out to find one of the elders in the church to see what it was all about.” Made no attempt to check the broadcast • “Over half of the people in this group were so frightened that they either stopped listening, ran around in a frenzy or exhibited behavior that can only be described as paralyzed.” Reasons for not checking • So frightened they never thought of checking – “We didn’t try to do anything to see if it were really true. I guess we were too frightened.” • Adopted an attitude of complete resignation – “I didn’t do anything. I just kept listening. I thought if this is the real thing you only die once-why get excited?” Reasons for not checking • Some felt they needed to take action. They prepared immediately for escape or death. – “My husband said we were here for God’s glory and honor and it was for Him to decide when we should die. We should prepare ourselves.” – “I couldn’t stand it so I turned it off. I don’t remember when, but everything was coming closer. My husband wanted to put it back on but I told him we’d better do something instead of just listen, so we started to pack. Some remained constantly tuned in to see how to escape The establishment of the ‘limited effects’ paradigm End of the ‘powerful effects’ model of media influence Limited effects model • Beginning in the mid-1940s and lasting until the mid-1960s a very different view of media effects came to prevail—one that assigned a much more limited impact to the mass media – The American Soldier – Mr. Biggott studies – The People’s Choice and Voting – Klapper’s Effects of Mass Communication Persuasion • Persuasion studies are really propaganda research that tends to take an effects approach – Persuasion really could be considered propaganda – Varies from single exposure individual effects studies to time-based campaign studies of population change WWII American Soldier studies • Part of a large-scale social science investigation of American soldiers recruited or drafted for service in WWII • This part especially interested in the effects of Frank Capra’s propaganda films on the morale of new recruits/draftees – Why We Fight – Documentary explanation of the buildup to and early years of the war WWII American Soldier studies • Part of a large-scale social science investigation of American soldiers recruited or drafted for service in WWII • This part especially interested in the effects of films developed to prepare soldiers for military duty – Why We Fight • Directed by Frank Capra – Documentary explanation of the buildup to, and early years of the war Hovland • Experiments on mass communication: Persuading the American soldier in World War II – Hovland, Lumsdaine and Sheffield (1949) • Studied the impact of showing Frank Capra’s propaganda films in “Why We Fight” series to American soldiers in a training camp – Many were draftees and did not want to fight Films were intended to foster: • A firm belief in the right of the cause for which we fight • A realization that we are up against a tough job • A determined confidence in our own ability and the abilities of our comrades and leaders to do the job that must be done • A feeling of confidence, insofar as possible under the circumstance, in the integrity and fighting ability of our Allies • A resentment, based on knowledge of the facts, against our enemies who have made it necessary to fight • A belief that through military victory, the political achievement of a better world order is possible Episodes taken from Why We Fight • Prelude to War • The Nazis Strike • Divide and Conquer • The Battle of Britain Method • Show groups of soldiers different treatments, ask questions before and after to see if exposure led to change – Manipulated exposure, type of media used, movie with and without additional content presentation, etc. • Before-after control-group design • Interested in learning of information, change in attitudes and change in interest in fighting Channel • The main study here tends to indicate that interpersonal channels are more effective at changing attitudes than are mass media channels. Performance by medium Avg % Correct 60% 50% 40% 30% Avg % Correct 20% 10% 0% Control Group Movie Group Filmstrip Group In-person presentation and movie effectiveness Avg % Correct 60% 50% 40% 30% 20% 10% 0% Control Group Movie Only Movie and Introduction Avg % Correct Movie and Review Battle of Britain • Men in two camps--some exposed to film, some not – 2100 in one camp (before/after control group) – 900 in another camp (before/after control group) – 1200 (after-only control group) – Sampling by company units • Units matched on several demographic variables Battle of Britain • Before and after questionnaires were slightly different – Tried to distract men from wondering why answering twice by writing “revised” on the questionnaire • One week between exposure and after measure • Anonymity was assured Results • The movies had a significant impact on factual knowledge • Ex. Why weren’t the Germans “successful at bombing British planes on the ground”? • Ans. “because the British kept their planes scattered at the edge of the field” • Experimental group: 78% correct • Control group: 21% correct Results: Learning from films 60 50 40 30 Test Score 20 10 0 Grade School High School College Battle of Britain • Men in two camps--some exposed to film, some not – 2100 in one camp (before/after control group) – 900 in another camp (before/after control group) – 1200 (after-only control group) – Sampling by company units • Units matched on several demographic variables Battle of Britain • Before and after questionnaires were slightly different – Tried to distract men from wondering why answering twice by writing “revised” on the questionnaire • One week between exposure and after measure • Anonymity was assured Results • The movies had a significant impact on factual knowledge • Ex. Why weren’t the Germans “successful at bombing British planes on the ground”? • Ans. “because the British kept their planes scattered at the edge of the field” % of answers correct 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Experimental Group % of answers correct Control Group Results: Opinions and interpretations • Effects were not as great • “the heavy bombing attacks on Britain were an attempt by the Nazis to . . .” • Answer: “invade and conquer England” Opinions and interpretations % of answers "correct" 70% 60% 50% 40% 30% 20% 10% 0% Experimental Group % of answers "correct" Control Group Results: General attitudes • Effects on general attitudes were slight • “Do you feel that the British are doing all they can to help win the war?” – Experimental group 7% greater than control – On several measures, only 2-3% positive difference was found Results: Overall morale • The series was ineffective – Question concerning whether trainees would prefer military duty in the U.S. or overseas Overall morale % preferring overseas deployment 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% Experimental Group Control Group Results: Attitudes • Subjects were asked whether unconditional surrender by Nazis was an important war aim Attitudes % Saying unconditional surrender by Nazis was an important war aim 70% 60% 50% 40% 30% 20% 10% 0% Experimental Group Control Group Results: “Sleeper effect” • 9 weeks after exposure – Factual material was forgotten • Retained only about 50% of factual items that 1-week groups remembered – On 1/3 of opinion issues, the long-term group showed less change – However, on more than half of the fifteen issues under study, the long term group showed greater change than the short-term group Yale School • Hovland et al. set up Yale School of research on persuasion • Studied the effect of: – Source characteristics – Message characteristics – Order of presentation – Psychological characteristics of audience Hovland et al. • Set up Yale school research on persuasion 1946-1961 under Hovland's direction • Analyzed interpersonal communication, speeches • Study the effect of: – Source characteristics – Message characteristics – Order of presentation – Psychological characteristics of audience Source characteristics • Credibility – Topic: Atomic submarines • Sources: J. Robert Oppenheimer/Pravda – Topic: Future of Movie Theaters • Sources: Fortune magazine/A woman movie gossip columnist – Greater persuasion with more credible source • However, after 4 weeks difference had disappeared High Credibility Source Low Credibility Source Should antihistamine drugs New England Journal of continue to be sold Medicine without a prescription? Magazine A (a mass circulation monthly pictorial magazine) Can a practicable atomicpowered submarine be built at the present time? J. Robert Oppenheimer Pravda Is the steel industry to blame for the current shortage of steel? Bulletin of National Resources Planning Board Writer A (an anti-labor, “rightist” newspaper columnist) As a result of television, Fortune magazine with there be a decrease in the number of movie theaters in operation in 1955? Writer B (A woman moviegossip columnist) Percent rating author as ‘fair’ 120% 100% 80% 60% 40% 20% 0% Antihistamines Atomic submarines High Credibility Steel shortage Low Credibility Future of movies Percent rating author as ‘justified’ 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Antihistamines Atomic submarines High Credibility Steel shortage Low Credibility Future of movies Net change of opinion 40% 35% 30% 25% 20% 15% 10% 5% 0% -5% Antihistamines Atomic submarines Steel shortage -10% High Credibility Low Credibility Future of movies Content • Fear appeals – The greater the fearfulness, the greater the effect on interest, tension – The less fearful messages had a greater effect on intention to change behavior – Fear appeals were thought to invoke some sort of interference Message: Fear Appeals • Janis and Feshbach (1954) used three different versions of a lecture on dental hygiene. The strong appeal provoked the most tension in the audience, but the greatest change in behavior in conformity with the message was produced by the minimal appeal to fear. • Suggested that when people feel they can do nothing about the threat then they are not likely to change their behavior. Fear appeal effectiveness 60% 50% 40% 30% 20% 10% 0% Strong fear Moderate fear Increased Conformity Minimal fear Decreased Conformity Control No Change Credibility • As research proceeded, foun that credibility tended to be connected with – Trustworthiness • Homophily • Perceived self-interest – Likeability • Homophily • Attractiveness – Expertise • Formal position Content • Fear appeals – Greater fear, greater effect on interest, tension – Lesser fear, greater effect on intension to change behavior – Thought to invoke some sort of interference • Drawing an explicit conclusion – Significantly greater effect if communicator drew an explicit conclusion Message: Fear Appeals • Janis and Feshbach (1954) used three different versions of a lecture on dental hygiene. The strong appeal provoked the most tension in the audience, but the greatest change in behavior in conformity with the message was produced by the minimal appeal to fear. • Suggested that when people feel they can do nothing about the threat then they are not likely to change their behavior. Message presentation • One-sided and two-sided presentations that USSR would not soon be able to develop a nuclear bomb were equally effective • However, when exposed to opposing view, those that had earlier been presented with two-sided version retained new opinion more than one-sided audience One-sided v. two-sided argument • Radio presentation saying war would be lengthy – Presented either as one-sided argument or with additional 4 minutes discussing view that it would be short – Before/after with control group Results • One-sided argument more effective with soldiers who: – Initially supported the idea that it would be a lengthy war – Had not completed high school • Two-sided arguments more effective with soldiers who: – initially felt the war would be short – had a high school degree education or greater Message factors: Drawing an explicit conclusion • Though the results are somewhat ambiguous, Hovland tends to assume that you should draw the conclusions for your audience where complex issues are involved. He also seems to believe that it depends on your assessment of the audience's intelligence. Inoculation • It is possible to inoculate audiences against certain views. If you present them with a weakened version of the arguments against your case, then they are likely to be resistant to stronger versions of those arguments that they may come across later. By giving them a weakened version, you allow them to formulate negative cognitive responses. Message factors: Order effects • Order effects – This question of ordering revolves around what is known as primacy and recency effects. The adage that 'first impressions count' states that the primacy effect is likely to dominate, whereas “whatever I heard last is what I believe” seems to say recency is dominant • Findings are mixed here Channel • The main study here tends to indicate that interpersonal channels are more effective at changing attitudes than are mass media channels. Audience factors • Scouts who valued group membership highly were least influenced by speaker who criticized wood craft learning Personality Personality variables such as self-esteem, anxiety and depression have an influence on persuasibility. Janis's research suggests that people with low self-esteem are likely to be relatively easily persuaded. Overall conclusions • Mass media are not as overwhelmingly persuasive as the propaganda theorists considered them • Persuasive effectiveness varied by source of the message, format of the message, audience factors such as personality and channel of communication Columbia School Paul Felix Lazarsfeld emigrated from Austria to the U.S. in the 1930s. A mathematician Interested in the application of mathematics, especially the new statistics, to study of social problems • Set up research programs at Princeton (Radio Research Bureau) and at Columbia (Bureau of Applied Social Research) that combine study of practical problems with academic methods of research • Some of his first studies focused on audience size, reactions to and use of radio programming Features of the approach • Interested in individual knowledge, attitudes and behavior resulting from exposure to media messages and campaigns • Focus groups and/or surveys were the most common methods of study • Lazarsfeld could predict with 76% accuracy which candidate someone would vote for based on his demographics. That was better than the people themselves could predict. – Prediction is taken as the criterion of validity – "Cross-Pressures" • Opinion Leaders (21%) – ("Have you tried to influence someone on a political issue recently?"; "Has anyone asked your advice recently on a political issue?") – opinion leaders were thought to be a relatively small group of influential people • Lacked evidence of actual influence The People’s Choice • Lazarsfeld, Berelson, and Gaudet • Went to Rockefeller foundation, Roper agency (provided money and interviewing), Life magazine (had money) • May be the most influential single research piece in media studies • Longitudinal analysis of campaign and election of 1940 – Roosevelt v. Willkie • Erie County, Ohio -- 43,000 • mix of urban and rural voters--no really big cities, not many poor, minorities • Sandusky, the main city, and Ohio both Republican • 1 strongly Republican newspaper, 1 neutral, 1 mildly and belatedly Democratic, Cleveland Plain Dealer endorsed Willkie • Radio stations came in from other places • National publications--Time, Life, Townsend farm journals • Stable politically • Relative lack of interest groups • Content Analysis: Research Questions 1. How do people decide to vote as they do? 2. What were the major influences on them? • "Social characteristics determine political preference." Method • • • • Innovative panel design Special analysis of changers Demographics Media use • Innovative Panel Design • May June October November • 600 (Panel) 600 600 600 • 600 (Control A)600 600 • 600 (B) 600 600 • 600 (C) 600 600 • 600 (D) 600 600 • 3000 3000 600 3000 July August September 600 600 600 600 600 600 600 600 1200 1200 1200 1200 1940 results Country still overwhelmingly for Roosevelt, but less than in 1936 • Vote overwhelmingly by SES, religion • SES: Hi 71% GOP • Lo 35% GOP • Religion: Protestant 60% GOP Catholic23% GOP • Rural • Urban • Respondent’s Level of Interest in October strongly predictive of voting. Findings • Lazarsfeld could predict with 76% accuracy which candidate the person would vote for by his/her demographics. That was better than the people themselves could predict. – Prediction is taken as criterion of validity of theory. – "Cross-Pressures" Political predisposition (SES, Protestant or Catholic and rural/urban) Voted Democrat Voted Republican 100 90 80 70 60 50 40 30 20 10 0 Strongly Rep Moderately Slightly Rep Slightly Dem Moderately Rep Dem Strongly Dem Looked at cross-over voters Criteria: 1) Had to vote for FDR in 1936 2) Had to vote for Willkie in 1940 3) Had to think Roosevelt had done a good job 4) Had to delay decision until after nominations Found only 15 people Only 9 recalled issue Only 6 mentioned 3rd term issue (very stringent requirements to find issue importance) Conclusion: Issues do not show much power in voter decision Political interest • “The more interested people are in the election, the more opinions they have on political issues, the more actively they participate in a campaign, and the more they expose themselves to campaign propaganda. Political interest 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Voted Non-voters Great interest Medium interest No interest Opinion Leaders • Opinion leaders tend to be upscale • Opinion Leaders 21% – ("Have you tried to influence someone on a political issue recently?"; "Has anyone asked your advice recently on a political issue?") – opinion leaders thought to be a rather small group of influential people, although influence not actually shown • Opinion leaders tended to be upscale Opinion leader media use 25 20 15 Newspaper Radio Magazine 10 5 0 Great Interest OLs Others Less Interest OLs Others When did they choose? • 1/2 made up their minds before May • Once they knew the nomination, another 1/4 made up their mind • 1/4 made up their mind between nomination and election Time of Final Decision • Interested/informed decide early • Cross-pressures lead to later decision • Columbia did another study in 1948, then got out of campaign studies – Voting, by Berelson, Lazarsfeld and McPhee Media influence Whenever a person in the sample changed his/her vote intention, the interviewer asked why: • Democrats mentioned radio most often – (30% vs. 20% for newspapers) • Republicans mentioned newspapers – (31% vs. 17% for radio). • Over half the voters said the media had the most important impact, 2/3 said news media were helpful • “Two-step flow” • Conclusion--interpersonal communication most important Major effects of media campaign • Activation of latent predispositions – Propaganda arouses interest – Increased interest brings increased exposure – Increased attention causes the voter to select information – Votes crystallize • Reinforcement • Conversion (rare) Percent intending to vote Republican 90% 80% 70% 60% 50% White collar 40% Union labor 30% 20% 10% 0% June August October November Percent intending to vote Republican 90% 80% 70% 60% 50% Protestant 40% Catholic 30% 20% 10% 0% June August October November The Four Steps of Activation • 1. Propagada arouses interest – interest increases as campaign gains momentum, propaganda increases in volume • 2. Increased interest brings increased exposure • 3. Attention is selective – “his predispositions come into play” • Votes crystallize • Over half the voters media most important, 2/3 news media helpful • Two-step flow • Conclusion--interpersonal communication most important Other research • Prejudice studies – Prejudiced and non-prejudiced individuals read anti-racist cartoons • Information campaigns – Cincinnati United Nations campaign • Persuasion studies • Political effects surveys Katz and Lazarsfeld • Personal influence: The two-step flow of communication – Katz and Lazarsfeld (1955) • Decatur study of opinion leaders conducted by the Bureau of Applied Social Research at Columbia Assessing opinion leaders’ role in four areas of influence: – – – – 1. marketing 2. fashion 3. public affairs 4. film choice • Talked to women in Decatur about where they went to get advice on the topics Personal Influence • Delineated the characteristics of opinion leaders – position in the life cycle, SES, social contacts • Marketing—middle-aged women, especially with families • Fashion—younger and single women • Public affairs—older and more educated • Film choice—younger and single • An excellent example of this is provided by Kendall and Woolf's analysis of reactions to anti-racist cartoons. The cartoons featured Mr Biggott whose absurdly racist ideas were intended to discredit bigotry. In fact 31% failed to recognise that Mr Biggott was racially prejudiced or that the cartoons were intended to be anti-racist (Kendall & Wolff (1949) in Curran (1990)). • Kendall and Woolf's (1949) analysis of reactions to anti-racist cartoons. The cartoons featured Mr Biggott whose absurdly racist ideas were intended to discredit bigotry. In fact 31% failed to recognise that Mr Biggott was racially prejudiced or that the cartoons were intended to be anti-racist (cited in Curran (1990)). • Hastorf and Cantril (1954). Subjects were showed film of a particularly dirty football match between Princeton and Dartmouth and asked to log the number of infractions of the rules by either side. The Princeton students concluded that the Dartmouth players committed over twice as many fouls as their team. The Dartmouth students concluded that both sides were about equally at fault. • Curran suggests it might be more accurate to say 'believing is seeing' rather than 'seeing is believing'. Research on information campaigns • Cincinnati United Nations campaign – Star & Hughes (1950) • Large-scale information campaign to generate knowledge of and support for UN • Very little effect on knowledge or support • Effects that did occur were concentrated among the already informed rather than those targeted by the campaign Some reasons why information campaigns fail • Hyman and Sheatsley’s (1947) review of findings from several surveys conducted by the National Opinion Research Center • Psychological barriers to information dissemination – The Chronic “Know-Nothing’s” in relation to information campaigns – The role of interest in increasing exposure – Selective exposure produced by prior attitudes – Selective interpretation following exposure – Differential changes in attitudes after exposure Klapper’s Five Generalizations • 1. Mass communication ordinarily does not serve as a necessary and sufficient cause of audience effects, but rather functions among and through a nexus of mediating factors and influences. Limited effects • Although people are exposed to propaganda and other forms of persuasive communication, their opinions are not often or significantly changed – They selectively expose themselves to media so that they do not often confront challenging content – They selectively perceive content with challenging messages so that they only remember the supportive parts of the message or even distort the oppositional content to be supportive – The selectively recall the messages so that over time even content that was challenging at first gradually comes to be remembered as supportive • Much of what people know is not taken directly from the media but filtered through opinion leaders and interpersonal networks – Opinion leaders tend to be more knowledgeable on topics and less easily swayed than their less knowledgeable friends and acquaintances • People just don’t change their minds all that easily – They have stable attitudes that flow from their socialization and experience • Can predict attitudes from demographics