Doris Graber of the University of Illinois at Chicago studied in detail how Americans of varied backgrounds interpreted the US news media (not only television) in 1976. Data-gathering techniques included in-depth interviews, personal media diaries, story-recall and news-attention tests.
Graber makes a distinction between initial screening and information processing.
Regarding exposure to the various news media, Graber's participants differed in their ease of access. She notes that apart from a few professionals their exposure to information was opportunistic. She outlines three key steps in the initial selection and rejection of information for processing and storing: attention arousal; information selection; and information decay and forgetting.
The participants seemed to scan the news unconsciously for items of interest. Much of the news was either ignored or given only partial attention. In relation to television news, the anchor's introduction to a story spared viewers the trouble of paying attention to the details (p. 97). Three main cues alerted the participants to whether particular news stories might be of interest to them: media cues, key words and social cues. Media cueing involved prominence, lengthy treatment and, most importantly, repetition (frequent coverage). Key words acted as verbal cues in scanning newspapers at least. Finally, topics which were current with people around them led participants to attend to related news stories. However, the participants paid little attention to many of the news stories and were unable to recall these.
The participants reported that their main reasons for paying attention to news stories were personal relevance, emotional appeal and an 'interesting story'. Their diary entries emphasized human-interest stories (such as those dealing with crimes and accidents) and stories related to personal lifestyle (such as those about health, sports, entertainment and celebrities) (pp. 102-3). Those lacking interest in politics paid more attention to human-interest stories. Foreign news appealed only to a very select group (p. 103).
Participants deliberately rejected unwanted information for personal reasons, because of the nature of the story or because of its style of presentation. Some stories were rejected because they contained disturbing information, some because they were perceived as being too remote or too complicated, some because of scepticism about sources (especially politicians), and some simply to save time and energy. 'When people do pay attention to news, they process stories that seem interesting, simple to understand, and believable' (p. 107).
Information decay and forgetting
Managing information also involves clearing the mind of information that is no longer needed. Some information is never stored in long-term memory. In addition, forgetting occurs because of long-term memory decay over time and problems in locating stored information. 'The ability to retain stories in memory and retrieve them varied widely, depending on the nature of stories, the use of visuals, and the concerns and life style of the audience' (p. 115). The participants who were particularly interested in the news could recall some stories which reflected major personal concerns almost as well after nine-months as after a few weeks. Women showed higher rates of forgetting than men. Graber suggests that this was because they saw little reason for committing news stories to memory (p. 111). Older people remembered more about stories than younger people of the same sex. Graber suggests that this was because of their greater knowledge of the world (p. 110). More educated people recalled more than less educated people. Interest in the news appeared to be the main reason for above-average recall. 'But at best, memory for news stories was quite limited when it came to retention of detail. For the most part, recall was hazy and incomplete' (p. 115). Whilst individual stories were forgotten, however, they may nevertheless contribute to developing general schemata.
Graber outlines 'a transactional model of communications effects': 'The incentives to learn or not to learn from media messages hinge on three factors' (p. 119). She lists these as message factors, audience motivations, and contextual factors.
In 'message factors' Graber includes both the form and the content of the material. 'Media cues about the importance of issues are accepted for selected issues only' (p. 125). If the audience is already very familiar with the topic (e.g. through direct experience or prior stories), further media treatment of it becomes 'obtrusive' and the audience becomes, in Raymond Bauer's term, 'obstinate': they have reached closure - their minds are made up. On topics which touched the lives of the participants less often they were more likely to accept guidance from the media.
According to uses and gratifications theory, 'people pay attention to information that is useful for them in their daily pursuits or that provides psychological gratifications' (p. 127). In Graber's study, psychological gratifications were the major reason for participants selecting stories (hence the attention paid to human interest stories and those relevant to personal life-style). Some stories were selected because of job relevance or because they 'satisfied the need to act as a "good citizen" interested in important public affairs' (p. 128). Another gratification was the reduction of uncertainty concerning pending decisions (e.g. election information, weather, stockmarket and conditions in places to which participants expected to travel). Often the expected gratifications were not forthcoming: stories were often criticized for not giving the participants what they wanted.
Cognitive balance theorists argue that people avoid dissonant information - information that conflicts with their existing knowledge, attitudes and feelings or which they find disturbing or threatening. 'They seek out information that is reassuring and congruent with their beliefs' (p. 130). Graber's study showed some limited evidence for this theory. The participants often (but not always) rejected information that they found annoying or disturbing. And when the issue had become obtrusive both supportive and contradictory information tended to be rejected. However, sometimes people did adjust their beliefs in response to discordant information.
According to agenda-setting theories, people accept guidance from the media in determining what information is important. Media cues include frequent coverage and prominence. Media cues often influenced the attention paid by the participants to news stories, although they did not guarantee attention. 'Media agenda setting, to the degree that it does take place, is a powerful force in determining which problems are taken seriously and in providing the context within which policies and individuals will be judged' (p. 133). However, people ignore media cues when their minds are already made up, when they doubt the arguments presented or when the information is disturbing.
The most important contextual factors in the interpretation of news include lifestyle and political socialization, prior knowledge and life experiences, current needs for various types of information, and attitudinal factors such as interest in news and perceived credibility of sources (pp. 119 & 133).
Early socialization establishes a hierarchy of purposes for media use (p. 136), tending to emphasize the primacy of, for example, informational or entertainment purposes. The participants were aware of widespread behavioural norms regarding media use. Childhood habits tended to persist into adulthood. 'Panelists whose current life-style made media use more difficult than in the past usually retained their original norms of how they ought to behave' (p. 134; my emphasis). The participants were aware of a stereotypical norm for males, adults and those in higher socioeconomic groups and with more education to take more interest in the news than others. 'In our panel, none of the women expressed high interest in politics compared to one-third of the men' (p. 135). Men and women claimed differences of concern which conformed to stereotyped patterns of use, but news attention test scores showed few statistically-significant gender differences in attention to particular news topics. However, women showed less inclination to participate in political discussions (which may help to explain why their recall of news stories was weaker than that of the men). For participants from small communities, shared social concerns also influenced news selection. Where participants had moved away from their home towns and had adopted quite different life-styles, 'social pressures linked to life-style' seemed 'to outweigh the effects of prior socialization' (p. 136).
As for prior learning, the availability of information is influenced by social, political and economic conditions. 'For most news stories the impact of prior information is profound. It affects the kinds of details that will be absorbed and the perspectives from which the story is viewed. It also determines which stories will be processed and which will be ignored' (pp. 136-7).
Prior knowledge provides schemata that facilitate the integration of new information. High prior knowledge amongst the participants led to high learning rates (p. 138). Prior knowledge stimulates learning until a saturation point is reached and then suppresses it.
The most important psychological predisposition affecting learning from the news was interest in it. Interested people learned more from the news, regardless of such factors as prior knowledge or educational level. Those with little interest in news could not focus their attention. Reasons for interest varied amongst the participants from relevance to daily life to detached curiosity. Specific purposes offer the greatest impetus for learning (e.g. needing to know more about candidates just before an election).
The final contextual factor is also attitudinal: it relates to credibility of sources. The participants routinely complained about general media distortions (such as sensationalism), imbalance and sometimes bias. But specific stories were usually regarded as credible except where they directly contradicted personal experiences (and even then they referred only to exaggeration or imbalance rather than rejecting the story as totally false). Lack of credibility of sources was not a major disincentive to learning.
'In addition to paring down the flow of information by ignoring large numbers of stories, people use a processing strategy that further reduces the amount of information that needs to be stored. This strategy is schematic thinking' (p. 250; my emphasis). Schematic thinking makes use of mental templates (schemata; sing. schema) for people and events.
Schemata also help in the recall of appropriate (if not specific) details.
After sensory data has been initially screened, new information is simplified for storage in short-term memory. It is then checked to determine whether an existing schema can be appropriately applied to it or whether it calls for the creation of a new schema. If so, it is integrated into the repertoire of schemata; otherwise it is quickly forgotten. As a result of childhood socialization, adults have a large repertoire of schemata for integrating information which is familiar within their culture or subculture. Graber offers this example: 'A story about a fatal fire in a Hispanic neighbourhood can be incorporated into schemata about tragic unavoidable accidents or into schemata about disregard for safety in minority neighbourhoods. The story might even lead to the creation of a new schema about fire prevention' (pp. 147-8).
Graber notes that whilst schematic thinking helps to reduce the problem of information overload, 'it does not lead to the retention of a large amount of factual data about specific events' (p. 250). In processing news stories, people do not store details and background information 'Details that do not seem essential at the time and much of the context of a story are routinely pared. Such levelling and sharpening involves condensation of all features of a story' (p. 161).
The schema process involves the incorporatation of inferences which may or may not be correct. The panelists frequently drew on their schemata to fill information gaps in news stories. For instance, their schemata about human behaviour permitted them to make inferences about character traits and behaviour to be expected from people even when stories were silent about these traits and behaviours. Politicians' actions were generally viewed as motivated by a desire to gain favourable publicity. (pp. 155-6)
Graber found that up to a third of her 48 subjects in one experiment made additions to both the verbal and visual information in each of 12 television news stories shown to them, deriving these from schemata stored in memory (p. 170).
It is hardly surprising that people have only a vague memory of news stories and are unable to recall details or to distinguish among various incidents. 'For most of the stories that our panelists recalled, they had processed meanings beyond a mere recapitulation of the mere facts' (pp. 193-4). Judgements about the significance of a story (motivations, implications etc.) may remain after the story is forgotten (p. 149). 'After people have forgotten specific events, the attitudes distilled from them are likely to be retained as part of a schema' (p. 259).
The participants used three major types of information processing strategies: relatedness searches, segmentation and checking.
Relatedness searches involve looking for similar situations amongst existing schemata. Most adults have a broad range of schemata relevant to events in news stories: this enables them to perform effective relatedness searches for large numbers of stories (p. 213). 'The choice of a particular schema seemed to hinge on the perceiver's interests and societal setting when the information was received, on the cues inherent in the information, and on the availability and accessibility of suitable schemata' (p. 153). The priming phenomenon (p. 159) was also important: 'When individuals had several suitable schemata available for filing particular information, schemata that had been used recently were most likely to be chosen again' (p. 154; my emphasis). Graber relates this to the context effect which complicates the interpretation of interviews or questionnaires: 'The cues inherent in the form of the question tend to influence which schemata are likely to be searched. Once a schema has been brough to the forefront of memory, it becomes part of the context that influences how subsequent questions are answered' (p. 159).
Relatedness searches were easier for participants who had a broad information base and who used it frequently. Knowledgeable participants learned new information far more easily than others did.
Relatedness searches often lead to mismatches. Robert E Lane (cited in Graber 1988: 158) argues that mismatches are generated by three factors:
Graber found that stories about third-world countries were often misinterpreted due to lack of information and appropriate schemata. Cultural stereotypes were drawn on. For instance, African politics is imagined as primitive and corrupt (p. 158).
Graber also discusses the importance of the ways in which news stories are presented:
'During relatedness searches, the information extracted from a news story may be integrated into a single schema, or it may be segmented and the segments embedded in several schemata. Alternatively, the whole story may become part of several schemata' (p. 250).
Segmentation is another matching strategy. Instead of seeking matches for a story as a whole, it involves seeking matches for selected aspects of it. Complex, multifaceted stories can be segmented into simpler ones. Graber offers this example: 'a complex story about parents boycotting a school to prevent entry of students bused from a distant location does not have to be related to stories involving nearly identical situations. Instead, the story can be segmented and integrated as an example of political participation, an incident that disrupts public education, an example of racial or ethnic prejudice, or a case of stubborn resistance to change. Each of these distinct perspectives, inherent in the story, provides an opportunity for integrating it into a different schema' (p. 161). Segmentation may lead either to multiple integration or to limited processing - the storing of selected segments only. Graber suggests three reasons for limited processing: limited knowledge, labour-saving and the prioritizing of information that is of special interest to individuals. She offers an example of limited processing based on labour-saving: 'several panelists automatically filed news of street crime on the basis of racial categories of likely offenders and victims. This segmentation allowed them to fit the news neatly into their established racial stereotypes' (p. 163). Multiple integration makes it far easier to retrieve stories than limited processing does (p. 161).
Graber's participants varied in their ability and inclination to use the segmentation strategy (p. 160).
Checking is the third major information processing strategy. 'It involves going beyond the first schema that comes to mind for storing or retrieving information and continuing to search for additional schemata' (p. 164). This is undertaken in order to find a better fit, to create a broader basis for judgements and to assist the retrieval of information from memory. Graber's participants examined only a limited number of options rather than running through many possibilities: this is a satisficing strategy (rather than an optimizing strategy). 'Their ability to locate appropriate schemata quickly varied with their general intellectual abilities and with the degree of familiarity with a particular situation' (p. 164).
'The individual's cognitive style determines which of the various cognitive procedures is undertaken, singly or in combination. Cognitive style, in turn is related to intellectual abilities and experience' (p. 165). The most common strategy seemed to be straightforward matching without segmentation or checking. But straight matching often frustrated the processing of information. Segmentation seemed much more effective. Checking seemed to be fairly rare. 'People who are able to segment information and engage in checking... tend to have more complex and richer schemata' than others (p. 189).
Information was rejected by participants for several reasons:
Sometimes new information led to some existing schemata being discarded or revised. However, 'major schema revisions are resisted unless there are strong social or circumstantial pressures for revision' (p. 176). Graber cites Robert Lane: 'Changing ideas is a strain not to be lightly incurred, particularly when these ideas are intimately related to one's self-esteem. The less education one has, the harder it is to change such ideas' (p. 176).
Occasionally, when participants had difficulty finding appropriate schemata to evaluate information, they resorted to generalized decision rules, sometimes expressed as proverbial commonsense: 'it's darkest before the dawn'; 'once a loser, always a loser'. When the accuracy of information was in question, the participant assessed the credibility of the source or (less often) tested the new information for logical consistency with existing information (p. 177).
Strategies for processing visual information
Graber notes that:
She adds that 'the prevalence of closeup views of people means that television news pictures are primarily sources of information about human actors' (p. 168). She found that visual cues were used to assess the credibility of news personnel. Closeups were also powerful in attracting and holding the viewer's attention. 'People often expressed a desire to help suffering fellow humans shown on television' (p. 168). In contrast to pictures of people, pictures of objects and places became important sources of information for the participants only when they depicted unfamiliar situations which they found it hard to imagine. Most memorable were closeups of famous people and closeups of unfamiliar people in exotic situations (p. 169). Viewers generally failed to recall: pictures used to clarify verbal statements; establishing shots; and distant shots of activities (p. 169). People focus on the anchor's lead-in for each news story and concentrate on pictures which are easy to make sense of; only extraordinary images challenged this pattern (p. 170).
Graber found that far fewer participants added visual details than embellished the audio text (p. 170); schematic processing of visuals was less frequent than schematic processing of text. She suggests several reasons for why the processing of visuals may be more data-based than schematic:
Graber's participants stressed several important functions of pictures in news stories:
Graber's participants routinely used six types of schema dimensions for processing news:
Graber notes that the two schema dimensions used most often were the cause-and-effect and human-interest and empathy dimensions, though the distribution of dimensions varied according to the types of issues presented (p. 214). She adds that the frequency of use of a schema dimension 'depended on personal idiosyncracies, educational differences, and differences in political outlook and interest' (p. 193). The use of schemata by all of those in her study revealed 'a good deal of shared stereotypical thinking' (p. 214).
Graber argued that her participants, and indeed 'average Americans' in general, did not organize their thinking and schemata in terms of an overarching ideology such as liberalism, conservatism, Marxism or capitalism (pp. 223 & 188). She suggests that 'contrary to cognitive consistency theories, schema theory suggests that people are not highly motivated to make their thinking consistent with general principles and to apply the same general principles to related matters' (p. 223). However, she does note that all of her participants had 'adopted culturally sanctioned values as the schematic framework into which schemata covering more specific matters were then embedded' (p. 252). 'Broad value principles are the closest thing to an overarching belief structure' (p. 254).
Graber rejects the 'hypodermic effect' whereby media messages are internalized exactly as presented (p. 258). However, in relation to topics on which people are dependent on media information, she acknowledges that the media are likely to influence the schemata formed (p. 264). The findings from this study undermine cognitive balance theories and suggest that whilst people may prefer information according to their established beliefs, this preference is not absolute (p. 258). She rarely found any automatic rejection of discordant information. She concludes that the cognitive processing of news follows patterns which 'are best explained in terms of schema theory': 'In the realm of political news, our panelists coped with the flood tide of information through a manageable, limited array of schemata that were simple and sparse in basic dimensions as well as in the number of themes included in the dimensions' (p. 245). She adds that 'despite substantial uniformity in the general shape of schema systems, much of it abetted by uniformity in media presentations, diversity among the schemata created by each panelist... remained a striking reality' (p. 245).
Doris Graber insists that:
Notes on: Doris Graber (1988): Processing the News: How People Tame the Information Tide (2nd edn.) New York: Longman
Daniel Chandler
August 1997