Loading “EBSCOhost”

Full Text Word Count: ISSN: Accession Number: Database: Full Text Database: Back 12 page(s) will be printed. Record: 1 Switching channels: The effects of television channels on the mental representation of television... Leshner, Glenn Reeves, Byron Nass, Clifford Journal of Broadcasting & Electronic Media; Winter98, Vol. 42 Issue 1, p21, 13p, 2 charts Article *TELEVISION broadcasting of news *TELEVISION viewers -United States UNITED States NAICS/Industry Codes515120 Television Broadcasting Examines the effects of television channels on the mental representation of television news by audiences in the United States. Rating on news attributes; Evaluation of generalist and specialist channels; Quality of television picture. 5392 0883-8151 903003 Academic Search Premier Communication & Mass Media Complete SWITCHING CHANNELS: THE EFFECTS OF TELEVISION CHANNELS ON THE MENTAL REPRESENTATIONS OF TELEVISION NEWS This study tests how people differentiate between television channels, conceptualized as having two important properties: channels differ in degree of specialized content and are positioned within arrays varying in size. Participants watched news stories identified as emanating from either specialist news channels or from generalist channels. News on specialist channels was rated higher on news attributes and evaluated more positively than identical news on generalist channels. News watched on one channel was rated as more similar than identical news watched on four channels. Participants who watched news on four channels rated the television picture quality higher than those who watched on one channel. This pattern of results is consistent with the notion of channel as a place where television programs--and the people and action in them--exist. * The rapid advances in communication technologies pose many new and interesting challenges for manufacturers, designers, program producers, policy-makers, and consumers. One of these new technologies includes the prospect of receiving 500 television channels in the home ("What Cable Could Be," 1991). This study is concerned with understanding what the meaning of huge numbers of channels might be to viewers, psychologically. That is, do the ways people mentally represent television channels influence how they think and feel about the programs? The goal of this experiment is to increase our understanding of how characteristics of 2/21/08 12:05 AM Loading “EBSCOhost” Page 2 of 11 http://web.ebscohost.com/ehost/delivery?vid=8&hid=116&sid=0f3893a7-4628-4dbe-af98-f40fb68af4b1%40sessionmgr103 television channels affect the ways people respond to the programs they watch. Specifically, this study experimentally examined the effects of television channels on the judgments and evaluations of program material, and the extent to which program segments were perceived as similar to each other. An important notion in this approach is that the concept "television channel" is assumed to be perceived as a place within the television set that can affect viewers' evaluative responses to the program content seen on the screen. Although no previous research was found relating channel as a place to television content, some literature exists that introduces the notion of place with respect to television sets and illustrates its effects on responses to the program content. This research is important for thinking about channel as a place[ 1] and also for suggesting criterion variables that might be affected when the sense of place is varied. Several studies have shown that viewers sometimes respond to television sets as places; that is, viewer expectations about what they will see encourage different responses to identical content. These studies have approached the box-as-place in essentially two ways: ( 1) the box as part of the spatial arrangement in a room, and ( 2) the box as a container for a role or function. The first approach demonstrates the influences of spatial arrangements between the box and viewer. Three experiments share the theoretical assumption that physical space is a critical cue guiding attention, memory, and behavior. Manipulations of viewing distance (close versus far) were treated as changes in the arrangements of physical space in two of the studies (Lombard, 1995; Reeves, Lombard, & Melwani, 1992). Because faces were shown on the screen, viewing distance represented differences in the arrangement of personal space. Both studies showed that closer viewing distances (just as closer interpersonal distances) encourage the evaluations of people on the screen to be more extreme. A third study demonstrated the relationship between physical space and memory. In an experiment in which participants watched a news and public affairs program, Rothkopf, Dixon, and Billington (1986) showed that the assignment of each of the program's discussants to his or her own television screen increased the likelihood that viewers would attribute statements to their proper source than if all discussants appeared on one screen. Such an effect was found presumably because viewers had a spatial cue to aid memory. These experiments suggest that the television box can define the place where people and action exist. Thus the placement of the box constitutes the physical relationship of people and action within it to other people and things in the viewing environment. Most notably, the box establishes the spatial relationship between the content of the screen and the viewer. The second approach--the box as a container for a role--considers the assignment of particular roles to different television sets, regardless of whether the assignment is made consciously or not. Nass, Reeves, and Leshner (1996) found that participants evaluated identical program content differently depending on the function of the TV. In their study, responses to television segments differed depending on whether viewers watched TVs that showed only one type of program or a TV that showed more than one type of program. Particular boxes encouraged different responses to identical content depending on the boxes' roles. If boxes serve as places where people and action exist, might channels also serve as such places? Defining Channels Television channels can be defined in a variety of ways, but most of the commonly accepted definitions attempt to summarize content. MTV as a channel means that its program content is primarily music videos. Similarly, the Weather Channel, BET, ESPN, etc., summarize a particular and fairly consistent type of content. A genre approach to defining television content can be daunting, especially given that most of the new channels are expected to be genre-specific. The more new channels that come on line, the greater the number of 2/21/08 12:05 AM Loading “EBSCOhost” Page 3 of 11 http://web.ebscohost.com/ehost/delivery?vid=8&hid=116&sid=0f3893a7-4628-4dbe-af98-f40fb68af4b1%40sessionmgr103 categories. In theorizing about channels in general, benefit would tend to diminish with the increase of new channels tied to specific genres. For this study, we were interested in finding channel attributes that are not tied to any particular program content, but rather, ones that can be descriptive of all channels--current ones and those yet to come. The two channel attributes we chose that might have consequences for how content is mentally represented are specialization and structure. Channel specialization describes the tendency for a channel to send homogenous messages. Although this attribute attempts to describe content, it is a dimension that is independent of an idiosyncratic specialty of any particular channel. A channel can vary along this dimension either by sending similar messages (e.g., CNN, which sends news only), or by sending dissimilar messages (e.g., NBC, which sends news, comedy, drama, documentaries, etc.). Structure means the access structure of any particular channel environment--the way channels are arranged and the way in which programs are selected by viewers. Rather than test a variety of structures, we first wanted to see if structure mattered at all. Therefore, we decided to test between television program content seen on the same channel or on different channels. We call this channel differentiation. We made separate predictions based on whether specialization or differentiation was manipulated. For channel specialization, we first asked: What is the extent to which channel specialization encourages television news stories to be perceived as representative of news? Then we asked how specialization affects the evaluation of news. Channel Specialization. The first question in effect asks: Is news more typical, or "newsier," if it comes from a specialist channel than if it comes from a generalist channel? Research has shown that people judge identical content differently depending on the role assigned to a computer screen (e.g., Nass & Steuer, 1993; Nass, Steuer, Henriksen, & Dryer, 1994; Nass, Steuer, Tauber, & Reeder, 1993). Nass and his colleagues concluded that as with people, once we learn to expect consistent behavior from a source, the source becomes specialized. This line of research suggests that people will respond to identical television content differently to the extent that television channels are specialists or generalists. Although there are no studies that examine the effects of channel specialization on viewer judgement of the representativeness of television programs, evidence suggests that the extent to which content is homogenous on a television set impacts viewer judgments about the program content seen on the screen. Nass, Reeves, and Leshner (1996) found participants who watched news segments on one TV and entertainment segments on another, but identical TV ("specialist TVs") rated the news higher on news attributes and the entertainment higher on entertainment attributes than participants who watched the same news and entertainment segments all on one TV ("generalist TV"). Their interpretation of these results is that people attributed a role to the television set that corresponded to the content type displayed. The expectation for this study is that a similar effect should be found for specialist news channels. So, a news story from a specialist news channel ought to encourage higher news-attribute judgments than a news story from a generalist channel. That is, stories on specialist channels ought to be more representative of news. H1: Specialist channels will encourage television news stories to be perceived as more representative of their specialty than generalist channels. Summary evaluations of television program content--story liking and perceptions of story quality--have also been shown to be affected by specialized television sets (Nass, Reeves, & Leshner, 1996), such that evaluations of stories from specialist channels should be more positive than from generalist channels. H2: Specialist channels will encourage more positive evaluations of television news stories than generalist channels. 2/21/08 12:05 AM Loading “EBSCOhost” Page 4 of 11 http://web.ebscohost.com/ehost/delivery?vid=8&hid=116&sid=0f3893a7-4628-4dbe-af98-f40fb68af4b1%40sessionmgr103 Channel Differentiation. What possible effects could seeing television program material on one channel vs. different channels have? If we think about the current structure of channels, some possibilities come to mind with respect to perceived similarity of stories. By its very nature, the current number-line address structure of channels may differentiate program content by placing it at different locations along the line. Put another way, television program sources occupy a place within an array. There is a growing literature in the study of numerical cognition that shows that some people mentally represent numbers in some sort of structured spatial array (e.g., Seron, Pesenti, Noel, Deloche, & Cornet, 1993). This research, when applied to channel arrays, suggests that the structure of these arrays--due to their number-line orientation--may influence how people mentally represent their channel environments. Such research may help describe important properties of numbered arrays and their consequences for channel as place. Seron et al. (1993) collected self-report data from people who claimed to visualize numbers in a structured, spatial configuration and were able to report those structures on paper. All of these structures had at least two spatial dimensions, and some even had colors or temperature associated with parts of their number structures. Although there was large variation in respondents' number structures, there was high consistency within people over time. Seron et al. (1993) also reported research more than a century old that found that some people had visual-spatial representations for other serial orders, such as hours of the day, days of the week, months of the year, and letters of the alphabet. Might people also have mental spatial representations for their channel environment? And might this mean that television program content can be spatially represented along a number line? Research has shown that people are predisposed to represent numbers along a number line. Several separate studies have demonstrated a phenomenon called the distance effect, which indicates the effect of proximity along a number line to judgments of objects along that line (Dehaene, Dupoux, & Mehler, 1990; Hinrichs, Yurko, & Hu, 1981; Moyer & Landauer, 1967). Moyer and Landauer (1967) showed that the time to decide which of two numbers is the larger (or the smaller) smoothly decreased with the numerical distance between them. Dehaene et al. (1990) had participants decide whether a given two-digit numeral, e.g., 59, was larger or smaller than a reference number, e.g., 65. They found that numerical distance had a significant effect on comparison time--reaction times were longer for numbers close to the reference and shorter for numbers far from the reference. They concluded that proximal numbers are harder to discriminate than distal numbers. Thus, a mental number line has properties that may affect how viewers psychologically respond to programs presented on different channels independent of the content of a particular program. That is, if proximal numbers are harder to discriminate than distal numbers, might programs located in the same place also be harder to discriminate, and thus, be seen as more similar than programs located at distal places on the number line? If so, then television program content may be mentally locatable along a number line. If a number line is a way people represent their channel environment, then the size of the array should produce certain effects. As a starting point, we compared array size, where a onechannel array should inhibit discriminability between program segments, and a four-channel array enhance it. The expectation is that channel differentiation will affect program similarity perceptions. H3: Undifferentiated channels will encourage television news stories to be perceived as more similar to one another than differentiated channels. Method Participants. Forty-eight undergraduate and graduate students enrolled in a communication class at Stanford University volunteered to participate in this study. Each participant received class credit. There were 18 females and 30 males. The mean age was 21.2 years with a range from 18 to 29. Overview and Design. The experiment was a between-subjects, full-factorial two-by-two 2/21/08 12:05 AM Loading “EBSCOhost” Page 5 of 11 http://web.ebscohost.com/ehost/delivery?vid=8&hid=116&sid=0f3893a7-4628-4dbe-af98-f40fb68af4b1%40sessionmgr103 design. The two factors were channel specialization (specialist/generalist channels) and channel differentiation (one/four channels). The experimental session lasted about 30 minutes for each participant. Procedures. Each participant was escorted into a viewing room and seated in a chair located about 3-and-one-half feet from a color television monitor. Participants were told by the experimenter that they were participating in a study about how people respond to television news programs, and that they would see several news stories and answer a few short questions about each one. Each participant was supplied with a question packet that contained explicit instructions, including the list of channel buttons each participant was required to press at specified times. In the generalist condition participants watched news stories, which were identified as either ABC, CBS, NBC, or Fox by both the experimenter during the pre-experiment protocol and by identifying graphics prerecorded onto the news stories. In the specialist condition, participants watched news stories, which were identified as either CNN, HNN (Headline News Network), SNC (Satellite News Channel), or RNN (Regional News Network), again, by both the experimenter during the pre-experiment protocol and by identifying graphics prerecorded onto the stories. In the one-channel condition, participants watched twelve news stories on one of the following channels: 12, 19, 61, or 68. In the four-channel condition, participants watched three news stories on each of four channels: 12, 19, 61, and 68. Each participant took part in the experiment in an individual session, and the experimenter was present only during the reading of the protocol. In the one-channel condition, participants were asked to select on the remote control a particular channel to begin watching the news stories, then to answer questions after each story on the pencil-and-paper questionnaire, which asked for responses to the news representativeness questions. After participants watched a story and answered the corresponding questions, they were instructed to hit the START button on the remote control to watch the next story, and each subsequent story, until all 12 stories had been watched, and all 12 sets of questions answered. In the four-channel condition, participants were asked to select a particular channel on the remote control to begin watching the news stories, then to answer the representativeness questions after each story. After participants watched a story and answered the questions, they were instructed to change the channel on the TV by pressing the button on the remote control that corresponded to the next channel indicated in their instructions. All channel number selections were predetermined in the instruction packet, and the channel number sequences for all participants in the four-channel condition were the same: first 12, then 19, then 61, then 68. That sequence was repeated so that each four-channel participant watched three stories on each of the four channels. The story sequences were selected so that only one source (e.g., ABC) would appear on each channel for all fourchannel participants. After participants completed watching all 12 stories and answering 12 sets of questions, the questionnaire instructed them to continue to a final set of questions--story evaluation, story similarity, and picture quality. The experimenter then debriefed participants on the intent and design of the experiment. Participants were queried to see if they were able to discern the manipulations of the experiment, and none could? The experimenter responded to all questions, then thanked and excused the participants. Stimulus Material. Various television news stories were acquired that had aired on local television news programs throughout the U.S. Twelve news stories were selected from this group. Stories were between 28 seconds and 1 minute and 40 seconds in length. Stories that could easily be identified with salient past news were avoided. Each story was recorded on 3/4-inch videotape and was electronically mixed with an identifying graphic to indicate the source of the story. 2/21/08 12:05 AM Loading “EBSCOhost” Page 6 of 11 http://web.ebscohost.com/ehost/delivery?vid=8&hid=116&sid=0f3893a7-4628-4dbe-af98-f40fb68af4b1%40sessionmgr103 Each graphic was shot individually with a studio color television camera and mixed into each story using the "wipe"[ 3] function of a video switcher. The graphics covered the lower quarter of the television screen in order to cover all preexisting identifying graphics. Eight versions of each story were re-recorded onto 3/4-inch videotape, generating 96 stories. The 48 generalist stories (12 stories x four generalist sources) were pressed onto a video laser disc. Also, the 48 specialist stories (12 stories x four specialist sources) were pressed onto a separate video laser disc. Four random orders were selected for presentation of stories to participants. Apparatus. Story sequences were created and stored with a computer and were played on a videodisk player and displayed on a 19-inch color video monitor. Audio was set to average -68 dB (Alten, 1990). A 22-key calculator-style numeric keypad served as the remote control. All function keys were covered with blank labels. The numeral keys were covered with the "channel" numbers used in this study, and the "Enter" key was covered with a "START" label. Dependent Measures. Four sets of dependent variables were measured in this study: ( 1) representative judgments of each story on five news attributes, ( 2) two overall evaluation ratings of the news stories, ( 3) an overall rating of how similar the stories were, and 4) three television picture quality ratings. Participants answered five representativeness questions immediately after seeing each news story. They were asked to judge how "important," "informative," "disturbing," "serious," and "interesting" each news story was. All of the questions were 10-point response scales, anchored by "Not at all" and "Very" at the extremes.[4] The last page of the questionnaire asked for the participants' summary evaluations of how much they liked the stories, their rating of the quality of the news stories, their overall rating of how similar the stories were, and their ratings of three attributes of the television picture: picture clarity, picture brightness, and color quality. All of these questions were also 10-point response scales, anchored by "Not at all" and "Very." Index Construction. Three indexes were created, one each for news representativeness, evaluation, and picture quality. The News Index was created by summing participant responses on each of the five news representativeness variables--"important," "informative," "disturbing," "serious," and "interesting"--across the twelve stories. A confirmatory factor analysis was performed on the News Index and all loaded on one factor (eigenvalue = 3.58; 71.5% of the variance explained, Cronbach's alpha = .90). An Evaluation Index was created by summing the liking and quality responses. The confirmatory factor analysis showed that these items loaded on one factor (eigenvalue = 2.61; 65.2% of variance explained, Cronbach's alpha = .81). The Picture Quality Index was created by summing responses to picture clarity, color quality, and picture brightness. The confirmatory factor analysis also showed that these items loaded on one factor (eigenvalue = 2.16; 72.0% of variance explained, Cronbach's alpha = .80). Results Hypothesis 1 predicted that news stories on specialist channels would be rated higher on news attributes than the same stories on generalist channels. A t-test was performed on the News Index, the results of which are shown in Table 1. As Table 1 shows, and strongly consistent with the first hypothesis, the difference between the means for generalist and specialist channels for the News Index is significant (t(46) = 1.70, p < .05, d = .45). News stories seen on specialist channels were rated higher on typical news attributes than were stories seen on generalist channels. Hypothesis 2 predicted that news stories seen on specialist channels will be evaluated more 2/21/08 12:05 AM Loading “EBSCOhost” Page 7 of 11 http://web.ebscohost.com/ehost/delivery?vid=8&hid=116&sid=0f3893a7-4628-4dbe-af98-f40fb68af4b1%40sessionmgr103 positively than stories seen on generalist channels. Table 1 shows that, for the Evaluation Index, participants who watched news on specialist channels evaluated the news more positively than participants who watched the generalist channels (t(46) = 1.56, p < .07, d = .25), showing some support for the second hypothesis. Although the first two hypotheses predicted only a main effect for channel specialization, analysis of variance (ANOVA) was performed on the News and Evaluation Indexes to determine if there were significant interactions for each with channel differentiation. No interactions were significant, nor was there a significant main effect for channel differentiation. Hypothesis 3 predicted that news stories seen on one channel will be rated more similar to each other than news stories seen on four channels. The data support this prediction. As Table 2 shows, participants who watched the stories on one channel rated the news stories as more similar to each other than participants who watched the stories on four channels (t(46) = 1.53, p < .07, d = .44). The interaction between channel specialization and channel differentiation for the similarity variable was not significant, nor was there a significant main effect for channel specialization. Picture Quality. Although there were no specific hypotheses for the Picture Quality Index, the data were analyzed based on the channel specialization manipulation. As shown in Table 2, mean ratings on the Picture Quality Index were significantly higher for participants who watched the news stories on four different channels than for participants who watched the news stories on one channel (t(46) = 3.31, p < .001, d = .96). It seems that participants used the channel differentiation attribute of the TV set to make attributions about physical qualities of the picture. In this study, at least, pictures on a TV that has many channels were perceived as better than those on a TV that has only one channel. There was no difference in picture quality based on channel specialization. Discussion The goal of this experiment was to explore how characteristics of television channels affect the way people think and feel about the programs they watch. Channel specialization--the degree to which the set of programs a source transmits is homogenous--was predicted to affect participants' judgments about how representative the program content was of a particular specialty, and was also predicted to affect participants' evaluations. Channel differentiation--whether program material was presented on one channel or on different channels--was predicted to affect perceived similarity of the content. For the most part, the results were consistent with these hypotheses. In addition, there is ample evidence that these two attributes of channel are independent of each other, because no significant interactions between channel specialization and channel differentiation were found for any of the dependent variables. Channel specialization had a significant effect on the perceived representativeness of the news stories. Overall, the news on specialist channels was perceived as more representative of a news specialty than the news on generalist channels. An alternative explanation could be that participants had previous attitudes about the existing generalist networks that were less positive than their attitudes about the specialty networks used in this study. As such, the more positive ratings of the stories seen on the specialist channels may simply be a reflection of preexisting attitudes toward some or all of the channels. Because participants' attitudes toward these networks were not measured, this explanation cannot be ruled out. However, the findings of this study closely correspond to previous research that showed that program content is more representative of a specialty when presented on a specialist TV than when presented on a generalist TV (Nass, Reeves, & Leshner, 1996). Hence, we are confident that the content representativeness effect is fairly robust and extends to television channels. Channel specialization was also hypothesized to encourage positive evaluations of the stories. Stories seen on specialist channels marginally encouraged a more positive viewer 2/21/08 12:05 AM Loading “EBSCOhost” Page 8 of 11 http://web.ebscohost.com/ehost/delivery?vid=8&hid=116&sid=0f3893a7-4628-4dbe-af98-f40fb68af4b1%40sessionmgr103 orientation to the stories than those seen on generalist channels. Undifferentiated channels were hypothesized to encourage higher story similarity ratings for participants who watched one channel than for participants who watched four channels. The data marginally support this hypothesis and are consistent with the notion of channel-asplace--that is, the mental representation of one channel showing 12 news stories is different from the mental representation of four channels showing three stories each. It appears that viewers discriminated less between stories when the stories were assigned to one channel number than between stories assigned to four channel numbers. This result is consistent with the analogical representation of numbers and suggests that the number line structure of television channels is one influence on how viewers judge program material. The criterion for judging story similarity appears to have been channel differentiation. The mean differences between the independent variables on both evaluation and similarity approached the conventional significance level, but it is likely that the limited number of participants in the study did not produce enough power for either analysis. Finally, channel differentiation significantly affected participant ratings of the physical attributes of the television picture. Although no hypotheses were made regarding picture quality, these results fit well with the notion that technology that can perform more functions (has more channels) produces better products. It seems that the perceived quality of the technology is global in nature such that it extends beyond the content and to the physical picture. Previous studies that used television found a significant relationship between the quality of the audio presentation and the perceived quality of the picture, even when there was no real difference in the picture (Neuman, Crigler, & Bove, 1991; Reeves, Detenber, & Steuer, 1993; Reeves & Nass, 1996). It seems that we can add channel differentiation to the list of attributes that enhance subjective viewing experiences. Implications. The advantages of specialist channels on judgment and evaluation of television program content are beginning to become clear. The advantages are especially important for program producers as they anticipate a huge channel environment. These findings might be generalizable to other communication technologies as well. Educational computer software or on-line newspapers, for example, could utilize the advantages of specialization via desktop environments or specialized agents. The results for similarity may have profound implications for the new multi-channel environment. Given that some sort of interface will be required for viewers to access individual programs, the extent to which these interfaces enhance or inhibit a sense of place where programs exist will impact viewers' judgments about how similar or dissimilar the program content is. For example, an interface that has many decision points along the way to accessing a program may enhance a sense of many places, and hence encourage perceptions of dissimilarity between programs. Compare that scenario to a voice-recognition television set where the set displays a program when given a simple verbal instruction. This might discourage a sense of many places--or encourage a sense of only one place--thereby increasing similarity perceptions among programs. In the latter case, the tendency to discriminate between program genres may decrease. Further research should examine the extent to which the similarity effect found in this study applies to between-genre judgments. To multimedia designers and others who think about a new media environment, this study suggests the utility of place, specifically the location of content in relationship to other content. Within that place, it also suggests the utility of specialization, based on a clearly defined role. Even though the benefits (and costs) of such a conception need to be explored in much greater detail, such issues may be as important to the user or viewer as are more common concerns, such as ease of operation or breadth of technical ability. Notes 1 The "sense of place" we refer to here can be distinguished from the "sense of place" Meyrowitz (1985) explicated. His "sense of place" refers to social situations and their effects 2/21/08 12:05 AM Loading “EBSCOhost” Page 9 of 11 http://web.ebscohost.com/ehost/delivery?vid=8&hid=116&sid=0f3893a7-4628-4dbe-af98-f40fb68af4b1%40sessionmgr103 on behavior. Our "sense of place" refers to psychological representations and their effects on behavior. 2 Most participants thought the experiment manipulated story types based on either the content of the stories or on production values. 3 A "wipe" is a video transition device where one picture seems to push the other off the screen. In this study, the wipe was stationary, so that the two video images--news story and graphic--were on the screen simultaneously during the story's duration. Table 1. Mean Ratings of News Attributes and Story Evaluation as a Function of Specialist or Generalist Channels