If College Students Are Appointment Television Viewers

This article was downloaded by: [University of Wisconsin - Madison]On: 21 March 2013, At: 06:55Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Journal of Broadcasting & Electronic MediaPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/hbem20

If College Students Are Appointment TelevisionViewers ...Suzanne Pingree , Robert P. Hawkins , Jacqueline C. Bush Hitchon , Eileen Gilligan , BarryRadler , LeeAnn Kahlor , Bradley Gorham , Gudbjörg Hildur Kolbeins, , Toni Schmidt &Prathana KannaovakunVersion of record first published: 07 Jun 2010.

To cite this article: Suzanne Pingree , Robert P. Hawkins , Jacqueline C. Bush Hitchon , Eileen Gilligan , Barry Radler , LeeAnnKahlor , Bradley Gorham , Gudbjörg Hildur Kolbeins, , Toni Schmidt & Prathana Kannaovakun (2001): If College Students AreAppointment Television Viewers ..., Journal of Broadcasting & Electronic Media, 45:3, 446-463

To link to this article: http://dx.doi.org/10.1207/s15506878jobem4503_5

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form toanyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses shouldbe independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly inconnection with or arising out of the use of this material.

If College Students Are Appointment Television Viewers . . .

Suzanne Pingree, Robert P. Hawkins, Jacqueline C. Bush Hitchon, Eileen Gilligan, Barry Radler,

LeeAnn Kahlor, Bradley Gorham, Gudbjorg Hildur Kolbeins, Toni Schmidt, and Prathana Kannaovakun

Whether television viewers are selective or passive has generally drawn its findings from two non-overlapping research traditions. Research showing little audience selectivity in aggregite audiences may stem from how the aggregate is defined-an idea pursued here with an examination of

Suzanne Pingree (Ph.D., Stanford University, 1975) is a Professor of Life Sciences Communication at the University of Wisconsin-Madison. Her research interests are cognitive processes in media use and effects, especially in television and on the World Wide Web.

Robert P. Hawkins (Ph.D., Stanford University, 1974) is a Professor in the School of journalism and Mass Communication at the University of Wisconsin-Madison. His research interests are cognitive processes in mass communication and the use of new technologies for health communication.

lacqueline C. Bush Hitchon (Ph.D., University of Wisconsin-Madison, 1991) is an Associate Professor of Life Sciences Communication and Human Ecology at the University of Wisconsin-Madison. Her research interests include the psychological effects of marketing communication and its social implications.

Eileen Gilligan (Ph.D. Candidate, University of Wisconsin-Madison) is an Assistant Professor in the Depart- ment of journalism and Technical Communication at Colorado State University. Her research interests include news production and reporters' processes and cognitive processes of media effects.

Barry Radler (Ph.D., University of Wisconsin-Madison) is a Senior Marketing Specialist for the Division of Information Technology at the University of Wisconsin-Madison. His research interests are the cognitive determinants of persuasion and human information processing.

LeeAnn Kahlor (Ph.D. Candidate, University of Wisconsin-Madison) is a graduate student in the School of Journalism at the University of Wisconsin-Madison. Her research interests are in health and environmental risk communication and information processing.

Bradley Gorham (Ph.D. Candidate, University of Wisconsin-Madison) is an Assistant Professor at the 5. 1. Newhouse School of Public Communications, Syracuse University. His research interests are the cognitive processes involved in interpreting media messages.

Gudbjorg Hildur Kolbeins (Ph.D., University of Wisconsin-Madison, 2000) is an Assistant Professor in the Faculty of Social Sciences at the University of Iceland. Dr. Kolbeins' research interests are the effects of televised violence on children.

Toni Schmidt (Ph.D. Candidate, University of Wisconsin-Madison) is a graduate student in communication at the University of Wisconsin-Madison with research interests in the persuasive effects of advertising messages used in social marketing campaigns.

Prathana Kannaovakun (Ph.D., University of Wisconsin-Madison, 1999) is a Lecturer in the Department of Western languages, Faculty of Humanities and Social Sciences, Prince of Songkla University-Paffani in Thailand. Dr. Kannaovakun's research interests are in political communication, public opinion and inter- cultural communication.

8 2001 Broadcast Education Association journal of Broadcasting & Electronic Media 45f3), 2001, pp. 446-463

446

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

Pingree et al./lF COLLEGE STUDENTS ARE APPOINTMENT TELEVISION VIEWERS. . . 447

college student viewing. Results from a week of viewing show evidence of both structural and individual determination of selection. These results suggest additional qualifications on use of student samples for communi- cation research.

Mass communication research and popular conceptions live with two opposing views of television audiences, each of which has considerable utility and research support. On the one hand, viewers are often characterized as essentially passive, reacting to and controlled by the content and scheduling structures they encounter (e.g., Gerbner, Gross, Morgan & Signorielli, 1986; Goodhardt, Ehrenberg & Collins, 1987; Kubey & Csikszentmihalyi, 1990). In contrast, viewers are just as often characterized as active, even obstinately idiosyncratic, in fulfilling their own indi- vidual needs and making their own interpretations (Dervin, 1980; Livingstone, 1991). Apparently, audience behavior i s complex and variable - both betM;een and within individuals and groups.

The active-passive opposition appears across a wide range of things viewers do: reacting emotionally, comprehending or constructing meaning, being affected by or making use of messages, etc., and devoting effort to each of these behaviors (see Biocca, 1988; Hawkins & Pingree, 1986; and Katz, 1996, for discussions). But at the heart of things are the selections or choices (or lack thereof) that produce viewing itself, which many writers have characterized as a two-stage process (Rouner, 1984; Webster & Wakshlag, 1983). Both in choosing to watch television instead of (or in favor of) doing something else, and in the choices of programs to watch, viewers have been characterized as both active and passive.

The two different pictures of audience choice come largely from different research traditions employing different methods. Research on aggregate audience flow - how the television audience behaves as a whole over evenings or weeks of viewing - shows that the “decision to use television is typically passive” and dependent on audience availability and programming structures (Webster & Wakshlag, 1983, p. 438; see also Goodhardt, Ehrenberg & Collins, 1987), and that television viewing seems a way to casually fill otherwise unoccupied time (Kubey, 1986). When it comes to program choices while viewing, the aggregate research indicates that staying with a channel (“loyalty”), repeat viewing and availability and inheritance effects are strong structural predictors of program choice among the viewing audi- ence (Barwise, Ehrenberg, & Goodhardt, 1982; Goodhardt, et al., 1987; Webster & Wakshlag, 1983; see also Webster & Phalen, 1997 for an overview and summary). Aggregate research has demonstrated that viewers manifest periods of continuous viewing, often staying on the same channel despite program change. Programmers try to use this by placing new programs in the “hammock” between two popular programs to build an audience on this inheritance effect.

In contrast, individual-level study of selective exposure (Zillmann & Bryant, 1985; 1994), uses and gratifications (Palmgreen, Wenner & Rosengren, 1985; Rubin,

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

448 Journal of Broadcasting & Electronic MedidSummer 2001

19941, or viewing styles (Hawkins, Pingree & Reynolds, 1991; Krotz & Hasebrink, 1998) often reports considerable audience activity in decisions about both whether and what to view. Other such research reports greater audience activity in content selection than in the decision to watch television, which can be a product of ritual or habit (Lometti & Addington, 1992; Rubin, 1984). Part of the reason for the contrasting findings may lie in the methods of the two approaches, an insight provided by Webster (Webster & Newton, 1988; Webster & Phalen, 1997). Individ- ual approaches do not generally measure the structural contextual factors of either the medium (e.g., program scheduling) or the individual (e.g., scheduled competing activities that determine availability to be in the television audience), instead measuring or manipulating only those individual characteristics such as gratifications sought or mood thought to drive preferences and choice. The aggregate approaches lend themselves far better to the study of structural processes, since their vantage point is ill-suited to measure the variety of small-scale psychological processes that would assume audience activity. For example, finding little average evidence of selection in the aggregate based on a researcher-defined genre does not justify an overall conclusion of nonselectivity, since individuals may be making selections using a variety of criteria or an idiosyncratic definition of genre. (Preston & Clair, 1994; see also Chaffee & Mutz, 1988 for a discussion of problems of inferred measures). Given this, Webster and Newton (1988) point out the differences be- tween and inherent biases of the two approaches, and call for researchers to pursue syntheses, recognizing that the two research traditions are complementary.

Perhaps the structural and individual characteristics of both medium and audience exert control over viewing (Webster & Phalen, 1997, p. 45). This approach suggests nothing about what the balance will be, however. It does appear that television viewing is often simply a thing people do with their leisure time, that the overall availability of people to watch television i s a key determinant of audience size, and that habit i s critical. With respect to the first-stage choice - the choice about whether to turn on the set - Rosenstein and Grant (1 997) suggest that the audience’s primary relationship with television may be with the medium itself, rather than with any specific channel or program. However, audience flow studies on transitioning and “seamlessness,” and on the ease of channel-changing with remote control devices also suggest that audiences may be more active and selective in exercising control over their viewing once the television has been turned on (Eastman, Neal- Lunsford and Riggs, 1995). Eastman, Newton, Riggs and Neal-Lunsford (1 997) go so far as to describe the relationship between programmers and audience as a ”tug of war” with each side attempting to exercise control.

In fact, some have suggested that the conclusions of the aggregate approach overall are vulnerable to technology and industry changes that free viewers from structural constraints and make it easier to exercise their preferences (e.g., remote controls, large numbers of channels, the Internet) (Litman & Kohl, 1992). In a media environment of many choices and easy access, audience fragmentation and polar- ization have the potential to make things like inheritance effects “a quaint historical

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

Pingree et al./lF COLLEGE STUDENTS ARE APPOINTMENT TELEVISION VIEWERS.. . 449

footnote” (Webster & Phalen, 1997, p. 99). Other researchers are skeptical, arguing that television is likely to remain a mass medium whose audience continues to follow patterns already observed (Barwise & Erhenberg, 1988). Research so far within rich cable environments suggests relatively little or slow change (Webster & Phalen, 1997).

Within the aggregate research approach, however, it may be possible to get a sense of the balance between active and passive aspects of program choice by subdividing the audience. If different segments are selective at different times or in different ways, analyses that simply combine them would smooth out the differences and produce a misleading average that would suggest nonselectivity. A direction that may produce effective segmenting is suggested by Reimer and Rosengren (1 990), who recommend that researchers examine how lifestyle affects television viewing behavior, because television viewing was an intermediary between lifestyle and values for their Swedish sample. “A person’s media use is connected to - and originates in - life-style . . . (And) the meaning of media use i s illuminated by taking its context into consideration” (p. 187). Social psychology employs varied perspec- tives on lifestyle. One such perspective, the “life course“ approach, uses cohort and age-graded analyses. Cohort analysis attempts to reflect historical and societal influences on people born during certain time periods and experiencing key events similarly (such as the Great Depression or the Vietnam War). In contrast, age-graded analysis employs “generalized divisions or categories, from childhood to old age,” (Elder & ORand, 1995, p. 457) through which all people pass; particular activities are expected to occur in that time period in one’s life.

One audience subgroup defined both by age and role is college students. College students are much studied in mass communication research, but few researchers have examined them in terms of audience flow. Of course, there is some individu- ally-focused research on television use that specifically examines college students’ television choices. For example, Rubin (1 985) reported both college students’ motives and use patterns for televised soap operas. O‘Keefe and Spetnagel (1 973) examined the patterns of news media use of college students, finding that males viewed more television and read more newspapers than females. More recently, as part of a uses and gratifications study, Vincent and Basil (1997) examined college students‘ news media use and found levels of television news viewing similar to O’Keefe and Spetnagel’s earlier findings - just under two hours per week for national and a similar amount per week for local news. But these and other studies have little comparability to aggregate studies on degree of selective viewing, so a study of college students’ aggregate viewing fills a potentially important gap. Furthermore, although college students are often used in academic communication research, standard measures of television audiences generally have problems cap- turing them: they live and view in housing situations largely overlooked by standard sampling methods, and the age range typical of college students is not specifically segregated by Nielsen reports.

Despite this neglect, a college student sample might well be an excellent subdi-

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

, 450 Journal of Broadcasting & Electronic MedidSummer 2001

vision of the population within which to examine selective viewing. Although returning adult and part-time students will provide some small variance in age style of life, a student population is typically quite homogeneous in age and in day-to-day lifestyle. Individual availability for viewing is certainly less tied to full-day routines of work and family than for the general population, since courses and part-time work structure briefer periods of non-availability. As a result, availability may occur at atypical times that do not necessarily correspond well to programming schedules. Furthermore, the student environment typically offers almost unparalleled opportu- nities for interactions with peers who share situations and interests. This offers students greater opportunity to approach television with these group interests being salient, rather than as isolated individuals. In addition, students generally have the independence from other generations to enact that shared approach. Given the above, it is quite possible that college students, even viewed as an aggregate, are less constrained by programming structures and their own availability than the general population, and therefore are likely to be more selective and active television viewers than the general population or even non-students of the same age.

There has been considerable debate lately over whether the widespread use of college students in communication research is appropriate (see Courtright, 1996; Sparks, 1995; and other articles in those two journal issues for a fuller discussion of the use of student samples in communication research). Despite acknowledging the convenience of student samples, critics argue that the unusual and homogeneous demographics of student samples, as well as the fact that the student lifestyle i s unlike normal life in many ways, makes generalization to the general population impossi- ble. Thus, they conclude, research on college students should generally be avoided. Defenders argue that college student samples are appropriate when the phenomena studied are believed to be similar for all people, when a research focus on internal validity only is justified, or when special characteristics of students make results from them more generally useful. Examining television selection by a student sample will also contribute to this debate because it can help researchers identify theory contexts in which it might be inappropriate to use them as subjects.

The study is a secondary analysis of data collected as part of a larger laboratory- based study of attention to television genres. A total of 731 freshmen and sopho- mores (64% women) enrolled in two semesters of an introductory communication class at a large midwestern university participated for class credit. Early in the semester (October, 1997 or February, 1998), all students in the class kept a diary of their television genre viewing for one week. These data were used as background for extra credit studies, but were also used in class to provide students information about their individual and collective viewing habits. Students were given an instruction sheet and a computerized form on which to record their data, and they were

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

Pingree et al./lF COLLEGE STUDENTS ARE APPOINTMENT TELEVISION VIEWERS. . . 451

instructed to record the genre of any television they viewed for each half hour between 6 a.m. and 2 a.m. for seven days. They were also instructed to leave blanks on the answer sheet for half-hours in which they did not watch .television.

Students were asked to classify the television content they watched into one of ten genres, using a classification scheme similar to content categorization schemes used in previous research (e.g. Potts, Dedmon & Halford, 1996; Preston & Clair, 1994; Rao, 1975; Tangney, 1988). Previous research from the professional literature on these “common sense” program types (Webster & Wakshlag, 1983) indicates that genre is a reasonably valid and reliable construct to use for categorizing television content (e.g. Frank, Becknell & Clokey, 1971; Gensch & Ranganathan, 1974; Rao, 1975). To aid reliability, students were also given examples of particular genres using shows the students were likely to be familiar with. The ten genres (and their examples) were: News, Sports, Soap Opera (e.g. Days of Our Lives, All My Children), Situation comedy (e.g. Friends, Seinfeld, The Simpsons), Drama (e.g. E.R., Party of Five, Star Trek Voyager, Walker Texas Ranger), News & Entertainment Magazines (e.g. Dateline NBC, 60 Minutes, Entertainment Tonight), Game & Talk Shows (e.g. Jeopardy, Late Night With David Letterman, The Rosie O’Donnell Show), Music Videos; Reality Television (e.g. Cops, America‘s Funniest Home Videos), and Other Television (e.g. QVC, Discovery Channel documentaries, A&E’s Biography, how-to shows, infomercials). Televised (and videotaped) movies were not included in this assignment because of the difficulty of assigning them to television-based genres. Thus, the overall television viewing figures may underestimate students‘ actual time spent in front of the television, although it i s important to note that this should not influence the genre-specific viewing figures. Note again that this method provided information about genres watched, not individual programs, to match the genre orientation of the laboratory study.

Because this large sample provided only diary data, more detailed information about sample characteristics is not available. However, we can report some char- acteristics of the 174 students who participated in the laboratory study. The sample was 67% women (quite similar to the 64% for the class overall). Almost all (94%) were 20 or younger, and nearly all were Caucasian (93%). Half (51 YO) described their family as “upper-middle class” (out of five choices, with ”middle class” next most frequent). They came from a greater range of environments than we had expected: 25% from big city suburbs, 24% from mid-size cities, 19% from small towns, 18% from small cities, and 13% from big cities (city size definitions were provided in the questionnaire). Three-quarters came from homes that received 26 or more television channels, and 86% reported that their family subscribed to a daily newspaper. Thus, the subsample (and probably the full sample as well) fits the stereotype of university student samples. They were extraordinarily homogeneous in age and race, predom- inantly middle and upper-middle class, exposed to relatively rich media environ- ments. They were diverse only in their places of residence.

That subsample also responded to three questions asking how well each of several statements about television viewing reflected their practice (using a five-point scale:

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

452 Journal of Broadcasting & Electronic Medidsummer 2001

not at all, a little, somewhat, a lot, very much "like me"). Changing channels in the middle of a program was fairly normally distributed across the five answers, with 36% saying it was "a lot" or "very much like me." Watching two programs at once was more unusual, with only 15% in the top two categories. But 55% said that knowing what program they would watch when they started viewing was "a lot" or "very much" like them. Thus, there is some suggestion that the students see themselves as at least moderately active and selective in their television viewing.

Results

For each day of the week, we constructed a trend chart to show the percentage of students viewing at each half hour over the day. When we examined these figures, they appeared to present three different patterns.

Figure 1 shows college students who reported they were in the viewing audience for each half hour Monday, Tuesday, and Friday. Very few students watched television during the morning (never more than 5%), but there was a sharp peak between noon and 1 p.m. (16% viewing) - with the sample almost entirely watching soap operas (All My Children and Days of Our Lives competed from 12-1 p.m.) - then back down again in the early afternoon with another peak (12-18% viewing) between 4 and 5 p.m. (this i s a talk show peak - Rosie and Oprah were both on from 4-5 p m ) . We suspect that Tuesday differed from Monday and Friday because of university course scheduling conventions that constituted an audience-

Figure 1 Ordinary Days

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

Pingree et al./lF COLLEGE STUDENTS ARE APPOINTMENT TELEVISION VIEWERS. . . 453

structural factor: 75-minute classes Tuesdays and Thursdays, most of which end by 3:30 p.m., and 50-minute classes Monday, Wednesday and Friday, some of which run later. Thus, more students were probably available to view at 4 p.m. Tuesday than on Monday or Wednesday. During the evening, viewing gradually built to a prime time peak (note that in the Midwest, prime time is between 7 and 10 pm.), with about 18% watching during the first two hours of prime time Monday and Tuesday, a 40% decline between 9 and 10 p.m., and a brief resurgence after local news at 10:30 p.m. Not surprisingly, fewer students watched prime time Friday, probably because of the competition with social events.

The viewing patterns for Wednesday and Thursday (Figure 2) show evenings when the students appeared to have favorite prime time programs. Viewing was very similar to Monday and Tuesday during the day (including the noon and 4 p.m. peaks), but prime time viewing shot up strikingly. On Wednesday, between 42% and 48% of the students watched during the 7 p.m. to 9 pm. period (corresponding with Beverly Hills 902 70 and Melrose Place), with viewing increasing from 14% in the audience at 6:30, and then returning to 13% at 9 p.m. After this period, the audience returned to the normal weekday pattern. Thursday‘s pattern was slightly different but perhaps even more dramatic: viewing went from 13% in the audience at 6:30 pm. to 40% at 7 p.m. (Friends), then it dropped precipitously to 24% at 7:30 p.m., but zoomed back up to 40% at 8 p.m. (Seinfeld). The audience declined again to 30% at 8:30 p.m., but rebounded again to 41 % at 9 p.rn. (Em. While this suggests that the students were only watching favorite programs, and doing something else during the

Figure 2 Viewing Peaks

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

454 Journal of Broadcasting & Electronic MedidSummer 2001

less-popular situation comedies aired in the "hammock" between hits, it should be noted that neither of the drops in the audience over this 7 pm. to 10 p.m. period on Thursday brought it as low as the largest prime-time audience on Monday, Tuesday, or Friday. At 10 p m , the audience was back to the normal weekday pattern.

The third distinctive pattern i s the weekend pattern presented in Figure 3. Both days began with extremely small audiences and had afternoon plateaus, with a large peak (averaging around 19%) early Sunday afternoon (heavily sports, due mostly to the fall half of the sample, when the local professional football team usually played at noon). As with Friday, the Saturday prime-time peak was small (only about 14%) and early, presumably as a result of social activities. In contrast, Sunday prime time began slowly, rising only from 12% to 15% at 6 p.m. (60 Minutes, Dateline), peaked briefly (25%) at 7 pm., and then stayed at 18% until 9 pm., when it followed the typical late-evening pattern.

As a first check on this weekend pattern, we separated the sample by gender (Table 1). Women watched more dramas and soap operas, while men watched more news and especially more sports. The overall difference in viewing between women and men (12.02 vs. 15.47 hours per week) was largely accounted for by men watching more sports. As a further examination, we separated total viewing by days of the week (not shown). Women and men viewed similar amounts until the weekend, when men viewed very much more, and this weekend viewing largely accounted for the 3.45-hour difference in total viewing time between men and women.

Figure 3 Weekend Viewing

SO%

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

Pingree et al./lF COLLEGE STUDENTS ARE APPOINTMENT TELEVISION VIEWERS . . . 455

Table 1 Means and Standard Deviations of Television Viewing Time

as Categorized by Genre

Whole Sample Female Male

Mean S.D. Mean S.D. Mean S.D. Sig. of Diff.

Total viewing Drama Sitcom News Magazine News Reality Television Soaps sports Game & Talk Show Video Others

12.96 9.10 12.02 7.85 15.47 10.59 ** 2.41 2.39 2.89 2.45 1.82 2.28 ** 2.68 2.79 2.63 2.80 3.09 2.94 0.36 0.93 0.41 0.96 0.34 1.00 0.68 1.37 0.63 1.33 0.87 1.61 * 0.15 0.50 0.17 0.56 0.12 0.34 1.02 1.77 1.35 1.93 0.48 1.36 ** 2.68 4.68 1.04 1.97 5.44 6.16 ** 1.38 2.00 1.49 1.86 1.44 2.24 0.82 1.87 0.78 1.96 0.92 1.75 0.77 1.69 0.71 1.73 0.96 1.77

Note: Means for viewing variables are in hours per week; *p < .05; **p < ,001.

Comparison with a Local Nielsen Sample

While much of the students’ viewing seemingly reflected times they were available to view (i.e., time not claimed by competing activities) or the medium structural factor of prime time, the Wednesday and Thursday pattern suggested something different and more selective. To explore this, we compared the student sample to ratings data (Nielsen, 1997). The Nielsen sample was 300 people (1 61 women and 139 men) in the 18-34-year-old age group in the area during May 1997. (The total sample size was 1,275 people from this Nielsen “designated market area.”). This is the closest age group Nielsen provides to the ages of our college student sample and it included 11 people living in college dormitories (although Nielsen says that they don’t use people in ”group living quarters” to make predictions).

Table 2 illustrates the collective focus of the two samples‘ viewing on Wednesday and Thursday evenings, when programs highly popular with the student age group were broadcast. The left pair of columns presents the percentage watching any television program at each half-hour, both for the student sample and the 18-34 age group in local Nielsen ratings, with significance of the difference in percentages tested by z-test. The right pair of columns reports the percentage watching particular programs popular at that time. However, the percentage reported for the student sample may be an over-estimate, because they were queried only about the genre they were viewing at any given time. That is, the 32% of students who watched dramas Wednesday at 7 p.m. were very probably watching Beverly Hills 902 10 (by

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

456 Journal of Broadcasting & Electronic Mediahummer 2001

Table 2 Comparing Viewing Patterns of College Students with a

local Nielsen Sample

Percent Viewing Percent Viewing Television Program/Genre

Nielsen Sample Students Nielsen Sample Students

Wednesday 7:OO p.m. 37% (Beverly Hills 902 7 0)

(Beverly Hills 902 7 0)

(Party of Five)

(Paw of Five)

(No dominant program)

(Friends)

(Union Square)

(Sein feld)

(Veronica's Closet)

Wednesday 7:30 p.m. 35%

Wednesday 8:OO p.m. 39%

Wednesday 8:30 p.m. 41 yo

Wednesday 9:OO p.m. 42%*

Thursday 7:OO p.m. 3 8%

Thursday 7:30 p.m. 35%*

Thursday 8:OO p.m. 44%

Thursday 8:30 p.m. 36%*

Thursday 9:00 p.m. (ER) 52%* Thursday 9:30 p.m. (ER) 52%*

Note: *p < .05

46%

48%

13%

40%

24%

40%

30%

41 '/o

38%

12%*

13%*

12%*

13%*

23%*

15%

25%*

1 1 Yo*

2 6% 2 8%

34%

18%

3 1 Yo

19%

2 7% 2 5 O/O

far the most popular program at that time for 18-34-year-olds), but The Sentinel was available first-run on UPN and a few other dramas were available as syndicated reruns. Only syndicated dramas were available at 8 p.m. to compete with Party of Five. On Thursday, Friends, Seinfeld, and E. R. were by far the most popular programs in their time slots among the 18-34 group, but there were Fox situation comedies competing with Friends and Union Square, and syndicated drama reruns at the same time as E.R. In other words, the program/genre percentages may be slight overesti- mates for the students, but they are probably st i l l quite close to the actual numbers.

For 7-9 p.m. Wednesday evening, the percentage of students viewing television was only slightly (and not significantly) higher than that of the age group as a whole. At 9 p.m., however, the students dropped away from television almost altogether (1 3% viewing), while 42% of the 18-34 Nielsen sample continued watching at roughly the same level. However, between 7 and 9 p.m. the student audience was

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

Pingree e t al./lF COLLEGE STUDENTS ARE APPOINTMENT TELEVISION VIEWERS. . . 457

far more concentrated on Beverly Hi l ls 90210 and Party of Five than the overall 18-34 audience. More than 75% of students watching television were apparently watching these programs, contrasting with about a third of the Nielsen group, even though these were the most popular programs at those times for this Nielsen sample. It is also worth pointing out that students dropped out of watching television at 9 p m . when there was no particularly popular dominant program to watch.

For the Thursday night NBC lineup of situation comedies and E.R., Table 2 shows a different pattern, but one that may make the same overall point. NBC attempted to build audiences for several new situation comedies in the "hammocks" after Friends, Seinfeld, and ER but was not very successful; viewing dropped off dramatically in each case both for students and the Nielsen sample. For three of the four comedies, significantly more students viewed than did the overall 18-34 age group. However, many of the students left television altogether for the 7:30 pm. half hour, so that significantly fewer of them viewed than of the larger age group. Thus, the overall viewing levels of the 18-34 group were steadier than the students' viewing levels. A similar but less dramatic pattern can be seen for Veronica's Closet at 8:30 p.m., when students were more likely to be watching that particular program than the Nielsen sample, but less likely to be in the audience at all. And at 9 and 9:30 p.m. the 18-34 age group was more likely to watch television than the students, but the percentages watching the most popular program, E.R., were essentially identical. Thus, on these two evenings, students' viewing appeared to be more concentrated on catching popular programs, and less on just watching television, than the viewing of the overall 18-34 age group.

Discussion

First, this study confirms previous findings (O'Keefe & Spetnagel, 1973; Vincent & Basil, 1997) that college students are lighter television viewers than the general population, or even other young adults. What is more interesting and important i s that aggregate examination of this group whose members generally share a lifestyle (Elder & O'Rand, 1995; Reimer & Rosengren, 1990) suggests both structural and individual ("active") determinants of television viewing. The university student lifestyle is one defined by life course: "age-graded life patterns embedded in social institutions and subject to historical change" (Elder & O'Rand, 1995, p. 453). The students in our data set are clearly part of a mainstream institution in our society, the state university, which has itself changed in recent history with the end of curfews and single-sex only dormitories. The physical location of television in students' environments has also changed. Twenty or 30 years ago, televisions were located in residence or union lounges, where students could gather to watch programs. Now most college dorm rooms contain at least one television, brought there by one of the residents, making casual, continuous viewing more possible than it used to be.

Students do show evidence of structural effects on their viewing (using Webster &

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

458 Journal of Broadcasting & Electronic Media/Summer 2001

Phalen's 1997 model of audiencehediurn, structural/individual predictors). They watched more during prime time than other parts of the day, although it i s difficult to determine whether this i s a result of audience structure (availability during less-scheduled evening time) or medium structure (the presence of prime-time programming). And the lure of prime time was apparently limited by weekend social activities on Friday and Saturday nights, a further consequence of audience avail- ability. However, it may be worth noting that the structural constraint of weekend social activities seems different than those posed by work or class schedules. The latter are typically fixed (whether by the individual or external actors) for relatively long periods of time, such as semesters, whereas weekend social activities are more varied and flexible. This somewhat blurs the boundary between audience availability and preference, since one could say that it i s individual preferences for a particular evening's social activities that decrease the weekend audience. Or one could say that, in the aggregate, social life on weekends has become so regularized as to constitute audience structure as much as work or school schedules. Programmers have responded to this structure with weekend schedules (medium structure) that assume the absence of young adults. Perhaps the best overall conclusion is that these dichotomies are difficult to maintain.

On the active side, student viewing also contains evidence of what might be called preference-based "appointments" with particular television programs, first in the regular hour-long weekday peaks for soap operas or talk shows. The fact that there is a viewing peak at noon could well be taken as availability at lunch time (few classes were scheduled between 12 and 1, many students return to their residence to eat lunch), and thus represent an effect of structure as well as preference for particular programs. However, the fact that almost all viewers watched one of two soap operas suggests the action of genre-related preferences, certainly in the choice of what to view, and perhaps in the initial decision about whether to view at all. The late-afternoon peak, again mostly concentrated in two talk shows, also suggests the operation of preference, but viewership is constrained on Monday, Wednesday, and Friday by course schedules that limit availability.

In addition, students seemed to have appointments with particular programs, since about twice as many watched prime time Wednesday and Thursday nights as on other evenings, and Sunday showed an unmistakable sports peak, particularly for men. Although the students reported what genre they watched, not what particular program, the limited numbers of other programs of the same genre makes this program attribution rather likely. It is also noteworthy that on these heavy-viewing evenings, students left the audience in large numbers when the popular programs ended. More than two-thirds of them left the audience at 9 p.m. Wednesday evening at the conclusion of two dramas targeted at young adults. Similarly, half of them stopped watching at 7:30 p.m. Thursday, and about a third stopped watching at 8:30 p m , thus skipping the less popular programs in the "hammocks" after Friends and Seinfeld. However, these latter two drops are a case of the glass being either half-full or half-empty, since they demonstrate some hammock structural effect as well as the

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

Pingree et al./lF COLLEGE STUDENTS ARE APPOINTMENT TELEVISION VIEWERS . . . 459

action of preference: as many or more students watched at low points (and they were sti l l predominantly watching the situation comedy genre) as watched at the peaks of prime time viewing on other weekdays.

These results provide enough information to allow us to characterize college students’ television viewing. Despite generally easy access to television, students clearly do not simply spend their free time casually viewing television. Television i s available to them, and they do respond to structural factors, being influenced both by their own schedules and television’s. But they also appear to make appointments with popular programs and genres for selective viewing. And unlike the “evening of viewing” characterization of the mass audience, college students then tune out to pursue some of the many other activities present in student life (such as homework, we hope).

These aspects of student viewing showed two important differences from the most-similar Nielsen data available - the age 18-34 cohort from the immediate geographical area. First, the Nielsen sample was much less likely than the students to leave television altogether during the two popular prime time evenings, as demonstrated both Wednesday at 9 p.m. and after the two most-popular Thursday situation comedies. By contrast, the proportion of the Nielsen sample viewing at each half hour was fairly steady, although it did increase dramatically for E.R. on Thursdays at 9 p.m. Second, although our confidence in this conclusion is limited by having genre data for students but program-specific numbers for the Nielsen sample, it appears that student viewing was much more concentrated than in the Nielsen sample on the most popular programs. Generally, more than 75% of the students viewing television seemed to be watching the most-popular program at the time period, whereas this was true only for a third to a half of the Nielsen sample.

This suggests that television viewing by those engaged in the student lifestyle (atypical as it is of adult life generally) may be more selective and less bound by both medium and personal structure than the general population. Student life often allows both a great deal of personal independence in media use (unconstrained by parents, partner, or dependents), and constant interaction with others of the same age with shared interests. All this occurs in an atmosphere of flexible and divisible time, in which class and work schedules vary from day to day, and often involve short commitments, rather than the 8-hour workday or the whole evening. Thus, the student lifestyle may in fact be the one most ideally suited to shared selective behaviors.

However, this may not be fair to the Nielsen 18-34 sample. As we noted earlier, aggregate examinations of the general population, or even of subsamples like this one, run a substantial risk of producing misleading averages of very different types of persons. This i s particularly true when seeking evidence of selective processes, since any “flattening“ of different kinds of selection would likely be taken as overall evidence of passive nonselection. It is instead likely that this Nielsen sample actually contains representatives of a number of different lifestyles (e.g., single worker living alone; single worker in a shared residence; married couple, both working, no

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

460 Journal of Broadcasting & Electronic MedidSurnmer 2001

children; married and working with preschool children, etc.). It could well be that within the Nielsen sample there are lifestyles that are each individually selective on some points of television viewing but structurally-determined on others. The lesson of this student sample is that researchers should address these differences instead of drawing active/passive conclusions from heterogeneous aggregates, and that I ifestyle would be a promising place to start (Reimer & Rosengren, 1990).

If college students are indeed sometimes appointment television viewers - if they view television as if they have "appointments" with specific programs so that they tune in based on what they wish to watch, rather than just watching to watch -then there are implications for researchers interested in television viewing variables who use student samples. Researchers often employ student samples, despite the diffi- culties in generalizing from them (for a rather wide-ranging debate, see Abel, 1996; Basil, 1996; Courtright, 1996; Lang, 1996; Potter, Cooper & Dupagne, 1993, 1995; Sparks, 1995). Despite some strong rhetoric at either end of the continuum in this debate, a developing consensus (replicating that developed earlier in other fields; Courtright, 1996), suggests that the researcher who uses student samples simply needs to carefully consider potential advantages and disadvantages and recognize the resulting limitations.

For example, one common argument i s that experiments can use college students if the processes studied are not unique or special to college students. Thus, in studies of processes of attitude formation or change (rather than of levels of attitudes), student samples are appropriate and resulting theory should be generalizable. Given the results here, caution is called for when experimental stimuli might have varying effects, depending on the activdpassive tendencies of experimental subjects. Survey researchers sometimes make a similar argument that their studies focus on relation- ships between variables, not on descriptions of general behaviors. This argument i s also sensible, although given the results of our study, it needs to be applied carefully and with attention to what these variables are.

However, when the relationships are between viewing and other variables, such as social effects, college students are probably not appropriate for several reasons. First, college students watch much less television than the general public. In cultivation studies, their viewing would be categorized as "light" (under two hours per day across a week). Hirsch (1 980) clearly demonstrated the problems of doing analyses of the relationship between television viewing and beliefs about social reality for light viewers - the relationships that result are quirky, unreliable, and often extremely difficult to interpret in the context of any theory. A critical but more subtle issue arises in studies of long-term processes, of which cultivation is again an example. In these studies, reports of current viewing are often used as indicators of a history of viewing amount and style. Unfortunately, college students' viewing behavior today is not likely to tell us much about their television viewing when they were growing up, or about their viewing when they leave the college lifestyle.

However, in contrast to these caveats, there may also be survey research situations where college students are a particularly appropriate sample. This is where college

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

Pingree e t al./lF COLLEGE STUDENTS ARE APPOINTMENT TELEVISION VIEWERS . . . 461

students themselves have some characteristic that is important for the conceptual issues of the research. One such situation occurs in studies of new communication technologies, such as the World Wide Web, where university settings provide a universal level of access not yet obtained by the general public (Ferguson & Perse, 2000; Authors, in press). Situations in which college students are a particularly suitable audience for an important communication issue, such as exposure to sexually-transmitted diseases or susceptibility to eating disorders, represent a second appropriate use of the student sample. The present study adds to this l ist the suggestion that researchers interested in viewers who watch television lightly or selectively might want to study college students because they constitute a hard-to- find subgroup with those characteristics.

Finally, this research also suggests that the two camps of researchers now need to seriously pursue two new takes on their old research agendas if the syntheses called for by Webster and Newton (1988) are actually to be achieved. Those doing aggregate research should continue to examine their traditional questions about inheritance, repeat viewing, genre and channel loyalty, etc., but should attempt to do so with subgrouped samples instead of the whole viewing audience. Given that a subgroup may be very selective some of the time, such research should provide a much more accurate picture of the balance between activity and passivity in television selection. In such research, the degree of subgrouping and the best method (e.g., combination of small demographic categories and lifestyles) will themselves constitute an additional issue to address.

Similarly, researchers studying individuals need to begin measuring the context of viewing decisions more directly, both for the medium structure that constrains choice and for the individual structure of work and other scheduled activities. They may well discover that typical survey measurement ("How often do you.. ."I is inadequate to the task; if so, a narrower focus on particular weeks or even days may allow much more accurate prediction. Taken together, the two research approaches should then be able to present a balanced picture of television viewing, instead of the irreconcilably conflicting one we now have.

References

Abel, R. (1 996). Can we generalize from Generation X? Not! lournal of Broadcasting and

Barwise, T. P. & Ehrenberg, A. S. C. (1 988). Television and its audience. London: Sage. Barwise, T. P., Ehrenberg, A. S. C., & Coodhardt, C. J. (1982) Clued to the box?: Patterns

of N repeat-viewing. /ourna/ of Communication, 32(4), 22-29. Basil, M. (1996). The use of student samples in communication research. journal of

Broadcasting and Electronic Media, 40(3), 43 1-440. Biocca, F. A. (1988). Opposing conceptions of the audience: The active and passive

hemispheres of mass communication theory. In J . A. Anderson (Ed.), Communication yearbook 71 (pp. 51-80), Newbury Park, CA: Sage.

Chaffee, S. & Mutz, D. (1988). Comparing mediated and interpersonal communication. In

Electronic Media, 40(3), 441-446. Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

462 Journal of Broadcasting & Electronic MedidSummer 2001

R. Hawkins, J. Wiemann, & S. Pingree (Eds.), Advancing communication science: Mergingmass and interpersonal. Thousand Oaks, CA: Sage, 1988.

Courtright, J. A. (1996). Rationally thinking about nonprobability. Journal of Broadcasting and Electronic Media, 40(3), 414-421.

Dervin, B. (1 980). Communication gaps and inequities: Moving toward a reconceptualiza- tion. In 8. Dervin & M. Voigt (Eds.), Progress in communication sciences, Vol. 2. Norwood, N.J.: Ablex.

Eastman, S. T., Newton, C. D., Riggs, K. E., & Neal-Lunsford, J. (1997). Accelerating the flow: A transition effect in programming theory!Journal of Broadcasting and Electronic Media,

Eastman, S. T., Neal-Lunsford, J., & Riggs, K. E. (1995). Coping with grazing: Prime-time strategies for accelerated program transitions. Journal of Broadcasting and Electronic Media, 39,

Elder, G. H. Jr., & O'Rand, A. M. (1995). Adult lives in a changing society. In K. S. Cook, C. A. Fine, & J. S. House (Eds.), Sociological perspectives on social psychology. Needham Heights, MA: Allyn and Bacon.

Ferguson, D. A. & Perse, E. M. (2000). The World Wide Web as a functional alternative to television. Journal of Broadcasting and Electronic Media, 44(2), 155-1 74.

Frank, R. E., Becknell, I., & Clokey, J. (1 971). Television program types. JournalofMarketing Research, 8, 204-21 1.

Gensch, D. H. & Ranganathan, B. (1974). Evaluation of television program content for the purposes of promotional segmentation. Journal of Marketing Research, 11, 390-398.

Gerbner, C., Gross, L., Morgan, M., & Signorielli, N. (1986). Living with television: The dynamics of the cultivation process. In Bryant, J. & Zillmann, D. (Eds.), Perspectives on media effects (pp. 17-40). Hillsdale, NJ: Lawrence Erlbaum Associates.

Goodhardt, C. J., Ehrenberg, A. S. C., & Collins, M. A. (1987). The television audience: Patterns o f viewing, an update. England: Cower House.

Hawkins, R. P. & Pingree, S. (1986). Activity in the effects of television on children. In J. Bryant and D. Zillmann (Eds.), Perspectives on Media Effects. Hillsdale, N.J.: Lawrence Erlbaum Associates.

.Hawkins, R. P., Pingree, S. & Reynolds, N. (1991). In search of viewing styles. journal of Broadcasting and Electronic Media, 35(3), 375-383.

Hirsch, P. M. (1980). The 'scary world' of the nonviewer and other anomalies: A reanalysis of Gerbner et al.'s findings of cultivation analysis. Communication Research, 7(4), 403-456.

Katz, E. (1996). Viewers work. In J. Hay, L. Crossberg, & E. Wartella (Eds.), The audience and its landscape (pp. 9-21). Boulder, CO: Westview Press.

Krotz, F. & Hasebrink, U. (1 998). The analysis of people-meter data: Individual patterns of viewing behavior and viewers' cultural backgrounds. Communications, 25(2), 151-1 74.

Kubey, R. (1986). Television use in everyday life: Coping with unstructured time. Journal of Communication, 36(3), 108-1 23.

Kubey, R. & Csikszentmihalyi, M. (1990). Television and the quality of life: How viewing shapes everyday experience. Hillsdale, N.J.: Lawrence Erlbaum Associates.

Lang, A. (1996). The logic of using inferential statistics with experimental data from nonprobability samples: Inspired by Cooper, Dupagne, Potter, and Sparks. Journal o f Broad- casting and Electronic Media, 4@3), 422-430.

Litman, B. R. & Kohl, L. S. (1992). Network rerun viewing in the age of new programming services. Journalism Quarterly, 69, 383-391.

Livingstone, S. (1 991). The resourceful reader: Interpreting television characters and narra- tives. Communication Yearbook, 15, 58-90.

Lometti, C. & Addington, T. (1 992, August). Testing opposing views of the audience: How the active television viewer selects programs. Paper presented to the Association for Education in Journalism and Mass Communication, Montreal, Canada.

A.C. Nielsen Company. (1 997). Nielsen Station lndex, Viewers in Profile. (Pamphlet) Madison, WI.

41(2), 265-283.

92-1 08.

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013

Pingree et al./lF COLLEGE STUDENTS ARE APPOINTMENT TELEVISION VIEWERS. . . 463

O’Keefe, G. J., & Spetnagel, H. T. (1 973). Patterns of college students‘ use of selected news media. lournalism Quarterly, SO, 543-548.

Palmgreen, P., Wenner L. A,, & Rosengren, K. E. (1 985). Uses and gratifications research: The past ten years. In Rosengren, Wenner & Palmgreen (Eds.), Media gratifications research: Current perspectives (pp. 11-40). Beverly Hills, CA., Sage.

Potter, W. J., Cooper, R., & Dupagne, M. (1993). The three paradigms of mass media research in mainstream communication. Communication Theory, 314), 31 7-335.

Potter, W. I., Cooper, R., & Dupagne, M. (1995). Reply to Sparks’ critique. Communication Theory, 5, 280-386.

Potts, R., Dedmon, A,, & Halford, J. (1 996). Sensation seeking, television viewing motives, and home television viewing patterns. Personality and lndividual Difference, 2 1, 1081 -1 084.

Preston, J. M. & Clair, S. A. (1 994). Selective viewing: Cognition, personality and television genres. British lournal of Social Psychology, 33,273-288.

Rao, V. R. (1 975). Taxonomy of television programs based on viewing behavior.]ournal of Marketing Research, 12, 355-358.

Reimer, B. & Rosengren, K. E. (1990). Cultivated viewers and readers: A life-style perspec- tive, In N. Signorielli & M. Morgan, (Eds.), Cultivation analysis: New directions in media effects research. Newbury Park, C A Sage.

Rosenstein, A. W. & Grant, A. E. (1997). Reconceptualizing the role of habit: A new model of television audience activity. lournal of Broadcasting and Electronic Media, 41, 324-344.

Rouner, D. (1 984). Active television viewing and the cultivation hypothesis. ]ournalism Quarterly, 61, 168-173.

’

Rubin, A. (1 984). Ritualized and instrumental television viewing. Iournal of Communica- tion, 34(3), 67-77.

Rubin, A. M. (1985). Uses of daytime television soap opera by college students. Journal of Broadcasting and Electronic Media, 29, 241-258.

Rubin, A. M. (1994). Media uses and effects: A uses-and-gratifications perspective. In. J. Bryant and D. Zillmann (Eds.), Media effects: Advances in theory and research (pp. 41 7-436). Hillsdale, N.J., Lawrence Erlbaum.

Sparks, G. (1995). Comments concerning the claim that mass media research i s ”prescien- tific”: A response to Potter, Cooper, and Dupagne. Communication Theory, 5, 273-280.

Tangney, J. P. (1988). Aspects of the family and children’s television viewing content preferences. Child Development, 59, 1070-1 079.

Vincent, R. C. & Basil, M. D. (1997). College students’ news gratifications, media use and current events knowledge. Journal of Broadcasting and Electronic Media, 41(3) 380-392.

Webster, J. G. & Newton, G. D. (1988). Structural determinants of the television news audience. lournal of Broadcasting and Electronic Media, 32(4), 381-389.

Webster, J. G. & Phalen, P. F. (1997). The mass audience: Rediscovering the dominant model. Mahwah, N.J.: Lawrence Earlbaum.

Webster, J. G. & Wakshlag, J . J. (1983). A theory of television program choice. Communi- cation Research, 10, 430-46.

Zillmann, D. & Bryant, J. (1985). Selective exposure to communication. Hillsdale, NJ: Lawrence Erlbaum.

Zillmann, D. & Bryant, J. (1994). Entertainment as media effect. In J. Bryant and D. Zillmann (Eds.), Media effects: Advances in theory and research (pp. 41 7-436). Hillsdale, NJ: Lawrence Erlbaum.

Dow

nloa

ded

by [

Uni

vers

ity o

f W

isco

nsin

- M

adis

on]

at 0

6:55

21

Mar

ch 2

013