Increasing the audience for educational telivision

Increasing the Audience Educational Television 1

for

PETER CLARKE 2

Although the promise of educational television seemed large a few years ago, numerous audience studies have since revealed that this new medium appeals mainly to people who are already relatively educated. This finding has proved disappointing to those interested in promoting wider general education in the interest of a more informed public, a

Educational television was also envisioned as a formal teach- ing mechanism; school children are exposed daily to lessons in geography, languages, and a variety of other subjects. However, the impact of evening educational TV broadcasts on youngsters of school age is less dear. The experiment reported here was an effort to discover whether children (in this case, tenth graders) could be attracted to some of this evening programing, and if,

This study was supported by a grant from the Office of Education, U.S. Department of Health, Education, and Welfare, under the provisions of Title VII of the National Defense Education Act.

2 The author appreciates the cooperation extended by the participating teachers and Minneapolis Public Schools officials.

The most recent and comprehensive studies of the ETV audience are found in: Schramm, Wilbur; Lyle, Jack; and Pool, Ithiel. The People Look at Educational Television. Stanford, Calif.: Stanford University Press, ~L963.

(Footnote continued on p. x84)

Peter Clarke is on the faculty of the School of Communications, University of Washington, Seattle.

~83

AV COMMUNICATION REVIEW : 18 4

indirecdy, their parents might be persuaded to join the audience tOO.

The experimental manipulations used in this study were designed to increase the audience for an evening ETV program focused on current events. The broadcast ("World Affairs," on Channel 2, Minneapolis-St. Paul) is produced specifically for children. 4 This study also sought to determine whether increased exposure to the "World Affairs" program would generalize to other ETV offerings.

Youngsters were not the only viewers of concern, however. The social context of family TV viewing makes it clear that persistent changes in children's audience behavior are more likely if parents can be attracted to reality-content programs too. 5

Preliminary audience research conducted among families with teen-agers revealed that television program viewing is a family affair; for example, 6 out of i o parents in one study reported they had watched their favorite program the evening before with their is-year-old child. Joint viewing was the rule whether the program was a fantasy or a reality offering. One-fourth of the parents also reported they accepted program-selection advice from their youngsters; again, there was no difference between reality and fantasy program viewers.

However, different types of programs do possess varying potentials for subsequent social interaction or, as Katz and Foulkes

(Continued from p. z83) See also: Schramm, Wilbur, editor. The Impact of Educational Television. Urbana: University of Illinois Press, ~96o; and: Carter, Roy E., Jr., and Troldahl, Verling C. "The Size and Composition of the Adult Evening Au- dience for Educational Television in the Twin Cities," Report No. 3- Studies of the Adult Evening Audience for Educational Television in the Twin Cities. Minneapolis: Communication Research Division, School of Journal- ism, University of Minnesota, ~96~.

"World Affairs" is a locally produced panel show broadcast Monday evenings at 9:3 o (opposite "Ben Casey"). The format is a newsman-moder- ator plus two or three guest experts discussing a different topic each week. During the final so minutes, a high school student usually joins the group to ask questions.

5 The label reality content has been adopted from: Schramm, Wilbur; Lyle, Jack; and Parker, Edwin. Television in the Lives of Our Children. Stanford, Calif.: Stanford University Press, z96x. pp. 63-64. These researchers distin- guish between reality and fantasy content by suggesting that the former usually refers the viewers to problems of the real world, requires effort, works chiefly through realistic materials, and offers enlightenment On the other hand, fantasy programs invite surrender and passivity, they play on emotions, attract interest by abrogating the rules of the real world, and offer pleasure through wish fulfillment and anxiety reduction.

INCREASING THE AUDIENCE EOR ETV : I 8 5

METHOD

(4) might put it, varying potentials for feedback to social roles. Reality viewers were more likely to talk with others (including their children) about "their favorite program last night." In addition, when tenth-grade children are queried, they frequently recount instances of talking with their parents about reality offerings on television.

Children normally find interaction with adult authority figures of higher status to be a rewarding experience; one experimental manipulation in this study was designed to provide these rewards to children for viewing the current program on ETV.

Fourteen volunteer social studies teachers in Minneapolis were asked to conduct class discussions each Monday for four weeks. The current events topic to be featured on that night's ETV program was introduced with special materials; teachers advised students that a class discussion would follow the next day. View- ing ass ignments were not made, and students were told they would not be tested on the material. However, the ETV program was recommended as a source of ideas for students to contribute to the discussion on Tuesday, and teachers were to maximize student interaction during these discussions. 6

A second, more pedestrian technique aimed at increasing the ETV audience was utilized. Direct-mail brochures were sent to some of the children at home, advertising each coming broadcast scheduled for the experiment. 7 Twenty-two out of 6o social studies teachers in Minneapolis high schools agreed to participate in the study. One class was selected randomly from classes taught by each participant. These classes (and teachers) were assigned by probability methods to the four experimental conditions--control group (no audience- increase attempt), discussion method alone, direct-mail method

e Tenth-grade children (mostly xs-year-olds ) and their parents were studied in this audience-increase experiment because evening ETV content about public affairs could easily be made relevant to class activities.

T A n audience-increase experiment using direct-mail stimuli has been reported by Goldstein and his colleagues. They achieved significant viewing increases, but the effect was due partly to a decline in exposure among control subjects. Other data also suggest that the pre-experimental interview may actually have been part of the experimental treatment. See: Goldstein, M. N., and Shepherd, J. R. An Experiment in Increasing the Audience for Educa- tional Television, Preliminary Report No. 5. Eugene, Oreg.: Institute for Community Studies, I963. In a much different vein, Parker has examined changes in ETV viewing by employing the principle of stimulus generaliza- tion. See: Parker, Edwin B. Increasing the Audience for Educational Tele- vision. Urbana, Ill.: Institute of Communications Research, i96o.

AV COMMUNICATION REVIEW : ~86

RESULTS

alone, and discussion combined with mail. Classes in a single school were assigned to the same treatment combination so that comparisons between methods would not be blurred if students exchanged notes on the way their social studies classes were being conducted, s

Approximately 2 3 students in each class were chosen randomly to be interviewed. One parent per family was interviewed as well since part of the experimental design called for charting the flow of child-to-parent influence in media behavior. Which parent was interviewed was determined by a quota system based on interviewer selection--in which half the girls and half the boys were interviewed with their fathers, and the remainder with their mothers. As a result, parents induded in the sample were those most likely to be home--a disadvantage in terms of sample bias, but an advantage in scheduling interviews for subsequent waves since the same parents were interviewed throughout the study.

Exposure to ETV and a variety of other variables were measured during three interview waves--immediately before experimental manipulations, immediately after the experiment, and again two months later. Briefly, the data-gathering strategy for each wave included separate child and adult questionnaires. Upon entering a respondent's home, the interviewer handed the des- ignated child a self-administered ballot, and he was sent to another room to complete it unaided while the parent was being interviewed. ~

Exposure to educational television was measured with a set of items developed by Carter and Troldahl (2). This battery asks for recency and frequency of viewing. Respondents who claim some exposure to ETV are asked a program recall item seeking the name of a broadcast, who was on it, or what it was about. 1~ The effects of these attempts to increase the ETV audience can be estimated for both children and parents by examining differences in January-March and January-May exposure levels. The first

8 Social studies is a required subject at the tenth-grade level. * Assigned interview completion rates during each of the waves exceeded

9 ~ percent. A more adequate discussion of experimental procedures can be found in the author's dittoed version of "An Experiment to Increase the Audience for Educational Television."

xo The recall criterion was included to reduce prestige response bias. Pro- gram mentions were scrutinized to make certain respondents were referring to evening broadcasts; vague program mentions, such as "world events," "sports," or "some sort of art," did not qualify respondents as viewers.

A qualified viewer is one who claimed to have watched ETV and who could name a program or describe it in detail.

INCREASING THE AUDIENCE FOR ETV : z8 7

test provides information concerning the immediate impact of experimental manipulations; the second test tells how persistent were changes in audience behavior.

Table I contains the analysis of differences between children's ETV exposure during January and March and mean percentage changes for the four treatment groups, xl

TABLE I

Changes in Percentage of

Qualified Viewers among

Children between

January and March*

Direct Mail

Discussion Absent Present

Present 1% 22% (4) (9)

Absent 0% 15% (4) (5)

Source d.f . M.S. F P

Discussion 1 .5714 5.92 Direct Mail 1 .0364 NS Disc. X Mail 1 .0002 NS Residual 18 .0965

< .05

Total 21

* Figures in parentheses are treatment-group Ns.

The discussion (social reward) method was successful in increasing the number of qualified viewers among children; a more potent technique was the combination of discussion at school and literature mailed to the children at home. In addition, teachers displayed wide individual differences in their ability to create rewarding conditions for viewing "World Affairs." Within the same school, for example, percentage gains in ETV viewing ranged from a 6 percent loss to a 55 percent increase, from z percent to 88 percent gains, and from 6 percent to 57 percent increases.

In a period of four weeks, then, significant numbers of children were introduced (or reintroduced) to evening educational television via a series of public affairs broadcasts. However, it is important to ascertain whether or not this behavioral change

n Analyses of variance were computed employing the zx degrees of free- dom afforded by the 22 social studies classes. Changes between interview waves in percentage of qualified viewers in each class have been normalized by arc-sine transformations.

ArC COMMUNICATION REVIEW : 1 8 8

TABLE 2

Recency of

continued. This was the purpose of the interviews in May, conducted two months after the experimental treatments and approximately six weeks after "World Affairs" concluded its season.

The children's ETV viewing at this later date was slightly greater than in March; the F-ratio for the discussion method equalled 6.88 (p~.o25). However, this analysis is again based on percentage of qualified viewers.

Accordingly, a second analysis of January-May exposure differences was performed, employing the recency-of-viewing question. The item asked: "About when was the last time you saw a program during the evening hours---within the last week, within the last month, or longer than a month ago?" This item defined four groups of children during each interview wave; the first three groups are qualified viewers (they could name a program), and the fourth group is made up of nonviewers. Table a shows the sample distribution during January, March, and May.

January March May

Viewing among Children

Viewed within last week Viewed within last month Viewed longer than month ago Could not name program

1% 12% 3% 9 14 8

13 12 30 77 62 59 lOO% lOO% lOO%

N -" 510 425 281

One can see that the apparent maintenance of gains in ETV exposure for two months after the experiment is largely the result of children who recalled programs they had seen earlier than April. In May, x1 percent of the tenth graders reported they had watched Channel 2 within the last month; before the experiment this figure was xo percent.

In order to test whether program-selection influence on children diffused to their parents, adult ETV (Channel 2) viewing behavior was analyzed. Still employing dasses as experimental units, percentage changes in qualified viewers among parents were computed and subjected to statistical examination. There were no significant gains in exposure between January and March or between January and May.

Parents' responses were studied more closely to learn whether any increases in watching "World Affairs" could be found. Be-

INCREASING THE AUDIENCE FOR ETV : 1 8 9

TABLE 3

Changes in Percentage of

Parents between

January and March Who

Mentioned "World Affairs"

Program

fore the experiment, 3 out of 51o adults mentioned the broadcast when asked to describe a recent ETV offering they had watched; during the March interview, 32 out of 425 parents mentioned the program. Within each class, January-March percentage changes were obtained for numbers of parents who mentioned "World Affairs," and these data were analyzed in the same way as children's data. Table 3 presents the variance estimates and mean percentage gains under the four treatment combinations.

Discussion Absent Present

Direct Mail

Present 3% 14% (4) (9)

Absent 2% 0% (4) (5)

Source d.f . M.S. F P

Discussion 1 .0418 NS Direct Mail 1 .1209 7.70 <.025 Disc. X Mail 1 .1454 9.26 <.01 Residual 18 .0157

Further Analysis of

Children Who Changed

Their Viewing Behavior

Total 21

Statistically significant changes were found for the experimental program, but this time the direct-mail technique displayed the only main effect. The interpretation of this outcome must be tempered, however, by recognizing the significant interaction. It was the combination of promotional mail (which the parents must have read) and children's desire to view ETV for class pur- poses that changed parents' Channel z viewing. The only children attracted to the ETV audience were students in discussion classes. Initially, 25 percent were qualified viewers, but after the experiment in March, viewing had doubled to 49 percent of the sample. Of particular interest are the children who were nonviewers in January (217 out of 29t in discussion classes). Many of these youngsters remained outside the ETV audience through March, but 82 said they had viewed educational television during the interim and provided sufficient evidence to back their claims, m

Boys were as likely as girls to join the ETV audience.

AV COMMUNICATION REVIEW : I 9 0

Whether or not these 227 tenth graders sought the social rewards attached to ETV viewing is highly correlated with the socioeconomic status characteristic of their families. This status dimension is based on chief wage earner's occupation and educational attainment.

Table 4 shows that children in high-status families (at least one college-educated parent) were twice as likely to start watching Channel 2 as those in the lowest-status falnilies. 18 This comes as no surprise to any student of TV program preferences. One expects to find more social reinforcement for reality-content viewing in families with parents likely to have intellectual inter- ests and capabilities.

TABLE 4

Socioeconomic Status and

Socioeconomic Status Low Medium High

Changes in ETV Viewing

Remained nonviewers Joined ETV audience

71% 65% 40% 29 35 60

~oo% aoo% 100% N=63 119 35

Of course, the concept of social status is shorthand for a variety of distinctive outlooks and behaviors. Child-rearing practices are among these, and Bronfenbrenner (2) has provided an analytic summary of the literature in this area. In addition, Mac- coby (7) has studied the relationships between social class posi- tion, parental socialization practices, and television viewing habits among kindergarten children.

Rather than socialization practices, the present study chose to focus on socialization values as an explanatory variable. This strategy follows Kohn's (5, 6) approach to the study of ways in which parents raise their children. A value is defined as a con- ception by parents of what is desirable in a child and as a factor influencing the choice of behavior thought to encourage that value. Parents in the present study were confronted with a list of Kohn's desirable traits, among which were two of particular in-

~8 Chi-square equals xo.4, 2 d.f.; p<.oi. One benefit of higher disposable income was not associated with ETV viewing changes; whether a family owned one or two or more operating TV sets was not related to the tenth grader's decision to watch or not watch Channel 2.

INCREASING THE AUDIENCE FOR ETV : "r9I

terest--"to be curious about things" and "to do well in school." Respondents were told:

Here is a list of things that have been mentioned by parents as being desirable in boys (girls) of - - age. Of course, parents disagree on how desirable each of these things is. Some parents rate a thing as most desirable, while other parents believe it is somewhat desirable.

Parents were asked to pick the three most important traits from a list of nine. Among these parents, about one-half thought "doing well in school" is one of the most desirable qualities a tenth grader should exhibit, while one-fifth of the parents accorded equal importance to "being curious."

The relative rankings of these attributes will be used to signify parents' socialization emphasis with respect to the intellectual development of their children. Parents were classified according to the typological scheme below as to whether they rated each trait of first or less importance.

Typological Label "To Do Well "To Be Curious Applied to Parents: in School" about Things"

Child-centered Less important First choice Parent-centered First choice Less important High intellectual concern First choice First choice Low intellectual concern Less important Less important

TABLE 5

Parental Child Values and Children's

High Low Child- P a r e n t - intellectual intellectual

centered cen tered c o n c e r n concern

ETV Exposure Remained nonviewers 56% 74% 56% 53%

Joined ETV audience 44 26 44 47

100% 100% "100% 100%

N = 25 96 18 77

Table 5 shows the relationship between the child-value typol- ogy and Channel 2 viewing increases; changes in ETV exposure were found equally among all groups of families, except those with parent-centered adults. 14 Among these families, only z6 percent of the children joined the ETV audience between January

14 Chi-square equals 9.9, 3 d.f.; p~.o2.

AV COMMUNICATION REVIEW : 192

DISCUSSION AND SUMMARY

and March. The barrier represented by this set of socialization values is given added weight by the fact that 44 percent of the families fell in this category.

Children in this experiment were more likdy to join the ETV audience if their parents were also qualified viewers. In families where at least one parent watched Channel 2, 54 percent of the tenth graders were drawn into the audience; but among families in which the interviewed parent did not watch, only 26 percent of the youngsters started watching ETV (p<,.ooz). These data support an earlier finding by Schramm and his colleagues (8) that "family influences make really spectacular differences . . . in some of the fringe behavior such as viewing educational television."

A closer look at the present data dramatizes this point more fully. Youngsters were asked how recently and how frequently they talked with their parents about TV in general. Families were classified into groups with high and low amounts of reported interaction between tenth graders and parents.

When ETV-viewing families are examined separately from those in which the parent does not watch Channd 2, the influence of social rewards (via social interaction) on ETV viewing is clear- ly seen. As Table 6 shows, social interaction is positively correlated with changes in children's ETV behavior among the viewer families, but interaction is negatively correlated among the nonviewer families. 15

The combined effects of parental ETV exposure and social interaction about television in general are quite marked. Almost 7 out of Io children who talk a great deal about TV with parents who watch Channel 2 began viewing the medium during Feb- ruary; however, this behavior change was cut to 2 out of Io among children who interact frequently with parents who do not view ETV. Of course, discussions about TV between children and nonviewing (ETV) parents are focused on programs compet- ing with Channel 2. Increased ETV viewing persisted among children as long as direct social rewards were available. When these experimentally induced

Chi-square for qualified viewer families (parents) equals 3.6, I d.f.; p~.zo. Chi-square for nonviewer families equals 6.7, x d.f.; p<~.ox. Ahhough the first test did not reach the .o 5 level, the interaction (statistical) between parental behavior and extent of discussion about "IV is significant.

Parents were classified as qualified viewers if they satisfied exposure cri- teria in January or in March.

INCREASING THE AUDIENCE FOR ETV : ~ 9 3

TABLE 6

Extent of TV Interaction,

Parental ETV Parents: Children:

Extent of TV Interaction with Parents

Low High

Behavior, and Changes in

Children's Viewing of Channel 2

Qualified viewers Remained nonviewers Joined ETV audience

Not qualified viewers Remained nonviewers Joined ETV audience

54% 32% 46 68

100% 100% (56) (31)

58% 80% 42 20

100% 106% (38) (90)

rewards were withdrawn (after four weeks), youngsters reverted to their pre-experiment television habits; program interest did not generalize to other educational television broadcasts.

The children's behavior is understandable, given the lack of ETV viewing increases among parents. The experiment failed to generate a set of social rewards for viewing within the family, and, contrary to expectations based on family viewing of other reality content on TV, parents and children did not watch "World Affairs" togetheruat least not in large numbers.

Perhaps parents thought "World Affairs" was only for high school students and inappropriate for adult viewing. It is quite possible that program recommendations from teachers can make reality content too educational for its own good. However, another possible explanation is that intrinsic program quality may have been too unattractive without some direct inducements. Certainly, this was the participating teachers' reaction as adult audience members. Most of them thought the program was dull and poorly moderated, and they were pessimistic about their classroom efforts to encourage viewing.

Of course, teacher ability is another important and unanalyzed variable in this experiment. Whether or not the social-reward method has practical utility remains unknown for situations in- volving teachers unlike those who were willing to participate in this study. One suspects that volunteering to take part in research is correlated with a broad range of personality and career in- volvement attributes.

However, the findings from this limited experiment do form a pattern. Children who responded to the experience at school were those most likely to receive additional reinforcement at

AV COMMUNICATION REVIEW : I 9 4

REFERENCES

home. For example, youngsters who joined the ETV audience were, predominandy, members of high socioeconomic families or had parents who viewed educational television. And among ETV families, the frequency of parental social rewards for television viewing in general was positively correlated with the probability that children would be encouraged to watch "World Af- fairs."

Children's viewing behavior was related to their parents' pri- orities in child-rearing values as well. Youngsters whose parents placed more importance on "doing well in school" than on "being curious about things" were less likely to respond to social rewards at school and to watch ETV.

The question of whether or not levels of taste in mass media consumption can be elevated has remained a lively topic throughout the history of communication research (3). The findings of the present study should underscore some of the difficulties one faces in seeking audience behavior changes. It is probable that stimuli for viewing changes should be administered over a much longer period of time than used here. In addition, the social system in which the target audience behaves must be capable of dispensing rewards for the desired behavior--in this case ETV viewing.

This reward capability depends on predisposing factors, such as parental education and value orientations, and on participation in the desired activity. If parents are not in the ETV audience-- naturally, or as a result of explicit 16 experimental manipulations --they are unlikely to provide their children with the appropriate rewards for watching the medium.

Throughout, this research has focused on reinforcements for viewing aside from whatever rewards might be yielded by program content. Reactions by the teachers in this experiment suggest that ETV programers need to continue their concern with this latter problem.

I . Bronfenbrenner, Urie. "Socialization and Social Class Through Time and Space." Readings in Social Psychology (edited by Eleanor Maccoby, et al.). New York: Henry Holt and Co., I958.

~o Explicit is used here to emphasize a shortcoming of the present study, which relied on natural family viewing patterns. One economical way for teachers to reach parents with the "watch ETV" message might be to encourage children to talk with their parents about the program and its contents before attending class discussions the following day.

INCREASING THE AUDIENCE FOR ETV : "[95

2. Carter, Roy E., Jr., and Troldahl, Verling C. "Use of a Recall Cri- terion in Measuring the Educational Television Audience." Pub- lic Opinion Quarterly 26:~4-21; ~962.

3. Himmelweit, Hilde T. "An Experimental Study of Taste Develop- ment in Children." Television and Human Behavior (edited by Leon Arons and Mark A. May). New York: Appleton-Century- Crofts, ~963.

4. Katz, Elihu, and Foulkes, David. "On the Use of the Mass Media As 'Escape': Clarification of a Concept." Public Opinion Quarterly 26:377-88; 2962.

5. Kohn, Melvin L. "Social Class and Parental Values." American Journal of Sociology 64:337-5I; ~959.

6. . "Sodal Class and Parent-Child Relationships: An Inter- pretation." American Journal of Sociology 68:47~-8o; ~963.

7. Maccoby, Eleanor E. "Why Do Children Watch Television." Pub- lic Opinion Quarterly I8:239-44; I954.

8. Schramm, Wilbur; Lyle, Jack; and Parker, Edwin. Television in the Lives of Our Children. Stanford, Calif.: Stanford University Press, �9 96z. PP. 47-49.

A POEM Some time will you tell me why A Latin "a" plurals "ae," And why we honor this with awe As an intellectual law.

Give it class, give it tone! English plurals, don't condone! Alumna, alumnae, Even encyclopediae.

Make it rich, smooth like satin. Status plurals--always Latin. No doubt I'll use before I die Some instructional mediae.

- - C . J. WALLINGTON NDEA Fellow in Instructional Technology University of Southern California

Documents

Increasing the audience for educational telivision