Introduction 
The purpose of this study is to look at Greek adolescents' television viewing and how it relates to their attitudes toward the family and traditional relationships between parents and adolescent children. The approach used here is the cultivation hypothesis.

The cultivation hypothesis states that the more television people watch, the more likely they are to hold a view of reality that is closer to television's depiction of reality. This is characterized by the work of George Gerbner and his colleagues (Gerbner et al., 1979). Their work starts with the cultural indicators project, which looks at the content of television programming. It then relates it to the perceptions of heavy versus light viewers about a variety of subjects. Cultivation research has showed that amount of viewing is associated with images of violence, sex roles, aging, politics, health, religion, minorities, jobs, the environment and other topics (Morgan and Shanahan, 1997).

Television constitutes a shared cultural environment made up of images and representations with which people grow up (Morgan, Leggett and Shanahan, 1999). Generally, the beliefs of heavy television viewers about the real world are consistent with the repetitive and emphasized images and themes presented on television (Gerbner, Gross, Morgan, and Signorielli, 1994). As such, heavy viewing cultivates a television-shaped view of the world (Hawkins, Pingree and Alter, 1987).

Cultivation theory generally assumes that "light viewers tend to be exposed to more varied and diverse information sources, while heavy viewers, by definition, tend to rely more on TV" (Signorielli and Morgan, 1990, p. 17). But in other societies, where there has not been as great a need to attract large audiences as in the United States, television content may not be as homogenized.

In Greece, for example, there is diversity in programming, not only in genres but also in terms of national origin, as viewers have the opportunity to watch Greek, US, British, French and Brazilian programs. Greek television has historically had a good deal of US programs. Up until the late 1980s, when there were only two (public) Greek television channels, they averaged between 38 and 48 percent in imported programming, most of which came from the United States (Zaharopoulos, 1990). The introduction of private television in Greece in 1989 revolutionized the market. Initially, imports from the United States were the main source of programming for the new private stations, sometimes making up over 50 percent of their total programming. Slowly, however, as the major Greek stations increased their local production, the share of foreign programming per major station decreased (Zaharopoulos and Paraschos, 1993).

According to Morgan (1990), cultivation is highly culture specific. "The symbolic environment of any culture reveals social and institutional dynamics, and because it expresses social patterns it also cultivates them" (p. 226). When this approach is used to study US television overseas, cultivation predictions cannot be as certain. For example, in an Australian study of over 1000 students, Pingree and Hawkins (1981) found that watching violent US television programs was more related to conceptions of reality in Australia than the reality of the United States.

One of the topics television portrays is the family and how family members relate to each other. For example, each family member can be a source of comfort, love, support, hostility, and pain for each other.

Skill et al. (1987) looked at US television families and found that over 65% of programs presented families that were rather conventional or traditional. Moore (1992) in a similar, but more encompassing, study found a similar result.

Cantor (1990) found that the families on television are generally presented as loving, supportive, and harmonious, as they rarely experience serious family conflict. Buerkel-Rothfuss, Greenberg, Atkin, and Neuendorf (1982) looked at what children learned by watching these television families. They found that those watching family programs were more likely to believe that real-life families show support and concern.

Signorielli (1991) found that adolescents who watched more television were more likely to want traditional family relationships such as getting married, staying married to the same person, and to have children (p. 145).

On the other hand, Morgan, Leggett, Shanahan (1999) found that heavy viewers tend to be less likely to endorse family values in the areas of illegitimacy and single parenthood (p. 54). However, age was an important intervening variable, as younger people were less traditional in their views. As they explained, this could possibly be "a result of, rather than a contributor to, non-conventional family views" (p. 58). In other words, young people who accepted illegitimacy and single parenthood turned to television more.

Many cultivation studies have shown that heavy viewers tend to hold more conservative values, reflecting the political and cultural mainstream presented on commercial television. Such values include support for traditional and harmonic family relationships (see Gerbner, Gross, Morgan and Signorielli, 1982; Morgan, 1986; Morgan and Shanahan, 1995).

Later studies of cultivation have gone beyond simply examining overall viewing, and have also looked at specific television genres (See Potter, 1993; Potter, 1994). Another important issue in this type of research is the perceived realism of television programs (Potter, 1986). For example, those who perceive television as more realistic are more likely to be influenced by its content. Again, one would expect that foreign viewers of US programs would differ with American viewers on this matter, because foreign viewers do not have a US experience with which to compare the realism of US television content. Elliott and Slater (1980), in a US study, for example, found that frequent viewers of certain programs tend to see them as more realistic, while those with direct positive experience (in this case with the police) perceived the programs as less realistic.

Another further refinement of traditional cultivation research is the consideration of the respondents' motivation to watch. For example, Stilling (1994) found that motivation and exposure to certain genres was a better predictor of television's acculturation effect than simply amount of viewing.

Generally, however, Gerbner et al. (1979) believe that "heavy television viewers perceive social reality differently from light TV viewers even when other factors are held constant" (p. 193), and this social reality is influenced by the amount of television viewing.

Given the overall findings that television, at least in the United States, tends to portray the family in more harmonious relationships, this study aims to test the following hypotheses:

1. Adolescent heavy television viewers will tend to favor more communication with their parents about love relationships and other problems. 
2. Heavy television viewers of US television will tend to favor more communication with their parents about love relationships and other problems. 
3. Heavy viewers of television will perceive the role of grandparents as more important in the family, than light viewers. 
4. Given that US television does not often portray extended families involving grandparents, heavy viewers of US television will perceive grandparents as less important in the family than light viewers. 
5. Heavy viewers of television will tend to believe that families are generally happy. 
6. Heavy viewers of television, as compared to light viewers, will tend to believe that children should obey their parents. 
7. Heavy viewers of television, compared to light viewers, will tend to adhere to more traditional views of Greek values, as expressed in their perception of "philotimo," and their church attendance. 
8. Heavy viewers of US television, compared to light viewers, will tend to hold less traditional views of Greek values, as expressed in their perception of "philotimo," and their church attendance. 
9. Heavy viewers of US television will hold more positive views about divorce.

Furthermore, in order to examine other television related variables, this study poses the following research question: What demographic, media consumption patterns, and other socioeconomic variables play a role in how viewers see traditional family relationships? Of particular interest are, such variables as, motivation to watch television, perception of realism of television programs, and specific types of media consumption.

Method 
Two Greek senior high schools, or lycea, were chosen for this research. One in a middle class section of Athens, and another in the agricultural town of Amaliada, which has a population of about 17,000 in southwestern Greece. These schools were chosen because they represent the urban/rural dichotomy of Greece, because they represent Greek society without extreme socioeconomic characteristics, and because access to these schools was easier, in terms of cooperating teachers and principals.

A survey questionnaire was designed first in English, using questions such as ones used in similar studies around the world, such as Kang and Morgan (1988). The instrument included Likert-type questions, as well as some open ended ones. This questionnaire was initially translated in the United States, and was later proofread and polished by professional proofreaders in Greece, and back-translated. Following approval of the proposed research by the Greek Ministry of Education and its Pedagogical Institute, which examined the questionnaire, the instrument was administered at the two schools. One teacher at each school was trained to instruct the other teachers on how to administer the questionnaire in their classes during the same day.

In order to test the hypotheses participants were split into two groups: heavy and light television viewers. Those watching television up to 150 minutes per day were classified as light viewers, while those watching more than 199 minutes were classified as heavy viewers. The average student watched 184 minutes per day, but the median was 174 minutes.

To answer the research question, a stepwise regression analysis was used to find predictor variables for each relationship. The following independent variables were used: Overall TV viewing; frequency of US program viewing; proportion of viewing devoted US and Greek programs; demographic and other socioeconomic characteristics of student and parents; amount of radio listening; frequency of newspaper reading; types of television viewing; specific program viewing; perceived realism of television programs; and motivation for viewing.

Findings 
Of the 508 usable surveys, 255 came from Amaliada, and 253 from Athens. Males make up 40.9 percent of the sample (208), and females make up 58.5 (297), while three did not respond to this question. The Greek senior high school is made up of three grades, and students were equally divided between the three: 170 from the tenth, 168 from the eleventh, and 170 from the twelfth grade.

Television and Other Media Consumption 
The students, on average, watch approximately 3 hours of television each day, including weekends, although they watch more on weekends than on weekdays. In terms of viewing US television programming, of the 403 students answering the open-ended question, "from which nation other than Greece do your favorite television programs come?" 323 or 63.6% said the United States. Furthermore, 18.5% of the respondents reported watching US television programs on a daily basis, 45.4% on a weekly basis, 18.1% on a monthly basis, and 18.1% reported that they rarely or never watch US shows. A Pearson correlation reveals that more television viewing (in minutes) is related to watching US programs more often during an average month [r (N = 476) = .25, p < .0001)]. Overall, respondents said they spend an average of a little over three hours a week watching US programs. Those who watch at least some US programs report that they spend an average 37.5% of their total viewing watching US television programs.

Generally males watch more than females (199 minutes per day vs. 176 for females) [F (1, 484) = 6.11, p < .014]. However, males spend 60% of their TV time watching Greek shows, while females spend 53.7% of their time doing the same [F (1, 434) = 8, p < .005].

As expected, males and females also differ in terms of what they watch. Males tend to watch NBA basketball and other sports programs, as well as Greek late night talk shows. Females watch more Greek sitcoms ("E Men kai E De," "Dis Ex' Amartin"), "Beverly Hills 90210," "Dr. Quinn: Medicine Woman," "Melrose Place," and soap operas like "Loving," and Greek telenovelas or social dramas ("Telefteo Antio," "Lampsi") [X2 (30, N = 432) = 126, p < .0001]. "Beverly Hills 90210" was the most popular program overall (including Greek shows), as 8% of all students made an effort to watch it each week.

Generally the students were divided as to whether US programs accurately portray real life in the United States-38% felt they generally do, 32.6% had no opinion or did not know, and 29.4% said US programs generally do not portray accurately life in the United States.

The students were asked to respond to a series of statements regarding family relationships and values. On the statement that "young people should talk about their love relationships with their parents," the students generally agreed (mean = 2.2 on a 5-point scale, with 1 being strongly agree). However, no significant difference exists between heavy and light viewers on their response to this statement [F (1, 382) = .38, p > .05]. Similarly on the statement that "young people should talk over problems with their parents" there was no significant difference between heavy and light viewers [F (1, 388) = .3, p > .05]. However, students were more supportive of this type of communication (mean = 1.7) than communicating about love relationships. Therefore, hypothesis one is rejected.

In terms of communication with parents about love relationships, no significant difference is found between heavy and light viewers of US television programs, either in estimated hours per week [F (1, 456) = .05, p > .05], or in the percentage of their viewing dedicated to US shows [F (1, 491) = .01, p > .05]. Similarly no difference was found regarding the statement regarding communicating about problems. Heavy and light viewers of US television programs do not differ, either in estimated hours per week [F (1, 463) = .24, p > .05], or in the percentage of their viewing dedicated to US shows [F (1, 498) = .69, p > .05]. Thus, hypothesis two is also rejected.

Responding to the statement "It is good to have grandparents living with the family," students were ambivalent (mean = 3). No significant difference exists between heavy and light viewers on this issue [F (1, 381) = .2, p > .05]. Thus, hypothesis three is rejected.

However, heavy viewers of US programs, in terms of spending a greater portion of their viewing time watching US shows, tend to feel more negative about living with the grandparents [F (1, 488) = 4.4, p < .04]. Thus, hypothesis four is retained.

Students were also ambivalent in responding to the statement "family members are generally happy" (mean = 2.96). Even though there was some of the hypothetized difference between heavy and light viewers, this was not statistically significant [F (1, 380) = 3.8, p > .04]. Hypothesis five is rejected.

In testing hypothesis six, students responded to the statement, "It is natural that young people should obey their parents. They generally did not agree (mean = 3.5). There is no significant difference between heavy and light viewers on this issue [F (1, 385) = .13, p > .05]. However, those spending a greater portion of their viewing time watching US shows tend to disagree more strongly on this issue about obeying parents [F (1, 496) = 6.9, p < .01].

To test hypothesis seven, students were asked to respond to the statement that "philotimo is still an important value to hold." Philotimo, which characterizes Greek value system and embodies a combination of pride and honor, seems to still be important to these young people (mean = 1.6). However there is no difference on this issue between heavy and light viewers [F (1, 381) = .12, p > .05]. Similarly there is no difference between heavy and light viewers on how often they attend church [F (1, 363) = .10, p > .05]. The students, on average, responded that they attend a few times each year. As such, hypothesis seven is rejected.

Hypothesis eight extended the two above statements to US television viewing. The study finds no difference between heavy and light viewers of US television programs, either in estimated hours per week [F (1, 454) = .59, p > .05], or in the percentage of their viewing dedicated to US shows [F (1, 489) = .25, p > .05], about the importance of Philotimo. However, those spending a greater portion of their viewing time watching US shows tend to go to church less often than those who spend a smaller portion of their viewing time viewing US programs [F (1, 468) = 13.8, p < .001]. Therefore, only a part of hypothesis eight holds true.

Finally, students were asked what they thought of divorce. Most replied that divorce is generally a mistake. However, there is no significant difference between light and heavy viewers on this issue [F (1, 384) = 2.8, p > .05]. Nor, is there a significant difference between light and heavy viewers of US television about the issue of divorce [F (1, 488) = .46, p > .05]. Hypothesis nine is rejected.

A regression analysis was used to answer the research question "what demographic, media consumption patterns, and other socioeconomic variables play a role in how viewers see traditional family relationships?" Generally, media related variables were found to be significant predictors on most of the tested relationships.

On the issue of young people talking to their parents about love relationships, four variables accounted for 18 percent of the variance: watching TV adventure shows, perceived realism of US shows, a student's mother's politics, and family income level (see Table 1). On the other hand, there is only one significant predictor for the attitude toward talking to parents about problems: This is the student's own political position, which accounts for only four percent of the variance (see Table 1).

Table 2 outlines significant predictors of attitudes toward obeying one's parents. Media related variables account for all 14% of the variance. These are: listening to radio, reading certain newspapers, and the perception of realism reflected in Greek television programs. Perceived realism of Greek television programs is the sole predictor of the perception of happy families, accounting for 14% of the variance (see Table 3).

On the attitude regarding the role of grandparents, there are three significant predictors. However, newspaper reading is the only media related one (see table 4). Twenty-one percent of the variance is accounted for regarding values such as "philotimo" (pride/honor/shame). Four of five significant predictors are media related variables: Radio listening, reading a certain newspaper, viewing television cartoons, and viewing music television programs (see Table 5).

Church attendance is predicted by four variables (see Table 6). Watching sitcoms and perceived realism of Greek television are the only media related predictor variables. Finally, perception of divorce is predicted by the sex of the respondent and reasons for watching television. Together they account for 11% of the variance (see table 7).

Discussion 
Aggregate television viewing seems to play a very limited role in the cultivation of adolescents' perceptions of Greek families, and on shaping their attitudes toward family relationships. Generally, this study finds that only on some issues do heavy television viewers differ from light viewers, and in all of these cases it is viewing US television that is related. For example, heavy viewers of US television programs tend to feel less supportive of the idea of including grandparents in their household, are less supportive of the idea that teenagers should obey their parents, and go to church less often than light viewers of US television programs. However, it is not clear from this research whether Greek teenagers have these attitudes because they watch US television, or they are attracted to US television because they hold these attitudes.

Nevertheless, television and other media, as well as media related variables, do play a role in the perceptions and attitudes of Greek adolescents. Perceived realism of television programs is especially related to cultivation effects. For example, perceived realism of television was found to be the lone predictor of the perception that families are generally happy; and a significant predictor of a positive attitude toward talking to parents about love relations, the belief that young people should obey their parents, and church attendance.

Specific TV genres also play a role in the students' attitudes toward other issues. For example, watching adventure programs is a predictor of the belief that young people should talk to their parents about love relationships. Watching cartoons and music television are significant predictors of attitudes toward the important Greek value of "philotimo." And finally, watching sitcoms is a predictor of church attendance.

Other media related variables such as newspaper reading and radio listening were also found to be significant predictors of certain attitudes. Finally, reasons for watching television was found to be a predictor of one's perception of divorce.

It seems that this study tends to reinforce findings from other studies, which point to the complexity of cultivation effects (Weaver and Wakshlag 1986; and Potter, 1986). This study also reinforces the idea that specific television content and other media consumption are factors in how young people perceive the world around them, and are related to attitudes they hold.

Obviously this study did not undertake a content analysis to determine the actual portrayal of families and family relationships on Greek television. This should be done in future studies. Also, this is a study of adolescents, and the results here cannot necessarily be generalized to the whole population. Future studies need to examine more closely how viewers actually process television entertainment and information, because television influence seems to be related to already held attitudes and perceptions. 

---

References

Buerkel-Rothfuss, N. L. Greenberg, B. S., Atkin, C. K. and Neuendorf, K. (1982). "Learning about the family from television." Journal of Communication, 32(3), 191-201.

Cantor, M. (1990). "Prime-time fathers: A study in continuity and change." Critical Studies in Mass Communication, 7, 275-285.

Elliott, W. R., and Slater, D. (1980). "Exposure, experience and perceived television reality for adolescents." Journalism Quarterly, 57 (3), 409-414, 431.

Gerbner, G., Gross, L., Signorielli, N., Morgan, M., and Jackson-Beeck, M. (1979). "The demonstration of power: Violence profile No. 10." Journal of Communication, 29 (3), 177-196.

Gerbner, G., Gross, L., Morgan, M., and Signorielli, N. (1982). Charting the mainstream: Television's contributions to political orientations. Journal of Communication, 32(2), 100-127.

Gerbner, G., Gross, L., Morgan, M., and Signorielli, N. (1994). "Growing up with television: The cultivation perspective." In J. Bryant and D. Zillman (eds.), Media effects, p. 17-41. Hillsdale, NJ: Lawrence Erlbaum Associates.

Hawkins, R. P. and Pingree, S. (1982). "Television's influence on social reality." In D. Pearl, L. Bouthilet, and J. Lazar (eds.), Television and behavior: Ten years of scientific progress and implications for the eighties. DHHS Publication No. ADM 82-1196, Vol. 2, p. 224-247. Washington, DC: US Government Printing Office.

Hawkins, R. P., Pingree, S. and Alter, I. (1987). "Searching for cognitive processes in the cultivation effect: Adult and adolescent samples in the United States and Australia." Human Communication Research, 13 (4), p. 553-577.

Kang, J. G. and Morgan, M. (1988). "Cultural clash: Impact of US television in Korea." Journalism Quarterly, 65 (2), 431-438.

Moore, M. L. (1992). "The family as portrayed on prime-time television, 1947-1990: Structure and characteristics." Sex Roles, 26, 41-61.

Morgan, M. (1986). "Television and the erosion of regional diversity." Journal of Broadcasting and Electronic Media, 30, 123-139.

Morgan, M. (1990). "International cultivation analysis." In N. Signorielli and M. Morgan (eds.), Cultivation analysis: New directions in media effects research, p. 225-247. Newbury Park: Sage.

Morgan, M. and Shanahan, J. (1995). Democracy tango: Television, adolescents, and authoritarian tensions in Argentina. Cresskill, NJ: Hampton.

Morgan, M. and Shanahan, J. (1997). Two decades of cultivation research: An appraisal and a meta-analysis. In B. Burleson (ed.) Communication Yearbook 20, p. 1-45. Newbury Park, CA: Sage.

Morgan, M., Leggett, S., and Shanahan, J. (1999). "Television and family values: Was Dan Quayle Right?" Mass Communication and Society, 2(1/2), 47-63.

Pingree, S., and Hawkins, R. (1981, Winter). "U.S. programs on Australian television: The cultivation effect." Journal of Communication, 31(1), 97-105.

Potter, W. J. (1986, Spring). "Perceived reality and the cultivation hypothesis." Journal of Broadcasting and Electronic Media, 30(2), 159-174.

Potter, W. J. (1993). "Cultivation theory and research: A conceptual critique." Human Communication Research, 19(4), pp. 564-601.

Potter, W. J. (Oct. 1994). "Cultivation theory and research: A methodological critique." Journalism Monographs, 147.

Signorielli, N. and M. Morgan (eds.) (1990). Cultivation analysis: New directions in media effects research. Newbury Park: Sage.

Signorielli, N. (1991). "Adolescents and ambivalence towards marriage." Youth and Society, 23, 121-149.

Skill, T., Robinson, J. D., and Wallace, S. (1987). "Portrayal of families on prime-time television: Structure, type, and frequency." Journalism Quarterly, 64, 360-367, 398.

Stilling, E. A. (1994). "The electronic melting pot hypothesis: The cultivation of acculturation among Hispanics through television viewing." Paper presented at the Broadcast Education Association convention, Las Vegas, NV.

Weaver, J. and Wakshlag, J. (1986). "Perceived vulnerability to crime, criminal victimization experience, and television viewing." Journal of Broadcasting, 30 (2), pp. 141-158.

Zaharopoulos, T. (March 1990). "The dimensions of Greek television." Paper presented at the Broadcast Education Association convention, Las Vegas, NV.

Zaharopoulos, T., and Paraschos, M. (1993). Mass media in Greece: Power, politics, and privatization. Westport, CT: Praeger.