January 8, 2005

Exploring Pathways From Television Viewing to Academic Achievement in School Age Children


The author's purpose in this study was to test 4 hypotheses that proposed different paths for the influences of children's television viewing on their academic achievement. Data were drawn from the 1997 Child Development Supplement (CDS) to the Panel Study of Income Dynamics (PSID). The population for this study included 1,203 children between the ages of 6 and 13 years from the CDS-PSID data set. The author used structural equation modeling to test pathways from children's television viewing to their academic achievement. The author assumed that children's television viewing hindered their academic achievement by reducing certain traits that related to academic achievement. Results showed that 3 hypothetical models fit the data-the time-displacement hypothesis, the mental effort- passivity hypothesis, and the attention-arousal hypothesis. A 4th hypothetical model, the learning-information hypothesis, which proposed that children's television viewing practices stimulate their academic achievement, was not supported. In sum, children who watched more television tended to spend less time doing homework, studying, and reading for leisure. In addition, their behaviors became more impulsive, which resulted in an eventual decrease in their academic achievement.

Key words: academic achievement, structural equation modeling, television viewing

RESEARCHERS HAVE RECOGNIZED THAT WATCHING TELEVISION occupies a larger amount of children's time than does any other single activity, with the exception of sleeping (J. L. Singer & Singer, 1983). Television is generally assumed to be one of the most important environmental factors that influences child development. Among the developmental domains, the relationship between cognitive development and television viewing has been the one most widely studied. Investigators disagree about the effects of this relationship (MacBeth, 1996). Researchers have found both positive and negative relationships between television viewing and cognitive functioning. There are a number of hypotheses regarding television viewing and children's cognitive development. As Anderson, Huston, Schmitt, Linebarger, and Wright (2001) showed, the hypotheses on the relationship between television and children's cognitive competence, such as academic achievement, have focused on the effects approach, which follows the tradition of communication research. Theories about the influences of television on children's academic achievement are divided into two broad categories: (a) stimulation hypotheses, which propose that watching well-designed programs enhances children's academic achievement because children learn from watching the programs and (b) reduction hypotheses, which concern the unique characteristics of television as a medium that inhibits the viewers' intellectual processing or leads to specific behaviors that may hinder children's academic achievement (Valkenburg & van der Voort, 1994).

Four groups of hypotheses have been proposed regarding the processes by which watching television stimulates or reduces academic achievement. First, the learning hypothesis (e.g., Bandura, 1994) and the information-processing hypothesis (e.g., Huesmann, 1986) are stimulation hypotheses. Both theories propose learning mechanisms by which children acquire information from television. According to these hypotheses, television, especially informative programs designed for children, may increase academic achievement because children can learn a wide range of academic content, such as letter and number recognition, from television (Anderson et al., 2001; D. G. Singer & Singer, 2001).

Second, the time-displacement hypothesis is one of the reduction hypotheses. According to the time-displacement hypothesis, watching television displaces or takes time away from intellectually demanding activities such as doing homework and studying (Beentjes & van der Voort, 1989; Harrison & Williams, 1986; Valkenburg & van der Voort, 1994). This displacement happens because television, which entertains the viewers with rapid-paced images and visual and auditory effects, is more attractive to children than are the school- related activities. Consequently, watching television presumably displaces comparable activities that involve learning opportunities (Anderson et al., 2001) and eventually decreases children's academic achievement (Koshal, Koshal, & Gupta, 1996).

Third, there is some evidence that watching television requires less mental effort than does reading (Beentjes, 1989; Salomon, 1984). In light of that issue, the mental-effort hypothesis and the passivity hypothesis propose that children become passive as they watch television (W. A. Collins, 1982). According to the mental- effort hypothesis (Kooltra & van der Voort, 1996), watching television leads to mental laziness. The more television that children watch, the less investment that they put in the mental effort required to master academic skills. Instead, they watch television programs that are entertaining and easy to understand. As a result, children are not likely to engage in activities that require intellectual functioning, such as reading. In a similar way, the passivity hypothesis proposes that the inner processing of information from television is passive and requires little mental effort compared with reading because rapid-paced programs leave children less room for reflection and hinder reflective thinking (Valkenburg & van der Voort, 1994). A low level of mental effort is elicited by television viewing and eventually leads to a tendency to expend little mental effort required to read or to solve arithmetic problems (Suedfeld, Little, Rank, Rank, & Ballard, 1986). Therefore, this propensity for mental laziness or passive-cognitive functioning can be considered to be a result of watching television, which may hinder academic achievement that requires more active intellectual efforts.

Last, the attention hypothesis and arousal hypothesis propose that television has negative effects on children's academic achievement. These hypotheses propose that television viewing encourages impulsive behaviors and may eventually decrease academic achievement, because television uses frequent movements and cuts that may discourage sustained activities (Andersen et al., 2001). Specifically, according to the attention hypothesis, television programming leads to superficial intellectual processing (P. A. Collins, 1991; MacBeth, 1996). As a result of growing up with television, children are less able to be task oriented. Thus, children may have difficulty sustaining attention in classroom situations and ultimately their academic achievement may be low (Greenfield, 1984; Healy, 1990; MacBeth; J. L. Singer, 1980; Valkenburg & van der Voort, 1994). Similarly, the arousal hypothesis argues that watching television with a format of rapidpaced and sophisticated visual pacing techniques has arousing effects on the viewers (J. L. Singer, Singer, & Rapaczynski, 1984; Valkenburg & van der Voort). The arousal produced by watching television leads to a restless and impulsive behavioral orientation, and eventually academic achievement that demands attention may be hindered (J. L. Singer & Singer, 1986).

In sum, there are a variety of hypotheses regarding pathways from children's television viewing practices to academic achievement based on different theories. Anderson and his colleagues (2001) argued that most of the hypotheses are beliefs that are contradictory at the conceptual level and that require more empirical evidence. Therefore, it is important to further explore the process from television viewing to academic achievement, which was proposed by the hypotheses.

My purpose in this study was to examine the processes through which children's television viewing is related to their academic achievement based on the four different hypotheses (see Figure 1). Researchers have hypothesized that children's television viewing is likely to stimulate their academic achievement based on the learning- information hypothesis (Anderson et al., 2001; Bandura, 1994; Huesmann, 1996; D. G. Singer & Singer, 2001). Television viewing also was assumed to hinder their academic achievement through: (a) decreasing the amount of homework and studying (the time- displacement hypothesis), (b) decreasing the amount of leisure reading (the mental effort-passivity hypothesis), and (c) increasing impulsive behaviors (the attention-arousal hypothesis). I used data from a national sample to test the different pathways of television viewing effects on their academic achievement (Child Development Supplement [CDS] to the Panel Study of Income Dynamics [PSID], 1997).

FIGURE 1. Conceptual model of the processes by which television viewing may affect academic achievement.



I obtained data for this study from the 1997 CDS-PSID, which was collected by the Institute for Social Research at the University of Michigan. The CDS is a supplementary component of the PSID, which was a longitudinal survey of a representative population of U.S. residents and their families. The objective of the CDS was to provide researchers with additional opportunities to study the dynamic process of early human capital f\ormation, which included family structure and living arrangements, neighborhood economic and social conditions, and school resources and programs, to children's developmental outcomes (CDS-PSID, 1997). Among 2,700 PSID families, 2,394 families, consisting of children 12 years old or younger, were selected and interviewed for the CDS. Because up to two children in a family were included, a total 3,563 children participated in the CDS-PSID. Although CDS-PSID provides only cross-sectional data collected in 1997, the data were obtained from various people who surrounded each participating child (such as the mother, a second caregiver, an absent parent, a teacher, a school administrator) and the CDS researchers who assessed the children.

To control for children's television-viewing environment in the present study, I selected children who attended school and whose family owned a working television set. The population for this study included 1,203 children between the ages of 6 and 13 years who were part of the 3,563 children in the CDS data set.

The children in my study were 50.3% boys and 49.7% girls with a mean age of 9 years. The ethnic composition of the population was 50.0% European American, 45.6% African American, 0.7% Hispanic, 0.1 % Asian American, and 0.7% Native American. Family income ranged from $0 to $577,000, with a mean income of $45,958 (SD = $45,692). The mean years of education of the family heads was 12.64 years (SD = 2.68), with a median of 12 years.


The predictors of academic achievement were (a) the amount of time the children spent watching television, (b) the amount of time the children spent completing homework and studying, (c) the amount of time the children spent reading for leisure, and (d) impulsive behaviors.

First, a time diary was used to measure the amount of time the children spent watching television, completing homework and studying, and reading for leisure. Primary caregivers reported a list of the target child's activities during one weekday and one weekend day. These caregivers were asked to record their child's activities for the entire 24-hr time period of both days, starting with midnight on the specified day and continuing until midnight on the next day. In the diary they reported what the child did, the actual time when each activity began and ended, the location of the child, the people who were directly and indirectly involved in each activity, and other activities that the child was doing simultaneously. If necessary, the target children themselves and other family members were allowed to report in the diary. Each of the reported activities in the time diary was coded so that the total time for activities done on a day equaled exactly 24 hr. Among a variety of activities, the definitions and activity coding instructions of three activities are as follows.

Amount of television viewing. Children's total time primarily spent watching television on one weekday and one weekend day was the sum of the minutes for the activities, which were coded as television in the time diary.

Homework and studying. Children's total time primarily spent on doing homework and studying one weekday and one weekend day was the sum of the minutes for the activities, which were coded as doing homework, studying, researching, or reading related to classes or profession except for current job in the time diary.

Leisure reading. Children's total time primarily spent on leisure reading on one weekday and one weekend day was the sum of the minutes for the activity, which were coded as reading or looking at books (not current job related; not professionally class related) or looking at books (even if the child cannot read) in the time diary.

In addition to the predictors from the time diary, I used two assessments of children in the present study-impulsive behaviors and academic achievement.

Impulsive behaviors. The measure of impulsive behaviors consisted of five items I selected. Three items were from the Behavior Problems Index (BPI; Peterson & Zill, 1986) and two items were from the Positive Behavior Scales (PBS; Polit, 1998). The primary caregivers completed the BPI and the PBS. The BPI was developed to measure the incidence and severity of child behaviors (Peterson & Zill). The primary caregivers responded on 3-point rating scales- not true (1), sometimes true (2), often true (3)-to items selected from the BPI. The three items I selected from the BPI were: (a) He or she has difficulty concentrating, cannot pay attention for long, (b) He or she is impulsive or acts without thinking, and (c) He or she is restless or overtly active, and cannot sit still.

The PBS was developed to measure the positive aspects of children's behaviors, including persistence, self-control, self- esteem, and obedience and compliance (Polit, 1998). For this study, I selected two items from the PBS, which were assessed on 5-point rating scales: not at all like child (1), to totally like child (5). The items were: Does neat, careful work and thinks before he or she acts, and is not impulsive. For high scores to indicate high impulsiveness, I reverse coded the two items of the PBS. To create a measure that combines the 5 items, I standardized and summed the scores on each item with a mean of O and a standard deviation of 1. Cronbach's alpha for the five items was .75 for this population.

Academic achievement. CDS-PSID researchers administered four subsets of the Woodcock-Johnson Revised Tests of Achievement (WJ-R; Woodcock & Johnson, 1989) at the child's school to assess academic achievement. The WJ-R test contains nine subtests that measure different aspects of academic achievement. The four subjects I chose for the CDS-PSID covered only the reading and math portion of the test. The measurement consisted of the letter-word (57 items), passage comprehension (43 items), calculation (53 items), and applied problem (60 items) tests, and each subtotal score was standardized. CDS-PSID reports explained that these tests were chosen because of the ease of administration as well as their brevity (CDS-PSID, 1997). On average, across all age groups, test administration took approximately 40 min. When applicable, I used the Spanish version of the WJ-R for children whose primary language was Spanish.

According to the CDS-PSID reports (1997), these scales can be used individually, or in the case of the four subscales, combined to create scores for Broad Reading and Broad Math. Because a measurement model with a latent variable-academic achievement with two indictors (broad reading and broad math) was not identified, I did not use the scores of broad reading and broad math in this study. Instead, I used the four standardized scores of the four subscales. Cronbach's alpha for this population was .83.

Measurement model. Prior to testing the initial model, I evaluated a measurement model. In the hypothetical model depicted in Figure 1, a measurement model for the present study consisted of one latent construct and four observed variables. The observed variables were academic achievement as indicated by the scores on the letter- word, passage comprehension, calculation, and applied problem tests. I used structural equation modeling to conduct confirmatory factor analysis of the model to evaluate whether the model fit the data. To measure maximum likelihood estimation, I used the covariance matrix as input data to estimate model parameters. For an analysis of the overall fit of the model I obtained, chi-square statistics, the Goodness-of-Fit Index (GFI), the Adjusted Goodness-of-Fit Index (AGFI), the Comparative Fit Index (CFI), the Normed Fit Index (NFI), and the Root Mean Square Error of Approximation (RMSEA).

Examination of parameter estimates showed that all indicator loadings and their critical ratios for the latent construct were large enough for significance (letter-word = .81, passage comprehension = .84, calculation = .80, applied problem = .82). Despite the highly significant factor loadings, the fit indexes of the measurement model were fair. Both statistics of model fits indicated good and prior values together, χ^sup 2^/df=25.61, GFI = .95, AGFI = .87, CFI = .96, NFI = .96, RMSEA = .16. GFI, CFI, and NFI indicated the measurement model fit the data well. The χ^sup 2^/df ratio, AGFI, and RMSEA, however, showed the model did not fit the data. The results of modification indexes (MI) indicated that the model fit can be improved when two covariance paths between letter-word and passage comprehension (MI = 61.62) and between calculation and applied problem (MI = 50.99) are added to the original measurement model. On the basis of the results, I added two paths between letter-word passage comprehension and calculation- applied problem to improve the adequacy of the measurement model. It revealed that the addition of two covariances resulted in significantly better values of all other indexes, χ^sup 2^/df = 0.42, GFI = 1.00, AGFI = .99, CFI = 1.00, NFI = 1.00, RMSEA = .00.

Aside from the empirical evidence from the statistics, the revised measurement model is also a better representation of the data conceptually. The scores of letter-word and passage comprehension tests represent reading achievement, and the scores of calculation and applied problems stand for math achievement. Therefore, I retained all the indicators in the original measurement model along with two paths between letter-word and calculation and between applied problem and passage comprehension.


Preliminary analyses

I first obtained means, standard deviations, and correlations for all measures for entire populations. Table 1 shows correlations among the amount of time spent watching television, the three mediating variables (the amount of time spent completing homework and studying, reading for leisure, and impulsive behaviors), and the four academic achievement scores (letter-word, passage comprehension\, calculation, applied problem). As shown, the hypothesized relationships among the amount of time spent watching television, the three mediators, and the four indicators of academic achievement were mostly significant; only the correlation between homework and studying and calculation was insignificant. Also, there was no significant association among the three mediators, which represent separate pathways from television viewing to achievement.

TABLE 1. Means, Standard Deviations, and Correlations for All Measures (N = 1,203)

In particular, the directions of relationships among television viewing, the mediators, and the indicators of academic achievement were mostly consistent with what each hypothesis proposed; only the directions of the associations between television viewing and indicators of academic achievement were different from the learning- information hypothesis. Because the learning and information hypothesis assumes a positive relationship between the amount of television viewing and academic achievement, the negative correlation between television viewing and four scores of academic achievement was contradictory to the hypothesis.

Structural Equation Modeling I

I analyzed the hypothesized structural relationships in Figure 1 using the structural equation model. The initial model included eight observed variables, (a) television viewing, (b) homework and studying, (c) amount of leisure reading, (d) impulsive behaviors, (e) letter-word scores, (f) passage-comprehension scores, (g) calculation scores, (h) applied-problem scores, and one latent variable-academic achievement (see Figure 2).

As shown in Figure 2, children's television viewing was likely to decrease their time doing homework, studying, and reading for leisure, and to increase their impulsive behaviors. In turn, the time the children spent doing homework, studying and reading for leisure were likely to increase their academic achievement, and their tendency to behave impulsively tended to decrease their academic achievement.

FIGURE 2. Full model of the processes of effects of television viewing on academic achievement. The estimates of factor loadings are standardized regression weights. The estimates between variances are correlations. *p

For analysis of overall fit of the initial model, J obtained chi- square statistics, GFI, AGFI, CFI, NFI, and RMSEA. The statistics of the fit measures of the initial model are shown in Table 2. All the indexes supported a good level of fit.

Structural Equation Modeling II

In addition to the test of the initial model that examined the fit of the competing hypotheses simultaneously, I completed additional analyses of individual models guided by theoretical considerations. I independently examined individual models that proposed different pathways from television viewing to academic achievement so I could examine the fit of the hypotheses separately. In the individual models, I assumed television viewing was completely mediated by a mediator depending on the hypothesis: (a) the model based on the time-displacement hypothesis represented the path from television viewing to academic achievement, mediated by the amount of studying and homework; (b) the model based on the attention-arousal hypothesis represented the path from television viewing to academic achievement, mediated by impulsive behaviors; and (c) the model based on the mental effort-passivity hypothesis represented the path from television viewing to academic achievement, mediated by leisure reading. I did not individually examine the learning-information hypothesis, which was not supported. Contrary to the hypothesis assuming positive effects of television viewing on achievement, the parameter of the path from television viewing to academic achievement was significantly negative.

The results of these supplementary structural models also are shown in Table 2. As shown in the table, each model was also a good representation of the data. Separately, all of the fit indexes for the three models-the time-displacement hypothesis, the attention- arousal hypothesis, and the mental effort-passivity hypothesis- could be statistically supported.


In this study, I reviewed and tested the hypotheses on the effects of children's television viewing on their academic achievement by using structural equation modeling.

The results of analyses of the initial full model and individual models showed that the more time children spend watching television, the less time they spend doing homework, studying, and reading for leisure. In turn, the decreased time spent doing homework, studying, and leisure reading contribute to less academic achievement. However, the more time children spend watching television, the more impulsively they behave, and eventually they show a decrease in academic achievement. In summary, the three hypotheses (the time- displacement hypothesis, the attention-arousal hypothesis, and the mental effort-passivity hypothesis) were supported for this population.

TABLE 2. Fit Indexes of Analyzed Models

Although the three hypotheses proposing negative effects of television viewing on achievement have been widely accepted, the criticism of the positive influences seem to be plausible but unproven (Anderson & Collins, 1988; Anderson et al., 2001; MacBeth, 1996). My findings in the present study provide empirical evidences supporting the three hypotheses; so, different pathways from children's viewing practices to achievement are supported. Therefore, it can be implied that potential influences of television viewing on achievement would be negative. In addition, there are multiple pathways that cannot be explained by a single theory.

In particular, the findings supporting the three hypotheses are meaningful to identify mediators between children's viewing and their academic achievement. Empirical evidence has supported the learning-information hypothesis, which assumed television-viewing effects on academic achievement in a simplistic form. Assumptions of the time-displacement hypothesis, the mental effort-passivity hypothesis, and the attention-arousal hypothesis are not simplistic (Anderson et al., 2001). In this study, we used structured equation modeling to test hypothetical pathways from television viewing to achievement, and we revealed a complex model beyond straightforward prediction of relationship between television viewing and achievement.

Although pathways from children's television viewing practices to academic achievement were revealed as multiple, indirect, and negative, the findings using structural equation modeling should be carefully interpreted. The correlations among the variables, for example, are statistically significant but relatively low. In structural equation modeling, good fit refers to a good expectation of hypothetical model in which relationships among variables match actual associations of the data. Even if the correlations are low in this study, the hypothetical-structural model, which anticipates the same patterns of correlations of the data, shows a good fit. Therefore, the final hypothetical model of this study should be understood as a representation showing the structural relationships that enable a clearer conceptualization of the theories, that is, the time-displacement hypothesis, the attention-arousal hypothesis, and the mental effort-passivity hypotheses (Bryne, 1994).

Of interest, the learning-information hypothesis was not supported, even though many past studies have shown positive effects of viewing practices on children's achievement. Rather, the association between the amount of time spent watching television and academic achievement was negative, even after the effects of the three mediating variables were controlled.

There is an important point that should be considered when interpreting the results. To test the learning-information model, I assumed that the total amount of time spent watching television did itself positively contribute to academic achievement. As Huston and Wright (1996) argued, however, the influences of children's television viewing on their academic achievement may be tied to the content of television programs watched in addition to the total amount of time children watch television. Television programs can be stimulating, creative, and of high quality; however, they also can be a source of portrayed violence and cruelty with unsophisticated techniques. Among a variety of programs, the past studies supporting the learning-information hypothesis tested the effects of watching well-designed shows that deliver educational messages for young children, such as Sesame Street (Ball & Bogatz, 1970; Bogatz & Ball, 1971; Zill, Davis, & Daly, 1994) and Mister Rogers' Neighborhood (Stein & Fredrich, 1975). In my study, however, I did not examine the influences of children's educational programs differently from the others because the children's television viewing diet was unavailable in the secondary data. The programs that children watched were not investigated. The learning-information hypothesis might have been rejected for this reason in the present research. Therefore, the influences of different programming should be considered as an important factor when the effects of television on achievement are investigated in future studies.

Another point that should be included in further research on the effects of television-viewing practices on achievement is children's uses of other mass media. Television is the most influential and accessible mass medium, but not the only one. Although the national data set I used in this study provided an excellent opportunity to test the hypotheses of effects of the most prevalent medium, there are many types of mass media, such as DVD recordings, video games, and the Internet, which young children also experience in addition to television. Because television is expected to be united \with other technologies, the use and effect of television can be affected by other media as a result of the convergence of different media (Funk, 2001 ). In this study, for example, I moderated the effect sizes of television viewing. The correlations and factor loadings were not large, although they are statistically significant. The inclusion of other mass media may result in larger influences of children's viewing on their academic achievement to some extent. In light of these arguments, other forms of mass media, in addition to television, should be considered in future studies.


Anderson, D. R., & Collins, P. A. ( 1988). The impact on children 's education: Television 's influence on cognitive development. Washington, DC: U.S. Department of Education.

Anderson, D. R., Huston, A. C., Schmitt, K. L., Linebarger, D. L., & Wright, J. C. (2001). Early Childhood television viewing and adolescent behavior. Monographs of the Society for Research in Child Development, 66(1, Serial No. 264).

Ball, S., & Bogatz, G. A. (1970). The first year of Sesame Street: An evaluation. Princeton, NJ: Educational Testing Service.

Bandura, A. (1994). Social cognitive theory of mass communication. In J. Bryant & D. Zillmann (Eds.), Media effects: Advances in theory and research (pp. 61-90). Hillsdale, NJ: Erlbaum.

Beentjes, J. W. J. (1989). Salomon's model for learning from television and book: A Dutch replication study. Educational Technology Research and Development, 37(2), 47-58.

Beenljes, J. W. J., & van der Voort, T. H. A. (1989). Television and young people's reading behavior: A review of research. European Journal of Communication, 4, 5 1-77.

Bogatz, G. A., & Ball, S. (1971). The second year of Sesame Street: A continuing evaluation (Vol. 1 & 2). Princeton, NJ: Educational Testing Service.

Bryne, B. M. (1994). Structural equation modeling with EQS and EQS/Windows. Thousand Oaks, CA: Sage.

Child Development Supplement to the Panel Study of Income Dynamics. (1997). Retrieved February 4, 2003, from http:// www.isr.umich.edu/src/child-developmcnt/

Collins, W. A. (1982). Cognitive processing in television viewing. In D. Pearl, L. Bouthilet, & J. Lazar (Eds.), Television and behavior: Ten years of scientific progress and implications for the eighties (Vol. 2; pp. 9-23). Washington, DC: Government Printing Office.

Collins, P. A. (1991, April). The impact of television on preschooler' perseverance and television viewing habits. Paper presented at the biennial meeting of the Society of Research in Child Development, Seattle, WA.

Funk, J. (2001, April). The effects of video games on youth. Paper presented at the Annual Meeting of the Society for Research in Child Development, Minneapolis, MN.

Greenfield, P. M. ( 1984). Mind and media: The effects of television, video games, and computers. Cambridge, MA: Harvard University Press.

Harrison, L. P., & Williams, T. M. ( 1986). Television and cognitive development. In T. M. Williams (Ed.), The impact of television: A natural experiment in three communities (pp. 87-142). Orlando, FL: Academic Press.

Healy, J. (1990). Chaos on Sesame Street. American Educator, 14(4), 22-29.

Huesmann, L. R. (1986). Psychological processes promoting the relation between exposure to media violence and aggressive behavior by the viewer. Journal of Social Issues, 42(3), 125-140.

Huston, A. C., & Wright, J. C. (1996). Television and socialization of young children. In T. M. MacBeth (Ed.), Tuning in to young viewers: Social science perspectives on television (pp. 37- 60). Thousand Oaks, CA: Sage.

Kooltra, C., & van der Voort, T. H. A. (1996). Longitudinal effects of television on children's leisure-time reading: A test of three explanatory model. Human Communication Research, 23, 4-35.

Koshal, R. K., Koshal, M. A., & Gupta, A. K. (1996). Academic achievement and television viewing by eighth graders: A quantitative analysis. Applied Economics, 28(8), 919-928.

MacBeth, T. M. (1996). Indirect effects of television: Creativity, persistence, school achievement, and participation in other activities. In T. M. MacBeth (Ed.), Tuning in to young viewers: Social science perspectives on television (pp. 149-220). Thousand Oaks, CA: Sage.

Peterson, J. L., & Zill, N. (1986). Marital disruption, parent- child relationships, and behavioral problems in children. Journal of Marriage and the Family, 48(2), 295-308.

Polit, D. (1998). The positive behavior scale. New York: Basic Books.

Salomon, G. (1984). Television is "easy" and print is "tough": The differential investment of mental effort as a function of perceptions and attributions. Journal of Educational Psychology, 76, 647-658.

Singer, D. G., & Singer J. L. (2001). The popular media as educators and socializers of growing children. In D. G. Singer & J. L. Singer (Eds.), Handbook of children and the media (pp. 1-6). Thousand Oaks, CA: Sage.

Singer, J. L. (1980). The power and limits of television: A cognitive-affective analysis. In P. Tannenbaum (Ed.), The entertainment function of television (pp. 31 -65). Hillsdale, NJ: Erlbaum.

Singer, J. L., & Singer, D. G. (1983). Psychologists look at television. American Psychologist, 38(7), 826-834.

Singer, J. L., & Singer, D. G. (1986). Family experiences and television viewing as predictors of imagination, restless, and aggression. Journal of Social Issues, 42, 107-124.

Singer, J. L., Singer, D. G., & Rapaczynski, W. S. (1984). Family patterns and television viewing as predictors of children's belief and aggression. Journal of Communication, 34(1), 73-89.

Stein, A. H., & Fredrich, L. K. (1975). The impact of television on children and youth. In E. M. Hetherington (Ed.), Review of child development research (Vol. 5, pp. 183-256). Chicago: University of Chicago Press.

Suedfeld, P., Little, B. R., Rank, A. D., Rank, D. S., & Ballard, E. J. (1986). Television and adults: Thinking, personality, and attitudes. In T. M. Williams (Ed.), The impact of television: A natural experiment in three communities (pp. 361-394). Orlando, FE: Academic Press.

Valkenburg, P. M., & van der Voort, T. H. A. (1994). Influence of television on day-dreaming and creative imagination: A review of research. Psychological Bulletin, 116(2), 316-339.

Woodcock, R. W., & Johnson, M. B. (1989). Tests of achievement, standard battery: Form B. Chicago: Riverside.

Zill, N., Davis, E., & Daly, M. (1994). Viewing of Sesame Street by preschool children in the United States and its relationship to school readiness. Rockville, MD: Westat.

Received March 11, 2004


Visiting International Professional Program

Michigan State University

Address correspondence to Nary Shin, Visiting International Professional Program, Michigan State University, #1 International Center, East Lansing, MI 48824; [email protected] (e-mail).