Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Journal of Communication Technology and Human Behaviors
(2013) 1: 1-24
doi:10.7726/jcthb.2013.1001
Research Article
______________________________________________________________________________________________________________________________
*Corresponding e-mail: [email protected]
Indira Gandhi Institute of Development Research, Mumbai, India
1
Television Exposure and Academic Skills of Children:
New Findings from India
Ashish Singh* and Sarthak Gaurav
Received 10 September 2012; Published online 20 October 2012
© The author(s) 2012. Published with open access at uscip.org
Abstract The studies associating television viewing to performance are primarily based on developed countries. The
effect of television on children in developing countries, like India, might be different from that of the
developed countries, owing to cultural and socioeconomic differences. Additionally, in the Indian case, no
study has systematically analyzed the relationship between children’s academic skills and television viewing.
We examined the association between academic skills (Reading, Mathematics and Writing) of Indian children
and television viewing using a sample of children (aged 8-11 years) who were in school or had ever attended
school, from a universe of all the households included in a micro unit recorded nationally representative
survey (IHDS, 2004-05). Relevant socioeconomic, demographic, parent/household-level controls were
included in the multivariate analysis. The impact of content of the television programs was also controlled in
the analysis. The analysis was done at two levels: first, for the full sample; and second for subsamples, formed
first by sex and second by economic status. Findings suggested positive correlation between overall viewing
and the reading skills of boys and the mathematics skills of girls, but negative correlation with the writing
skills of both. The correlation of educational content with academic skills also differed by gender, with a
positive correlation with the three skills of girls but a positive correlation with only the writing skills of boys.
Moreover, educational content was positively correlated with the skills of children belonging to the lowest
wealth class but had no correlation with the skills of the children belonging to the highest wealth class.
Keywords: Television exposure; Academic abilities; Children; India
1. Introduction
The research on possible relationships between exposure to television and a child’s development
dates back to 1950s (Greenstein, 1954). The bulk of research on the topic points to the deleterious
effects of children’s television viewing on outcomes such as obesity, inactivity, attentional
problems, sleep disorders, aggression, and high-risk behavior (Liebert, 1986; MacBeth, 1996;
Robinson et al., 1998; Sargent et al., 2001; Villani, 2001; Earles et al., 2002; Dalton et al., 2003;
Christakis et al., 2004; Sargent et al., 2004). However, there are a few studies which found that
television viewing can be good for children. An extensive review of this body of literature can be
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
2
found in Kondo and Steemers (2010). These conflicting perspectives make it difficult to generalize
the influence of television viewing on child outcomes.
The generalization is even more difficult in the domain of cognitive ability where the evidence is
considerably more ambiguous; some studies have shown overall television viewing time and
viewing of non-educational programs to be harmful for children’s school readiness and cognitive
outcomes (Friedrich and Stein, 1973; Gadberry, 1980; Singer, 1990; Voort and Valkenburg, 1994;
MacBeth, 1996; Geist and Gibson, 2000; Wright et al., 2001; Hancox et al., 2005; Sharif and Sargent,
2005; Zimmerman and Christakis, 2005), but others have shown educational television to be
beneficial, with a number of experimental studies reporting that children who watched episodes of
educational television demonstrated improvements in educational domains immediately
afterwards (Bogatz and Ball, 1971; Singer and Singer, 1981; Davis, 1989; Hall et al., 1990; Rice et al.,
1990; Huston, 1992; MacBeth, 1996; Singer and Singer, 1998; Crawley, 1999; Anderson et al., 2001;
Naigles and Mayeux, 2001; Wright et al., 2001; Zill, 2001; Buckingham and Sefton-Green, 2004;
Fisch, 2005).
Further, there are few studies that found a positive influence of overall viewing time (including
non-educational programmes) on educational achievement: Schramm et al. (1961) reported that
heavy television viewing was associated with higher mental ability in young children; Johnson et al.
(1983) found higher math grades associated with a preference for sports, family, game and cartoon
shows; Gentzkow and Shapiro (2006) found that children who watch television perform marginally
better in reading and general knowledge at school. There are also studies which did not find any
relationship between time spent watching educational TV and educational domains, particularly
reading (Huston et al., 1999; Lyle and Hoffman, 1972; Schramm et al., 1961).
The ambiguity gets even more complicated by the results of studies that found different influences
in different settings or on different abilities. For example, Zimmerman and Christakis (2005) found
a detrimental effect of overall TV viewing on reading recognition, reading comprehension and
memory for digit scan for television viewing before age 3 but a beneficial effect of TV viewing on
reading recognition at ages 3 to 5 years. Schramm et al. (1961) reported that heavy television
viewing was associated with higher mental ability in young children and lower mental ability in
adolescents, but they found no relationship between viewing time and ability among older,
preadolescent children. Fetler (1984), Huston and Wright (1996) and Wright et al. (2001) found
“effect modification” of the effects of television viewing on educational outcomes by the
socioeconomic status of parents. Jordan (2004) proposed an ecological paradigm in which the
effects of television viewing might be different within distinct sociocultural niches and received
support from Joensson (1985) which reported that “high achievers” used television as a
complement to school learning whereas “low achievers” used television as a substitute for it. There
are also studies like Johnson et al. (1983), Lemish and Rice (1986), Singer and Singer (1998) and
Sharif and Sargent (2005), which reported positive influence of parental restriction and adult
involvement on academic achievement. Anderson et al. (2001), Wright et al. (2001), and Linebarger
and Walker (2005) argued that content is a stronger predictor of developmental outcomes than
total time spent watching television. Last but not least, Lablonde (1966), Strouse and Buerkel-
Rothfuss (1987) and Anderson et al. (2001) found that increased television screen time as well as
media content had different effects on girls’ and boys’ attitudes, behaviours and grades. If the
above review of literature is considered carefully, it indicates that the studies exploring the
relationship between television exposure and school performance have focused primarily on
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
3
exposure during the preschool years (Huston et al., 2007) and have yielded conflicting results.
Given the plethora of contesting evidence in the literature it became imperative to ponder over the
value addition of our study over and above the existing ones.
1.1 Importance of Our Study
It can be observed that all the aforementioned studies are based on developed countries, primarily
the United States. Our literature search revealed a dearth of studies on relationship between
television viewing and academic performance of children in developing countries. We have reasons
to expect that the effect of exposure to television on children in developing countries might be
different for reasons mentioned below.
In developing countries (for example, India), children either do not have a room of their own or, if
they have a room, they do not have a separate television for themselves. Since the family sizes are
relatively larger (than developed countries) and the number of rooms is limited, most of the family
members, including children, watch television together (in the sample, the average number of
individuals per room per household was 3.3). So, the question of small children watching “R-rated”
programmes or adult programmes becomes irrelevant. To be specific, watching adult programmes
(together) is a taboo in India and the television is normally kept in a common room where
everybody can watch it. Therefore, it is highly improbable that children would be able to watch “R-
rated” programmes. Also, children in developed countries have multiple choices (other than
watching television, for example, piano lessons), some of which might be superior to watching
television, for spending their leisure time. Unfortunately, the same is not true for the majority of
children from developing countries. Therefore, the relationship between television exposure and
academic skills of children should be examined in proper context (Chernin and Linebarger, 2005, p.
688) and the present study is an attempt in this direction.
Moreover, most of the studies discussed above suffer from some shortcoming or other. The
problems range from: (i) small sample sizes (for example, Clarke and Kurtz-Costes (1997) is based
on a sample size of 30 preschool children); (ii) nonrandom and non-representative samples (for
example, Sharif and Sargent (2005), which is one the most complete and methodologically sound
studies (based on a sample of 4508 students), has 95% white students in the sample with most of
the parents having completed high school); (iii) inadequate controls for content, personality and
parenting characteristics (Chernin and Linebarger 2005, p.688); (iv) self-reported or parental-
reported measure of school performance (Sharif and Sargent, 2005); and, (v) problems of
measuring television viewing accurately, where viewing measures sometimes consist of one or a
few questions with dubious validity asking individuals how much time they spend watching
television (Wright et al., 2001, p. 1350).
Our study (to the best of our knowledge) is the first study on India that systematically analyzed the
relationship between academic skills and exposure to television for children and does not suffer
from any of the aforementioned problems. Two studies that attempted to establish some relation
between television viewing and child behavioural/health outcomes in the Indian context need to be
mentioned. Gupta et al. (1994) studied a non-representative sample of 250 children (aged 3 to 10
years) in a hospital of an Indian city (Jaipur in the state of Rajasthan), and found an increase in
weight for 20% of children, a decrease in physical activity for 30% of cases, decreased interest in
study for 18% of children and a decrease in school performance for 10% of children. Since they
neither controlled for any background characteristics (demographic, socioeconomic, family, etc.)
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
4
nor conducted any statistical tests or analysis, their results cannot be considered reliable. Arya
(2004) studied a sample (non-representative) of 150 children (aged 6 to 12 years) from two
government and two private schools in a block of Kangara (a city in the Indian state of Himachal
Pradesh) and examined the impact of television viewing on gender-based differences regarding
democratic, social, economic, hedonistic, health, religious, aesthetic and knowledge values of
children. The results were mixed with gender-based differences apparent in a few of the values and
not in others. These results were again doubtful because the analysis did not control for relevant
background characteristics and lacked methodologically sound statistical analysis.
In the present study, we used a nationally representative sample of more than 12,000 children to
analyze the association between television viewing time and observed reading, mathematics and
writing skills. The sample of children included all the children in the age group of 8 to 11 years who
were in school or ever attended school, from a universe of all the households included in a micro
unit recorded nationally representative survey conducted in 2004-05. Relevant socioeconomic,
demographic, parent/family-level and household-level controls were included in the multivariate
analysis. The impact of content of the television programs was also controlled in the analysis. For
the first time in this stream of literature, we controlled for other forms of media exposure, namely,
newspapers. The analysis was done at two levels using ordinal logistic regressions. The first level
involved analysis on the full sample. The second level of analysis involved the same examination at
two sub-levels: first, on subsamples formed by sex of children, and second, on subsamples formed
by economic status of the households to which children belonged. The results were mixed and
correlation of television on observed academic skills of children varied by gender, socioeconomic
status and by academic skill itself.
2. Data and Methods
2.1 Data
We used the publicly available data from the Indian Human Development Survey (IHDS), conducted
by the National Council of Applied Economic Research, New Delhi, India in collaboration with the
University of Maryland, in 2004-051. The survey is a micro unit recorded, nationally representative
survey based on a stratified multistage sampling procedure. The survey was spread over 33 states
and union territories of India and covers 26,734 households in rural and 14,820 households in
urban areas respectively. This survey was unique in the sense that it was designed to measure
different dimensions of human development with modules on education, health, employment,
income, gender empowerment and media exposure. This study made use of the modules on: (i)
media exposure which reports the extent and type of media exposure for men, women and children
of every household surveyed, and (ii) education which assesses reading, writing and mathematics
skills for children aged 8 to11 years.
2.2 Outcome Measures
The outcome measures are the results of tests conducted for measuring reading, mathematics and
writing skills of children. The tests were administered at home in order not to miss children who
are absent from school and were designed in a way to measure basic skills with relative ease of
administration and with low anxiety levels on the part of children. For this module IHDS worked
1 The data and the questionnaires are available at http://www.icpsr.umich.edu/icpsrweb/ICPSR/studies/22626.
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
5
with PRATHAM, which has pretested similar tools on more than 250,000 children in India.
PRATHAM is a voluntary organization that has worked in the field of elementary education in India
for many years and has developed simple assessment tools to measure the effectiveness of their
training programs. The tests were simple and intuitive and were translated (with similar difficulty
levels) into 13 languages in addition to English and the children were asked to take the test in
whichever language they were most comfortable in (Desai et al., 2010, p.79). Interviewers were
trained by PRATHAM volunteers using specifically developed films to differentiate between things
like a child’s shyness and inability to read and to develop rapport with children.
The focus was on children aged 8 to 11 years because “all of these children should have acquired
the basic reading, writing, and arithmetic skills” (Desai et al., 2010, p.79). Children’s reading skills
were divided into five categories: (i) cannot read at all; (ii) can read letters but not form words; (iii)
can put letters together to read words but not read whole sentences; (iv) can read a short
paragraph for 2-3 sentences but not fluent enough to read a whole page; and (v) can read a one
page short story. The arithmetic skills were divided into four categories: (i) cannot read numbers
above 10; (ii) can read numbers between 10 and 99 but not able to do more complex number
manipulations; (iii) can subtract a two digit number from another; and (iv) can divide a number
between 100 and 999 by another number between 1 and 9. In terms of writing, the children were
tested for whether they were able to write a simple sentence, such as, “My mother’s name is
Madhuben”, with two or fewer mistakes (Desai et al., 2010). Though the module was administered
to 12,837 children, the information was consistent for 12,198 children (more than 95% of the total)
and these were the children who have been included in our study.
2.3 Explanatory Variables
The explanatory variable of primary interest is time spent by a child watching television. It was
measured as the number of hours per day (on an average day) a child watches television (0 = less
than 1/2 hour, 1 = 1/2 – 2 hours and 2 = more than 2 hours). The categories were formed for
meaningful interpretation and discussion. They were formed by taking into account the distribution
of television viewing time across children. Since, the survey reported less than or equal to half an
hour (per day) as zero, we had no choice but to take half an hour per day of viewing as the lowest
cutoff. Another explanatory variable of interest is how often a child watches educational content
(includes News, talk shows, educational or agricultural shows) and was categorized as “never”,
“sometimes” and “regularly” as reported in the survey. This variable for educational content was
included to separate the effect of educational television viewing from the effect of overall television
viewing. Of note is the fact that the survey itself provides the above categorization of the
educational content variable and we have used it as provided by the survey. Since the survey does
not provide any information on the number of hours of watching educational content, we do not
have any other option but to follow the survey’s categorization.
There are factors other than television viewing, notably family/parental, demographic,
socioeconomic, and some other characteristics that influence the outcomes of interest. An immense
body of past research has demonstrated that all these variables exert significant influence on child’s
academic outcomes (LaBlonde, 1966; Fetler, 1984; Joensson, 1985; Dombusch et al., 1987; Huston
et al., 1990; Steinberg et al., 1992; Truglio et al., 1996; The Probe Team, 1999; Anderson et al., 2001;
Wright et al., 2001; Jordan, 2004; Govt. of India, 2006; Sharif and Sargent, 2006; Desai et al., 2008a;
Desai and Kulkarni, 2008). Control variables included in the multivariate models were:
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
6
i. Highest educational attainment of adult female in the household – Highest standard
completed by female adult in the household (continuous variable)
ii. Highest educational attainment of adult male in the household – Highest standard
completed by male adult in the household (continuous variable)
iii. Age of the child – In years (continuous variable)
iv. Sex of the child – Male or female
v. Caste – Scheduled Castes/Scheduled Tribes (SC/ST), Other Backward Castes (OBC) and
other socioeconomically advanced castes (General/Upper castes). Individuals belonging to
SC/ST category are historically disadvantaged and have suffered severe social exclusion and
discrimination.
vi. Religion – Hindu (majority religious group in India), Muslim and Other religions
vii. Economic status – Measured by household “assets score” which is reported in the survey
and is measured on a scale of 30 household consumer goods and housing assets (continuous
variable).2
viii. Place of residence – Household classified as being in an urban or a rural area
ix. Newspaper reading by child – To capture exposure to other forms of media and is measured
by how often a child reads newspaper (never; sometimes; regularly).
It is important to point out that none of the studies covered in the literature survey controlled for
other forms of media exposure when analyzing the impact of television viewing on academic
performance of a child. We believed that it is important to control for other forms of media
exposure while evaluating the impact of one particular media on a child’s academic skills. Because
of the aforesaid rarity, as well as to capture the impact of watching educational content, we
estimated three sets of models for every outcome variable; (i) a basic model with the above
mentioned explanatory variables except educational content and newspaper reading habit of child,
(ii) a second model with the explanatory variables of the basic model and educational content
viewing by the child, and (iii) a final model with all of the explanatory variables.
2.4 Statistical Analysis
Since our outcome variables are ordinal, we used ordinal logistic regression analysis to calculate
odds ratios (ORs) for the relationship between observed academic skills and television use
variables while adjusting for family, demographic and socioeconomic characteristics. This model
gives cumulative ORs modeling the probability of being in any higher category on the academic
skills given a baseline category. With an ordered outcome variable, these models had the advantage
of retaining information that would be lost by combining the data into two arbitrary groups, as is
done using logistic regression.
Since for each dimension of academic skills of the N children, we had J achievement levels naturally
ordered in a meaningful way (ranked) for the ordinal outcome variable, the ordinal logistic model
can be specified as follows (Desai et al, 2008a):
iii xy εβ +=∗
)1(
kyi = if
∗
− ≤ ik y1δ kδ< for ,...,1=k J ;,...,1; Ni =
∞=−∞= ko δδ ; )2(
2 The assets score is based on 30 dichotomous items measuring household possessions and housing quality. The
details are available in Desai et al. (2008b, p.18). The measure of assets has a Cronbach’s reliability coefficient
alpha of 0.914.
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
7
where the ordered outcomes are modeled to arise sequentially as a latent variable, y*, with lower y*
indicating lower level of achievement (Cameron and Trivedi, 2005). In our case, y* is an unobserved
measure of child’s academic skills as indicated by measured test scores in reading skills,
mathematics skills and writing skills.
Though the tests were administered by trained interviewers to specifically distinguish among
children at varying levels of reading, mathematics and writing ability, the same child may well be
classified, for example, by one interviewer as being able to read letters and not words and by
another interviewer as being able to put the letters together in words. So, the outcome variable in
this case was better classified as a propensity to read rather than a specific skill level (Desai et al.,
2008a). Observed reading levels (as coded in the survey responses) are only observed; yi are tied to
this latent variable (yi*) (which is not observed) by the measurement model where the underlying
events are cumulative logits:
( )readnotdoesyi 1= if
∗≤−∞= io yδ 1δ< )3(
( )lettery i 2= if
∗≤ iy1δ 2δ< )4(
( )wordyi 3= if
∗≤ iy2δ 3δ<
)5(
( )paragraphyi 4= if
∗≤ iy3δ4δ< )6(
( )storyyi 5= if
∗≤ iy4δ
∞=< 5δ )7(
For mathematics skills the observed levels were: yi =1 (does not recognize written numbers), yi = 2
(can read numbers), yi = 3 (can subtract a two digit number from another two digit number) and yi
= 4 (can divide a three digit number with a one digit number).
Since the outcome variable for writing ability was dichotomous, the model became a simple logistic
regression (logit) model with yi = 1 (cannot write a simple sentence with two or less mistakes) and
yi = 2 (can write a simple sentence with two or less mistakes).
As some existing studies on the impact of television viewing on children’s academic performance
documented differential impact by gender and socioeconomic status, we carried out the analysis at
two levels. At the first level, the estimation was done on the full sample of children. At the second
level, the full sample was divided into subsamples of male and female children with each subsample
analyzed separately, followed by division of the full sample into tertiles based on economic status
(as measured by household assets) and the analysis carried out separately for the poorest and
richest tertiles.
3. Results
3.1 Descriptive Statistics
Table 1 presents the distribution of television viewing, family, demographic, and socioeconomic as
well as newspaper reading characteristics of the children. It also reports the mean and standard
deviation of the observed reading, mathematics and writing skills by the aforesaid characteristics.
It can be observed that the majority (43%) of children fell into the category of half an hour to 2
hours per day of television watching, but when it came to watching of educational programmes,
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
8
about 76% of children never watched educational content (news, talk shows and other educational
or agricultural programmes). Also, a little more than half of the children belonged to households
where there was no adult female who had completed even one year of schooling. The condition was
little better when it came to the education of male adults in the households, where more than 36%
of children belonged to households with at least one adult male with schooling between 6-10 years.
Another observation of interest was related to the newspaper reading habit of children. About 78%
children “never” read newspapers compared to only 6% who read newspapers “regularly”.
Table 1 Observed Academic Skills by Background Variables of Indian children, IHDS 2004-05
Reading Mathematics Writing N (%)
Mean Std. Dev. Mean Std. Dev. Mean Std. Dev.
Television viewing by children
(hours/day)
<= 1/2 2.3 1.4 1.3 1.0 0.6 0.5 40.9
1/2 – 2 2.8 1.3 1.7 1.0 0.8 0.4 43.3
> 2 3.0 1.2 1.8 1.0 0.7 0.4 15.8
Educational content viewing by
children
Never 2.5 1.4 1.5 1.0 0.7 0.5 76.1
Sometimes 3.0 1.2 1.9 1.0 0.8 0.4 18.5
Regularly 3.1 1.1 2.0 1.0 0.8 0.4 5.4
Highest schooling of female
adult in household (years)
No formal schooling 2.2 1.4 1.3 1.0 0.6 0.5 50.5
1 – 5 2.7 1.3 1.6 1.0 0.7 0.5 15.8
6 – 10 3.1 1.1 2.0 0.9 0.8 0.4 24.7
> 10 3.4 0.9 2.2 0.8 0.9 0.3 9.0
Highest Schooling of Male
Adult in Household (years)
No formal schooling 2.1 1.4 1.2 1.0 0.6 0.5 28.0
1 – 5 2.4 1.4 1.4 1.0 0.6 0.5 16.7
6 – 10 2.8 1.2 1.7 1.0 0.7 0.4 36.4
> 10 3.2 1.1 2.1 0.9 0.8 0.4 18.9
Age of Child (years)
8 2.1 1.3 1.2 0.9 0.6 0.5 25.1
9 2.6 1.3 1.5 1.0 0.7 0.5 22.2
10 2.8 1.3 1.7 1.0 0.7 0.4 33.5
11 3.0 1.2 1.9 1.0 0.8 0.4 19.2
Sex of Child
Male 2.7 1.3 1.6 1.0 0.7 0.5 52.9
Female 2.6 1.4 1.5 1.0 0.7 0.5 47.1
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
9
Caste
General 2.9 1.2 1.9 1.0 0.8 0.4 30.8
OBC 2.6 1.4 1.5 1.0 0.7 0.5 40.0
SC/ST 2.4 1.3 1.3 1.0 0.6 0.5 29.2
Religion
Hindu 2.7 1.3 1.6 1.0 0.7 0.5 79.1
Muslim 2.3 1.4 1.4 1.0 0.6 0.5 13.8
Others 2.8 1.3 1.8 1.0 0.7 0.4 7.1
Economic Status of Household
(Asset Quantiles)
1 2.1 1.4 1.2 1.0 0.6 0.5 36.7
2 2.6 1.3 1.6 1.0 0.7 0.5 32.6
3 3.2 1.0 2.1 0.9 0.8 0.4 30.7
Place of Residence
Rural 2.5 1.4 1.5 1.0 0.7 0.5 69.9
Urban 2.9 1.2 1.9 1.0 0.8 0.4 30.1
Newspaper Reading (child)
Never 2.5 1.3 1.5 1.0 0.7 0.5 78.1
Sometimes 3.0 1.1 1.9 1.0 0.8 0.4 15.7
Regularly 3.3 1.1 2.1 0.9 0.9 0.3 6.2
Overall 2.6 1.3 1.6 1.0 0.7 0.5 100
Notes: (1) Total number of observations = 12,198; (2) Reading Skills: cannot read at all = 0; can read
letters but not form words = 1; can put letters together to read words but not read whole sentences =
2; can read a short paragraph but not fluent enough to read whole page = 3; and can read a one page
short story = 4; (3) Mathematics Skills: cannot read numbers above 10 = 0; can read numbers between
10 and 99 but not able to do more complex number manipulation = 1; Can subtract a two digit number
from another = 2; Can divide a three digit number by another number between 1 and 9 = 3; (4)
Writing Skills: Cannot write a simple sentence with 2 or less mistakes = 0, writes with 2 or less
mistakes = 1; and (5) Educational content includes News, talk shows and other educational and
agricultural shows.
The pattern of academic skills by television viewing shows that all the observed academic skills
improved with both overall television watching time and watching of educational content. The
observed reading, mathematics and writing skills also improved with the increase in the level of
formal schooling of adult females and adult males in the household. However, the correlation with
educated adult females seems more than the correlation with educated adult males in the
household. It can be observed that the mean scores (in any academic skill) for a child was higher for
any highest level of schooling completed by adult females in the household compared to the same
level of schooling completed by adult males in the household.
The observed academic skills also improved with the age of the children. This was not surprising
since older children were more likely to have higher academic skills compared to their younger
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
10
counterparts for the same set of tests being administered. The male children performed relatively
better than their female counterparts but the difference was very small. Further, the performance
of children in all the academic dimensions by social and economic status of the household was as
expected. Region of residence had a bearing on the test scores as well, with the urban children
having higher average scores in the three skills than their counterparts. It can also be seen that the
observed reading, mathematics and writing skills systematically increased with the increase in time
spent reading newspapers.
While the descriptive statistics presented above are of interest and give an idea of the general
pattern of distribution of academic skills by background characteristics, they do not control for
other independent (explanatory) variables while investigating the role of a particular independent
variable in explaining variations in a child’s academic performance. It therefore becomes important
to pay more attention to the multivariate results that control for other factors while examining the
effect of a particular independent variable on an outcome variable.
3.2 Multivariate Results – Full Sample
The main multivariate results for the full sample are documented in Table 2, which reports the
cumulative odds of being in any higher level on the academic skills given a baseline level.
3.2.1 Reading Skills
The first estimated model (Model 1r) showed a positive correlation between total television
watching time and the observed reading skills of children. It did not include the controls for
television content and other forms of media exposure. It can be observed that the odds of achieving
any higher level of reading skills for children who watched half an hour to two hours of television
daily was 18% (significant at p < 0.01) more than that of the children who never watched or
watched less than half an hour of television per day. Similarly, children who watched more than
two hours of television daily had 15% (significant at p < 0.05) higher odds of achieving any higher
level compared to those who never watched or watched less than half an hour of television per day.
It can be noted that the correlation between reading skills and watching half an hour to two hours
of television daily was stronger (both in terms of magnitude and significance) than that with
watching more than two hours daily.
When we introduced the viewing of educational content in the model (Model 2r), the influence of
overall television watching was suppressed, but it still remained positive and significance levels did
not change. Watching of educational content “sometimes” also had a positive correlation
(significant at p < 0.05) with a child’s reading skills. Children who watched educational contents
“sometimes” had 11% higher odds of achieving any higher reading level compared to those who
“never” watched educational content. However, watching educational content “regularly” did not
have a significant correlation (though the nature was positive) with the reading skills of children. It
should be kept in mind that educational content here included news, talk shows and agricultural
shows in addition to educational shows. If another control in terms of exposure to other forms of
media (reading newspapers) is introduced (Model 3r), the positive correlation between overall
television watching and the reading skills diminishes further. Also, the significance level of positive
correlation between watching more than two hours of television daily and reading skills fell and
was significant at p < 0.1. With the introduction of newspaper reading habit of children, the effect of
watching educational content became insignificant (though the nature remained positive), while
reading newspapers “sometimes” had a strong positive correlation (significant at p < 0.01) with the
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
11
reading skills of children. However, reading newspapers “regularly” did not explain reading skills of
children significantly.
3.2.2 Mathematics Skills
As per the basic model (Model 1m), watching half an hour to two hours of television daily was
positively correlated (significant at p < 0.01) with the mathematics skills of children. The children
who watched this duration of television had 13% higher odds of achieving any higher mathematics
level than those who never watched or watched less than half an hour of television per day.
Conversely, watching more than two hours of television per day had a negative (but insignificant)
correlation with the mathematics skills.
Introducing the viewing of educational content (Model 2m) suppressed the positive correlation
between watching half an hour to two hours of television daily and mathematics skills (both in
terms of magnitude as well as significance). Also, the detrimental effect of watching more than two
hours of television increased minutely (the odds ratio changed from 0.97 to 0.94) but remained
insignificant. As in the case of reading skills, watching educational content “sometimes” had a
strong positive correlation (significant at p < 0.01) with the mathematics skills of children. Once
again, watching educational content “regularly” did not come out to be significant (though the odds
ratio was greater than 1), as was the case in reading skills. When newspaper reading habit was also
controlled for (Model 3m), the positive correlation between watching half an hour to two hours of
television daily and mathematics skills plummeted further (now significant at p < 0.1). The negative
correlation between watching more than two hours of television and mathematics skills increased
even more, but still remained insignificant. Unlike the case of reading skills, the inclusion of
newspaper reading habit of children did not affect the significance of watching educational content
“sometimes”, but it decreased the odds of achieving a higher level from 19% to 14%. Reading
newspapers “sometimes” had a strong positive correlation (significant at p < 0.01) with the
mathematics skills of children, a case similar to that of reading skills. However, reading newspapers
“regularly” did not have any significant effect on the mathematics skills of children.
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
12
Table 2: Odds ratios from the ordinal logistic regression: academic skills (full sample)
Reading Mathematics Writing
Model 1r Model 2r Model 3r Model 1m Model 2m Model 3m Model 1w Model 2w Model 3w
TV watching, 1/2 – 2 hrs/day 1.181*** 1.151*** 1.149*** 1.128*** 1.089** 1.085* 1.126** 1.059 1.062
(0.049) (0.049) (0.049) (0.047) (0.047) (0.047) (0.056) (0.054) (0.055)
TV watching, > 2 hrs/day 1.149** 1.121** 1.112* 0.971 0.939 0.929 0.837** 0.787*** 0.783***
(0.065) (0.065) (0.064) (0.055) (0.054) (0.053) (0.061) (0.058) (0.058)
Highest schooling of female 1.067*** 1.067*** 1.067*** 1.064*** 1.063*** 1.063*** 1.106*** 1.104*** 1.103***
adult in household (yrs.) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) (0.007) (0.007) (0.007)
Highest schooling of male 1.030*** 1.030*** 1.029*** 1.035*** 1.035*** 1.034*** 1.024*** 1.024*** 1.023***
adult in household (yrs.) (0.004) (0.004) (0.004) (0.004) (0.004) (0.004) (0.005) (0.005) (0.005)
Age (yrs.) 1.597*** 1.595*** 1.593*** 1.598*** 1.597*** 1.594*** 1.415*** 1.413*** 1.409***
(0.026) (0.026) (0.026) (0.026) (0.026) (0.026) (0.029) (0.029) (0.029)
Female 0.850*** 0.851*** 0.851*** 0.747*** 0.748*** 0.749*** 0.844*** 0.844*** 0.847***
(0.029) (0.029) (0.029) (0.025) (0.025) (0.025) (0.035) (0.035) (0.036)
OBC 0.847*** 0.850*** 0.853*** 0.797*** 0.802*** 0.806*** 0.831*** 0.839*** 0.843***
(0.036) (0.036) (0.036) (0.034) (0.034) (0.034) (0.046) (0.047) (0.047)
SC/ST 0.738*** 0.740*** 0.741*** 0.668*** 0.671*** 0.674*** 0.769*** 0.774*** 0.778***
(0.036) (0.036) (0.036) (0.032) (0.032) (0.033) (0.048) (0.048) (0.049)
Muslim 0.589*** 0.592*** 0.591*** 0.645*** 0.649*** 0.648*** 0.717*** 0.723*** 0.723***
(0.030) (0.030) (0.030) (0.034) (0.034) (0.034) (0.046) (0.046) (0.046)
Other Religions 0.929 0.924 0.919 1.044 1.040 1.032 0.976 0.966 0.958
(0.062) (0.062) (0.062) (0.066) (0.065) (0.065) (0.088) (0.087) (0.087)
Household assets score 1.073*** 1.072*** 1.071*** 1.079*** 1.078*** 1.077*** 1.053*** 1.050*** 1.047***
(0.005) (0.005) (0.005) (0.005) (0.005) (0.005) (0.006) (0.006) (0.006)
Urban 0.968 0.966 0.969 1.075* 1.072* 1.077* 1.078 1.071 1.078
(0.041) (0.041) (0.041) (0.045) (0.045) (0.045) (0.059) (0.059) (0.059)
Watching educational 1.112** 1.073 1.189*** 1.138*** 1.310*** 1.220***
contents sometimes (0.053) (0.052) (0.054) (0.054) (0.085) (0.081)
Watching educational 1.105 1.097 1.009 1.014 1.420*** 1.344**
contents regular (0.089) (0.088) (0.076) (0.078) (0.167) (0.161)
Newspaper reading 1.193*** 1.251*** 1.371***
Sometimes (0.059) (0.062) (0.095)
Newspaper reading 0.982 0.934 1.147
Regular (0.085) (0.077) (0.145)
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
13
Notes: (1) Figures in parenthesis are robust standard errors; (2) ***p < 0.01, **p < 0.05, *p < 0.10; (3) The
reference categories for the independent variables are: For overall television watching – watching <
1/2 hours (hrs.)/day; Sex of child – male; Caste – general (upper castes); Religion – Hindu; Place of
residence – rural; Watching educational content – never; Reading newspapers – never; and
(4) We conducted a Wald test for joint significance of the coefficients in the models and a Brant test
for proportional odds assumption. For each of the three models (in each of the skills), the Wald test
provided sufficient evidence for rejecting the null, which is joint insignificance of the coefficients in
the model, thus justifying the inclusion of the independent variables in the model. One of the
assumptions underlying ordinal logistic regression is that the relationship between each pair of
levels of outcome variable is the same, that is, odds being the same across adjacent categories. This
proportional odds assumption or the parallel regression assumption was tested using the Brant test
(omnibus likelihood ratio test). Though the models (reading ability and mathematics) could not
reject the proportional odds assumption for specific coefficients of interest, the assumption was
rejected for all the coefficients together. Therefore, we conducted generalized ordinal logistic
estimation, which does not assume the proportional odds assumption, for reading ability and
mathematics ability. The results showed that for all the independent variables, the odds ratios for the
lowest level of outcome variable (for all outcome variables) to the higher ones, next lowest to the
remaining higher ones and so on, were very close to each other and also very close to the odds ratios
of the ordinal logistic estimation for both the reading ability as well as mathematics ability.
Moreover, the significance of coefficients remained the same (the results can be provided upon
request). Therefore, we keep the estimates of ordinal logistic regressions for a simple but meaningful
interpretation and discussion. Since there are only two levels of outcome for writing ability, the
proportional odds assumption does not apply.
3.2.3Writing Skills
The results of writing skills were more interesting and quite different from those of reading and
mathematics skills. In the absence of controls for educational content and newspaper reading
(Model 1w), watching half an hour to two hours of television daily had a positive and significant
correlation (at p < 0.05) with a child’s writing skills. But watching more than two hours of television
per day had a significant (at p < 0.05) detrimental correlation with the same skills.
When watching of educational content was introduced in the model (Model 2w), the positive
correlation between watching half an hour to two hours of television daily and the writing skills
became insignificant. The negative relationship between watching more than two hours of
television per day and the writing skill increased, both in terms of magnitude as well as level of
significance. The odds of writing a simple sentence with two or less mistakes decreased from 0.84
to 0.79 and were now significant at p < 0.01. In a manner different from the case of reading skills
and mathematics skills, watching educational content “sometimes” as well as watching educational
content “regularly” had a positive and significant correlation (at p < 0.01) with a child’s writing
skills.
With the introduction of newspaper reading habit of the children in the model (Model 3w), the
negative effect of watching more than two hours of television daily increased even further and
remained significant at p < 0.01. Also, unlike reading and mathematics skills, watching educational
content “sometimes” or “regularly” had a positive correlation with a child’s writing skills at p <0.01
and p < 0.05, respectively. It is also evident that the effect of reading newspapers was similar to the
case of reading and mathematics skills.
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
14
3.3 Multivariate Results – Boys vs. Girls
For the analysis by the sex of children, we divided the full sample of children into two subsamples
based on the sex of children. Since we already showed the importance of including educational
content as well as newspaper reading habit of children in the estimation, we have presented the
results (Table 3) of only the full models for reading, mathematics, and writing skills.
3.3.1 Reading Skills
The analysis by sex of children showed a very contrasting picture as far as the correlation between
television viewing and a child’s observed academic skills was concerned. If reading skills were
considered (Model 1m and Model1f), watching half an hour to two hours of television daily had a
positive and significant correlation (at p < 0.01) with the reading skills of boys, whereas it did not
affect the reading skills of girls significantly. When watching more than two hours of television daily
was taken into account, it affected the reading skills of boys positively and significantly (at p < 0.1)
but had an insignificant effect on the reading skills of girls.
Watching educational content did not have any significant correlation with the reading skills of
boys, but was significantly correlated with the reading skills of girls. Watching educational content
“sometimes” increased the odds of achieving any higher level of reading skills in girls by 26%
(significant at p < 0.01), whereas watching educational content “regularly” increased the odds of
achieving any higher level of reading skills by 23% (significant at p < 0.1). In both cases, the
reference category was of those girls who “never” watched educational content.
3.3.2 Mathematics Skills
The results on mathematics skills (Model 2m and Model 2f) also differed by sex of children. While
watching television was not correlated with the mathematics skills of boys, watching half an hour to
two hours of television daily had a positive and significant correlation (at p < 0.1) with the
mathematics skills of girls.
In a manner similar to that of reading skills, watching educational content was not significantly
correlated with the mathematics skills of boys, but watching it “sometimes” was positively and
significantly correlated with the mathematics skills of girls. Watching educational content
“sometimes” increased the odds of achieving any higher level of mathematics skills in girls by 18%
(significant at p < 0.01). For both boys and girls, watching educational content “regularly” was not
correlated significantly with mathematics skills.
3.3.3 Writing Skills
The results for male and female children (Model 3m and Model 3f) were somewhat similar as far as
writing skills were concerned. Watching half an hour to two hours of television daily neither
affected the writing skills of boys nor the writing skills of girls significantly, whereas watching more
than two hours of television daily was negatively and significantly correlated with the writing skills
of both boys and girls. However, the detrimental correlation was relatively higher for boys
compared to girls, where the boys who watched more than two hours of television daily were only
0.74 times (significant at p < 0.01) as likely to write a simple sentence with two or fewer mistakes
as the boys who did not watch or watched less than half an hour of television daily. The
corresponding figure was 0.82 (significant at p < 0.1) for female children.
Considering the effect of watching educational content, we found that watching educational content
“sometimes” was positively and significantly correlated with the writing skills of both boys as well
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
15
as girls. However, the positive correlation was relatively higher for girls compared to boys. The girls
who watched educational content “sometimes” had 26% (significant at p < 0.05) higher odds of
writing a simple sentence with two or fewer mistakes compared to the girls who “never” watched
educational content. The corresponding figure was 17% (significant at p < 0.1) for boys. Also, it was
noteworthy that watching educational content “regularly” did not have any significant correlation
with the writing skills of boys, while it had a positive and significant correlation (at p < 0.01) with
the writing skills of girls.
Table 3 Odds ratios from the ordinal logistic regression: academic skills by sex of child
Male Children Female Children
Reading
Model1m
Mathematics
Model2m
Writing
Model3m
Reading
Model1f
Mathematics
Model2f
Writing
Model3f
TV watching, 1/2–2 hrs/day 1.198*** 1.050 1.109 1.087 1.117* 1.009
(0.071) (0.062) (0.079) (0.067) (0.070) (0.076)
TV watching, > 2 hrs/day 1.147* 0.925 0.744*** 1.060 0.925 0.823*
(0.092) (0.073) (0.075) (0.089) (0.079) (0.089)
Highest schooling of female 1.057*** 1.046*** 1.100*** 1.077*** 1.081*** 1.106***
adult in household (yrs.) (0.007) (0.007) (0.010) (0.008) (0.008) (0.011)
Highest schooling of male 1.039*** 1.046*** 1.023*** 1.017*** 1.022*** 1.023***
adult in household (yrs.) (0.006) (0.006) (0.007) (0.006) (0.006) (0.008)
Age (yrs.) 1.629*** 1.666*** 1.432*** 1.556*** 1.521*** 1.386***
(0.037) (0.038) (0.040) (0.037) (0.036) (0.041)
OBC 0.910 0.830*** 0.899 0.784*** 0.773*** 0.786***
(0.053) (0.049) (0.069) (0.048) (0.048) (0.063)
SC/ST 0.713*** 0.660*** 0.778*** 0.771*** 0.686*** 0.775***
(0.047) (0.044) (0.069) (0.055) (0.049) (0.071)
Muslim 0.592*** 0.634*** 0.709*** 0.588*** 0.663*** 0.742***
(0.042) (0.046) (0.067) (0.044) (0.050) (0.069)
Other Religions 0.894 0.998 1.024 0.958 1.086 0.892
(0.079) (0.082) (0.128) (0.098) (0.105) (0.118)
Household assets score 1.057*** 1.072*** 1.045*** 1.089*** 1.084*** 1.049***
(0.007) (0.007) (0.008) (0.008) (0.007) (0.009)
Urban 0.870** 0.954 1.002 1.089 1.226*** 1.162*
(0.050) (0.056) (0.076) (0.067) (0.075) (0.093)
Watching educational 0.941 1.098 1.177* 1.255*** 1.183** 1.266**
contents sometimes (0.061) (0.070) (0.107) (0.093) (0.084) (0.121)
Watching educational 0.980 0.988 1.261 1.232* 1.040 1.422*
contents regular (0.111) (0.102) (0.215) (0.140) (0.121) (0.240)
Newspaper reading 1.258*** 1.278*** 1.516*** 1.127 1.230*** 1.231**
sometimes (0.085) (0.084) (0.146) (0.082) (0.091) (0.124)
Newspaper reading regular 1.116 1.111 1.161 0.875 0.788* 1.143
(0.128) (0.123) (0.199) (0.116) (0.097) (0.216)
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
16
Notes: (1) Figures in parenthesis are robust standard errors; (2) ***p < 0.01, **p < 0.05, *p < 0.10; (3)
The reference categories for the independent variables are: For overall television watching – watching
< 1/2 hours (hrs.)/day; Sex of child – male; Caste – general (upper castes); Religion – Hindu; Place of
residence – rural; Watching educational content – never; Reading newspapers – never.
3.4 Multivariate Results – by Economic Status
For analyzing the differential effect of television viewing on children by the economic status of the
households they belong to, we divided the full sample into three quantiles (tertiles) based on
household assets and compared the correlation between television watching and skills of the
children falling into the first tertile (henceforth referred as “poor”) to the correlation between
television watching and skills of children who fell into the third tertile (henceforth referred as
“rich”). Table 4 reports the results (odds ratios) of this comparison.
3.4.1 Reading Skills
Results of Model 1p and Model 1r need to be compared for observing the contrasts in reading skills
between the poor children and the rich children. Though the total television watching time was not
significantly correlated with the reading skills of children of either group, it can be observed that
watching educational content “sometimes” had a very significant positive correlation with the
reading skills of poor children, while it did not have any significant correlation with the reading
skills of rich children. Among the poor children, the odds of achieving any higher level in reading
skills for the children who watched educational content “sometimes” were 42% higher (significant
at p < 0.01) than the children who “never” watched educational content. Another finding worth
noticing here is that sex of children did not have any significant bearing on the reading skills of
children in the set of rich children but it had a significant correlation with the reading skills of
children falling into the poor category. Female children in the group of poor children were only 0.63
times (significant at p < 0.01) as likely to achieve any higher level of reading skills compared to
their male counterparts.
3.4.2 Mathematics Skills
Model 2p and Model 2r report the estimates for mathematics skills for the aforesaid two groups of
children. In mathematics skills, similar to the case of reading skills, the total television watching
time was not correlated significantly with the children of either group. Again, as in the case of
reading skills, it can be observed that watching educational content “sometimes” had a significant
and positive correlation with the mathematics skills of the poor children but failed to have any
significant correlation with the mathematics skills of rich children. Among the poor children, the
odds of achieving any higher level in mathematics skills for the children who watched educational
content “sometimes” were 39% higher (significant at p < 0.01) than the children who “never”
watched educational content.
3.4.3 Writing Skills
Unlike reading and mathematics skills, watching more than two hours of television daily was
negatively and significantly correlated with the writing skills of the children of both groups (Model
3p and Model 3r). Again, watching educational content “sometimes” had a very significant and
positive correlation with the writing skills of poor children but failed to have any significant effect
on the writing skills of rich children. Among the poor children, the odds of writing a simple
sentence with two or fewer mistakes for the children who watched educational content
“sometimes” were 73% higher (significant at p < 0.01) than the children who “never” watched
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
17
educational content. Compared to reading and mathematics skills, this positive correlation was
much higher.
Though the effect of newspaper reading had not been discussed in the analysis based on
subsamples, it is worthwhile to mention that reading newspapers “sometimes” had a significant and
positive correlation with the reading, mathematics and writing skills of children of all the
subgroups discussed so far, except for the reading skills of girls and rich children. For these two
subgroups, the correlation was positive but nonetheless not significant.
Table 4 Odds ratios from the ordinal logistic regression: academic skills by assets quantiles
Bottom one-third Top one-third
Reading
Model1p
Mathematics
Model2p
Writing
Model3p
Reading
Model1r
Mathematics
Model2r
Writing
Model3r
TV watching, 1/2 – 2 hrs/day 1.105 1.080 1.089 1.031 0.988 0.825
(0.073) (0.075) (0.085) (0.116) (0.104) (0.136)
TV watching, > 2 hrs/day 1.072 0.920 0.485*** 1.010 0.841 0.660**
(0.159) (0.136) (0.081) (0.121) (0.095) (0.116)
Highest schooling of female 1.103*** 1.103*** 1.127*** 1.049*** 1.040*** 1.079***
adult in household (yrs.) (0.012) (0.013) (0.017) (0.009) (0.008) (0.012)
Highest schooling of male 1.028*** 1.025*** 1.019** 1.025*** 1.051*** 1.025**
adult in household (yrs.) (0.007) (0.008) (0.008) (0.009) (0.009) (0.012)
Age (yrs.) 1.501*** 1.479*** 1.365*** 1.680*** 1.776*** 1.465***
(0.039) (0.039) (0.041) (0.054) (0.056) (0.068)
Female 0.694*** 0.633*** 0.776*** 1.100 0.890** 0.871
(0.037) (0.035) (0.049) (0.071) (0.055) (0.081)
OBC 0.786*** 0.734*** 0.803** 0.896 0.840** 0.879
(0.064) (0.063) (0.081) (0.065) (0.059) (0.093)
SC/ST 0.759*** 0.666*** 0.843** 0.641*** 0.724*** 0.638***
(0.066) (0.061) (0.081) (0.060) (0.065) (0.081)
Muslim 0.545*** 0.689*** 0.759*** 0.672*** 0.574*** 0.688***
(0.049) (0.064) (0.081) (0.068) (0.058) (0.093)
Other Religions 0.897 1.143 1.094 1.147 0.959 1.116
(0.109) (0.154) (0.182) (0.123) (0.088) (0.174)
Household assets score 1.111*** 1.114*** 1.078*** 1.043*** 1.057*** 1.030
(0.017) (0.017) (0.019) (0.013) (0.012) (0.019)
Urban 1.153 1.169 1.103 1.038 1.156** 1.243**
(0.112) (0.115) (0.124) (0.071) (0.078) (0.120)
Watching educational 1.427*** 1.385*** 1.732*** 1.009 1.093 1.055
contents sometimes (0.156) (0.147) (0.253) (0.076) (0.080) (0.115)
Watching educational 0.758 0.775 1.013 1.093 0.996 1.327
contents regular (0.192) (0.224) (0.312) (0.123) (0.103) (0.231)
Newspaper reading 1.575*** 1.500*** 1.593*** 1.127 1.258*** 1.266**
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
18
sometimes (0.156) (0.163) (0.217) (0.089) (0.094) (0.146)
Newspaper reading regular 1.892 0.921 1.472 0.941 1.017 1.219
(0.733) (0.339) (0.543) (0.097) (0.102) (0.193)
Notes: (1) Figures in parenthesis are robust standard errors; (2) ***p < 0.01, **p < 0.05, *p < 0.10; and
(3) the reference categories for the independent variables are: For overall television watching –
watching < 1/2 hours (hrs.)/day; Sex of child – male; Caste – general (upper castes); Religion –
Hindu; Place of residence – rural; Watching educational content – never; Reading newspapers –
never.
Further discussion on these results has been provided in the subsequent section which also
concludes this paper.
4. Discussion and Conclusion
As far as India is concerned, this paper provides the first systematic and reliable evidence on the
association between watching television and a child’s observed academic skills. The age group of
the children in the study varied from 8 - 11 years, which is a deviation from most of the previous
studies, which were based on preschool children. This age group was important from the
perspective of the different stages of cognitive development that are related to television watching
and that a child experiences (Piaget, 1969; Lemish, 2007). Among the different stages of cognitive
development experienced by children, the “concrete operational stage” is associated with the age
group of 7 to 12 years. This is the stage when children begin to engage in abstract thoughts that
allow them to understand the medium’s codes and conventions sufficiently to follow story lines.
They develop levels of perception (televisual literacy) that allow them to understand the chunks
and segments that constitute a television programme and how they are linked (Signorielli 1991,
p.28). From the age of 12, children are assumed to understand television in a similar way to adults
(Hodge and Tripp, 1986, p. 80-81; Lemish, 2007, p. 39). Since this is a stage where the attraction for
television is very high among children as they start to comprehend the programmes, it became
imperative to study the effect of watching television on this specific age group.
We studied the correlation between watching television and the reading, mathematics and writing
skills of children and found substantial variation in level and nature of correlation with the three
skills. In case of reading skills, watching television had a positive correlation irrespective of
controlling for educational content or other forms of media exposure. Results in the case of
mathematics skills were a little different, where watching television for small durations (1/2 – 2
hours daily) had a positive correlation, whereas watching more than two hours of television daily
had a detrimental but insignificant correlation. In contrast to reading and mathematics skills,
watching more than 2 hours of television had a strong and significant positive correlation with the
writing skills of children. These differential effects have also been documented in earlier studies:
Zimmerman and Christakis (2005) reported a beneficial effect (at age 3-5 years) of television for
reading recognition scores, but found modest adverse effects on other scores; Johnson et al. (1983)
associated lower grade (and I.Q.s) with more television watching but higher math grades with a
preference for family, game and cartoon shows (fifth to eighth grade children); Schramm et al.
(1961) correlated higher mental ability with heavy television viewing in young children but no
relationship between viewing time and ability among older, preadolescent children.
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
19
Watching of educational content did not have any significant correlation with the reading skills of
the children in the presence of effective controls, while watching educational content “sometimes”
was positively correlated with mathematics skills as well as writing skills. In the case of writing
skills, even watching educational content “regularly” had a positive correlation. The insignificant
correlation of watching educational content “regularly” with the reading and mathematics skills
should be seen in the light that “educational content” included news, talk shows and agricultural
shows in addition to the direct education programmes. If a child watched this content too much it
might have resulted in the neglect of regular studies, which may have led to insignificant difference
in the skills of children who did not watch educational content and the children who watched too
much of it. Since the exact contribution of the different components of educational content is
unknown, no further insightful analysis can be done. It must be mentioned here that there is ample
evidence in support of the positive correlation between educational content and a child’s academic
performance, as has already been discussed; we do not repeat it here.
The findings related to the differential impact of television on the academic skills of children
belonging to different sexes were more interesting. While overall television viewing was positively
correlated with the reading skills of boys, it did not have any significant correlation with the
reading skills of girls. Conversely, the overall television viewing time did not have any significant
correlation with the mathematics skills of boys, but watching 1/2 - 2 hours of television daily was
positively correlated with the mathematics skills of girls. The influence of watching educational
content on reading and mathematics skills also differed across boys and girls, where it did not have
any significant effect on boys, but watching “sometimes” was positively correlated with both the
reading as well as mathematics skills of girls. These contrasts are not surprising if seen in the light
of LaBlonde (1966), which found a significant negative relationship between television viewing and
I.Q. for girls but not for boys, Anderson et al. (2001), which related increased television screen time
to higher grades for boys but lower grades for girls and Sharif and Sargent (2006), who found that
girls were more vulnerable if there were more cable movie channels in the home, whereas boys
were more vulnerable if “R-rated” movie viewing was restricted (p. e1068-69). The difference in
media use patterns by gender (Huston et al., 1999) and the development of gender identities by the
age of 6-7 years with categorization of programmes by own distinctive tastes (Davis et al., 2000)
have been attributed as possible reasons for the differential impact of television viewing on the
academic skills of boys and girls. Only in the case of writing skills, the influence of watching
television was similar for both boys and girls, and watching more than 2 hours of television daily
had a strong detrimental correlation with writing skills, but watching educational content
“sometimes” had a significant positive association with writing skills.
We also deviated from most of the extant studies by analyzing the differential influence of watching
television on the academic skills of children belonging to different wealth classes. The separate
analysis stands justified if we note the results that watching educational content “sometimes” had a
significant positive correlation with the reading, mathematics and writing skills of children
belonging to the lowest wealth class but failed to have any significant correlation with any of the
three academic skills of the children belonging to the highest wealth class. These differences in the
influence of television on children with different socioeconomic status are in accordance to the
studies by Fetler (1984), Anderson et al. (2001) and Wright et al. (2001), who suggested that
television use by children of lower socioeconomic status may be particularly beneficial to their
school performance. In the study by Fetler (1984) among children in lower socioeconomic class
homes, those who watched little or no television had worse academic performance than those who
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
20
watched moderate amounts of television. Conversely, among children from higher socioeconomic
class homes, children who watched little or no television had better academic performance than
those who watched moderate amounts (Anderson et al., 2001; Jordan, 2004). It is worthwhile to
mention here that watching too much television (more than two hours daily) had a significant
negative correlation with the writing skills of both sets of children.
Though the effect of newspaper reading on children’s academic skills is not the focus of this study, it
is important to note that reading newspapers “sometimes” had a positive correlation with the
academic skills of children irrespective of their gender or socioeconomic status. It is also important
to note that heavy exposure of either educational content or reading newspapers failed to
significantly affect the academic skills of children, for which the possible reason of interference with
regular studies has already been discussed.
This study has several strengths, including systematic measure of television exposure, large sample
size, adequate and appropriate controls for a number of family, demographic and socioeconomic
characteristics and no geographic area limitation. Moreover, departure from self-reported
measures of academic performance as the main outcome is an added advantage of this study. At the
same time, our study is also subject to a few limitations. Correlations of television viewing and
achievement may be a function of other factors. Though we controlled for a host of demographic
and family characteristics (e.g., education, caste, religion and wealth) and the controls are informed
by earlier literature on this subject as well as the relevance of factors in Indian context, it is always
possible that unmeasured variables account for part of the associations. Another limitation comes
from the cross-sectional and non-experimental nature of our study. In the absence of longitudinal
evidence and appropriate counterfactuals, it becomes difficult to establish significant causality
between the variables concerned. Clearly, a major way to take this research agenda (the effect of
television viewing on academic skills of Indian children) forward will be to evaluate the effects of
television viewing on the academic skills of children in India under an experimental setup (e.g.,
randomized control trials). This will help in establishing the causal relationship between watching
of television and academic skills and will help in providing additional evidence on how television
viewing can impact the academic skills of Indian children. Although the study is subjected to the
above limitations, grounding of the analysis in appropriate theory, methodological soundness,
robustness of results and support for the results in existing literature justifies its contribution.
On a concluding note we will like to mention that demonizing television on the basis of incomplete
studies with inadequate controls seems unjustified. Watching television in an uncontrolled manner
might be detrimental to the academic skills of children, but watching educational content in a
regulated manner seems to have a positive correlation with children’s academic skills. As a final
remark, we quote:
“Television can be of general benefit to children. It can bring them into contact with aspects of
life they would not otherwise become aware of. It can provide a valuable tool in the home and at
school not simply to keep children occupied but also, if used appropriately, as a constructive way to
use their time… Television is not a ‘one-eyed monster’ lurking impishly in the corner of the living
room, kitchen or bedroom waiting to exert an evil influence over young members of the household.
It is a channel through which a range of entertainment, drama and learning can be obtained and
experienced and increasingly these days it is under the control of the viewer.”
-Gunter and McAleer [1997, p. xii-xiii]
.
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
21
Acknowledgement
The authors thank the referee and the editorial team for helpful comments.
References Anderson, D. R., Huston, A. C., Schmitt, K. L., Linebarger, D. L., Wright, J. C., Larson, R., 2001. Early childhood
television viewing and adolescent behaviour: the recontact study. Monographs of the Society for
Research in Child Development 66, i-viii + 1-154.
Arya, K., 2004. Time spent on television viewing and its effect on changing values of school going children.
Anthropologist 6, 269-271.
Bogatz, G., Ball, S., 1971. The Second Year of Sesame Street: A Continuing Evaluation. Educational Testing
Service, Princeton.
Buckingham, D., Sefton-Green, J., 2004. Structure, agency, and pedagogy in children's media culture, In:
Joseph Tobin (Ed). Pikachu's Global Adventure. Duke University Press, London.
Cameron, A. C., Trivedi, P. K., 2005. Microeconometrics: methods and applications. Cambridge University
Press, New York.
http://dx.doi.org/10.1017/CBO9780511811241
Chernin, A. R., Linebarger, D. L., 2005. The relationship between children's television viewing and academic
performance. Archives of Pediatrics and Adolescent Medicine 159, 687-689.
http://dx.doi.org/10.1001/archpedi.159.7.687
PMid:15997006
Christakis, D. A., Zimmerman, F. J., DiGiuseppe, D. L., McCarty, C. A., 2004. Early television exposure and
subsequent attentional problems in children. Pediatrics 113, 708-713.
http://dx.doi.org/10.1542/peds.113.4.708
PMid:15060216
Clarke, A. T., Kurtz-Costes, B., 1997. Television viewing, educational quality of the home environment, and
school readiness. The Journal of Educational Research 90, 279-285.
Crawley, A. M., Anderson, D. R., Wilder, A., Williams, M., Santomero, A., 1999. Effects of repeated exposures to
a single episode of the television program Blue's Clues on the viewing behaviors and comprehension of
preschool children. Journal of Educational Psychology 91, 630-638.
http://dx.doi.org/10.1037/0022-0663.91.4.630
Dalton, M. A., Sargent, J. D., Beach, M. L., Titus-Ernstoff, L., Gibson, J. J., Ahrens, M. B., Tickle, J. J., Heatherton, T.
F., 2003. Effect of viewing smoking in movies on adolescent smoking initiation: a cohort study. Lancet 62,
281–285.
http://dx.doi.org/10.1016/S0140-6736(03)13970-0
Davies, H., Buckingham, D., Kelly, P., 2000. In the worst possible taste: children, television and cultural value.
European Journal of Cultural Studies 3, 5-25.
Davies, M. M., 1989. Television is Good for Your Kids. Hilary Shipman Limited, London.
Desai, S., Adams, C. D., Dubey, A., 2008a. Segmented schooling: inequalities in primary education. IHDS Paper
6, India Human Development Survey.
Desai, S., Dubey, A., Joshi, B., Sen, M., Shariff, A., Vanneman, R. D., 2008b India Human Development Survey
(IHDS), ICPSR22626-v1, University of Maryland and National Council of Applied Economic Research, New
Delhi. Inter-university Consortium for Political and Social Research, Ann Arbor, MI.
Desai, S., Kulkarni, V., 2008. Changing educational inequalities in India in the context of affirmative action.
Demography 45, 245-270.
http://dx.doi.org/10.1353/dem.0.0001
PMid:18613480 PMCid:2474466
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
22
Desai, S., Dubey, A., Joshi, B., Sen, M., Shariff, A., Vanneman, R. D., 2010. Human development in India:
challenges for a society in transition. Oxford University Press, New Delhi.
Dornbusch, S. M., Ritter, P. L., Leiderman, P. H., Roberts, D. F., Fraleigh, M. J., 1987. The relation of parenting
style to adolescent school performance. Child Development 58, 1244–1257.
http://dx.doi.org/10.2307/1130618
PMid:3665643
Earles, K. A., Alexander, R., Johnson, M., Liverpool, J., McGhee, M., 2002. Media influences on children and
adolescents: violence and sex. Journal of National Medical Association 94, 797–801.
PMid:12392043 PMCid:2594155
Fetler, M., 1984. Television viewing and school achievement. Journal of Communication 34, 104–118.
http://dx.doi.org/10.1111/j.1460-2466.1984.tb02163.x
Fisch, S., 2005. Children's learning from television. TELEVISION 18, 10-14.
Friedrich, L. K., Stein, A. H. 1973., Aggressive and pro-social television programs and the natural behavior of
preschool children. Monographs of the Society for Research in Child Development 38, 1-64.
http://dx.doi.org/10.2307/1165725
PMid:4756996
Gadberry, S., 1980. Effects of restricting first graders' TV-viewing on leisure time use, IQ change, and
cognitive style. Journal of Applied Developmental Psychology 1, 45-57.
http://dx.doi.org/10.1016/0193-3973(80)90061-1
Geist, E. A., Gibson, M., 2000. The effect of network and public television programs on four and five year olds
ability to attend to educational tasks. Journal of Instructional Psychology 27, 250-261.
Gentzkow, M., Shapiro. J., 2006. Does Television Rot Your Brain? New Evidence from the Coleman Study.
NBER Working Paper 12021, National Bureau of Economic Research, Washington DC.
Government of India, 2006. Social economic and educational status of Muslim community in India.
Government of India, New Delhi.
Greenstein, J., 1954. Effect of television upon elementary school grades. The Journal of Educational Research
48, 161-176.
Gunter, B., McAleer Jill, J., 1997. Children and Television. Second edition. Routledge, London.
Gupta, R. K., Saini, D. P., Acharya, U., Miglani, N., 1994. Impact of television on children. Indian Journal of
Pediatrics 61, 153-159.
http://dx.doi.org/10.1007/BF02843606
PMid:7927612
Hall, E. R., Esty, E. T., Fisch, S. M., 1990. Television and children's problem-solving behavior: a synopsis of an
evaluation of the effects of Square One TV. Journal of Math Behavior 9, 161-174.
Hancox, R. J., Milne, B. J., Poulton, R., 2005. Association of television viewing during childhood with poor
educational achievement. Archives of Pediatrics and Adolescent Medicine 159, 614-618.
http://dx.doi.org/10.1001/archpedi.159.7.614
PMid:15996992
Hodge, R., Tripp, D., 1987. Children and Television. Polity Press, Cambridge.
Huston, A., Bickham, D., Lee, J., Wright, J., 2007. From Attention to Comprehension: How children watch and
learn from television, In: N. Pecora, J. Murray, E. Wartella (Eds). Children and Television: Fifty Years of
Research. Lawrence Erlbaum, Mahwah NJ.
Huston, A. C., 1992. Big World, Small Screen: the Role of Television in American Society. University of
Nebraska Press, Lincoln.
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. Sage Publications, Thousand Oaks.
Huston, A. C., Wright, J. C., Marquis, J., Green, S. B., 1999. How young children spend their time: television and
other activities. Developmental Psychology 35, 912-925.
http://dx.doi.org/10.1037/0012-1649.35.4.912
PMid:10442861
Huston, A. C., Wright, J. C., Rice, M. L., Kerkman, D., St Peters, M., 1990. Development of television viewing
patterns in early childhood: a longitudinal investigation. Developmental Psychology 26, 409-420.
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
23
http://dx.doi.org/10.1037/0012-1649.26.3.409
Joensson, A., 1985. TV—a threat or a complement to school? A longitudinal study of the relationship between
TV and school. Educational and Psychological Interactions 82, 2-14.
Johnson, R. D., Cooper, H., Chance, J., 1983. The relation of children's television viewing to school achievement
and I.Q. The Journal of Educational Research 76, 294-297.
Jordan, A., 2004. The role of media in children's development: an ecological perspective. Journal of
Developmental and Behavioral Pediatrics 25, 196-206.
http://dx.doi.org/10.1097/00004703-200406000-00009
PMid:15194905
Kondo, K., Steemers, J., 2010. Can television be good for children? URL:
http://www.savekidstv.org.uk/wp.../d/cantelevisionbegoodforchildren.doc.
LaBlonde, J. A., 1966. A study of the relationship between the television viewing habits and scholastic
achievement of fifth grade children. Dissertation Abstracts International 27, 22-84.
Lemish, D., 2007. Children and Television: A Global Perspective. Blackwell, Oxford.
PMid:17419335
Lemish, D., Rice, M., 1986. Television as a talking picture book. A prop for language acquisition. Journal of
Child Language 13, 251-74.
http://dx.doi.org/10.1017/S0305000900008047
PMid:3745331
Liebert, R. M., 1986. Effects of television on children and adolescents. Journal of Developmental and
Behavioral Pediatrics 7, 43–48.
http://dx.doi.org/10.1097/00004703-198602000-00008
PMid:3949990
Linebarger, D., Walker, D., 2005. Infants and Toddlers Television Viewing and Language Outcomes. American
Behavioural Scientist 48, 624-645.
http://dx.doi.org/10.1177/0002764204271505
Lyle, J., Hoffman, H. R., 1972. Children's use of television and other media. In: E. A. Rubinstein, G. A. Comstock,
J. P. Murray (Eds). Television and social behavior (Vol. 4); Television in day to day life: Patterns of use.
Government Printing Office, Washington DC.
MacBeth, T. M. (Ed)., 1996. Tuning in to Young Viewers: Social Science Perspectives on Television. Sage
Publications, Thousand Oaks.
Naigles, L. R., Mayeux, L., 2001. Television as Incidental Language Teacher, In: J. Singer, D. Singer (Eds).
Handbook of Children and the Media. Sage, London.
Piaget, J., 1969. The Origins of Intelligence in the child. International Press, New York.
Rice, M. L., Huston, A. C., Truglio, R., Wright, J. C., 1990. Words from Sesame Street: learning vocabulary while
viewing. Developmental Psychology 26, 421-428.
http://dx.doi.org/10.1037/0012-1649.26.3.421
Robinson, T. N., Chen, H. L., Killen, J. D., 1998. Television and music video exposure and risk of adolescent
alcohol use. Pediatrics 102, e54.
http://dx.doi.org/10.1542/peds.102.5.e54
PMid:9794984
Sargent, J. D., Beach, M. L., Dalton, M. A, Mott, L. A., Tickle, J. J., Ahrens, M. B., Heatherton, T. F., 2001. Effect of
seeing tobacco use in films on trying smoking among adolescents: cross sectional study. British Medical
Journal 323, 1394–1397.
http://dx.doi.org/10.1136/bmj.323.7326.1394
PMid:11744562 PMCid:60983
Sargent, J. D., Beach, M. L., Dalton, M. A., Ernstoff, L. T., Gibson, J. J., Tickle, J. J., Heatherton, T. F., 2004. Effect of
parental R-rated movie restriction on adolescent smoking initiation: a prospective study. Pediatrics 114,
149–156.
http://dx.doi.org/10.1542/peds.114.1.149
PMid:15231921
Ashish Singh and Sarthak Gaurav / Journal of Communication Technology and Human Behaviors (2013) 1: 1-24
24
Schramm, W., Lyle, J., Parker, E. B., 1961. Television in the lives of our children. Stanford University Press,
Stanford.
Sharif, I., Sargent, J. D., 2006. Association between television, movie, and video game exposure and school
performance. Paediatrics 118, e1061-e1070.
http://dx.doi.org/10.1542/peds.2005-2854
PMid:17015499
Signorielli, N., 1991. A Sourcebook on Children and Television. Greenwood Press, London.
Singer, D. G., 1990. The House of Make-Believe: Children's Play and the Developing Imagination. Harvard
University Press, Cambridge.
Singer, J., Singer, D., 1981. Television, imagination and aggression: A study of preschoolers. Lawrence
Erlbaum, Hillsdale NJ.
Singer, J., Singer, D., 1998. Barney and Friends as entertainment and education: evaluating the quality and
effectiveness of television series for preschool children, In: Joy A. K., Gordon L Berry (Eds). Research,
Paradigms, Television and Social Behaviour. Sage, London.
Steinberg, L., Lamborn, S. D., Dornbusch, S. M., Darling, N., 1992. Impact of parenting practices on adolescent
achievement: authoritative parenting, school involvement, and encouragement to succeed. Child
Development 63, 1266–1281.
http://dx.doi.org/10.2307/1131532
PMid:1446552
Strouse, J. D., Buerkel-Rothfuss, N. L., 1987. Media exposure and the sexual attitudes and behaviors of college
students. Journal of Sex Education and Therapy 13, 43–51.
PMid:12143703
The Probe Team, 1999. Public report on basic education in India. Oxford University Press, New Delhi.
Truglio, R. T., Murphy, K. C., Oppenheimer, S., Huston, A. C., Wright, J. C., 1996. Predictors of children's
entertainment television viewing: why are they tuning in? Journal of Applied Developmental Psychology
17, 475-493.
http://dx.doi.org/10.1016/S0193-3973(96)90012-X
Villani, S., 2001. Impact of media on children and adolescents: a 10-year review of the research. Journal of
American Academy of Child Adolescent Psychiatry 40, 392–401.
http://dx.doi.org/10.1097/00004583-200104000-00007
PMid:11314564
Voort, T. H., and Valkenburg, P. M. 1994. Television's impact on fantasy play: a review of research.
Developmental Review 14, 27-51
http://dx.doi.org/10.1006/drev.1994.1002
Wright, J. C., Huston, A. C., Murphy, K. C., St Peters, M., Pinon, M., Scantlin, R., Kotler, J., 2001. The relations of
early television viewing to school readiness and vocabulary of children from low-income families: the
early window project. Child Development 72, 1347-1366.
http://dx.doi.org/10.1111/1467-8624.t01-1-00352
PMid:11700636
Wright, J. C., Huston, A. C., Vandewater, E. A., Bickham, D. S., Scantlin, R. M., Kotler, J. A., Caplovitz, A. G., Lee, J.
H., Hofferth, S., Finkelstein, J., 2001. American children's use of electronic media in 1997: a national
survey. Journal of Applied Developmental Psychology 22, 31-47.
http://dx.doi.org/10.1016/S0193-3973(00)00064-2
Zill, N., 2001. Does Sesame Street enhance school readiness? evidence from a national survey of children, In:
S.M. Fisch, and R. Fisch (Eds). "G" Is for Growing: 30 Years of Research on Children and Sesame Street.
Erlbaum, Mahwah NJ.
Zimmerman, F. J., Christakis, D. A., 2005. Children's television viewing and cognitive outcomes: a longitudinal
analysis of national data. Archives of Pediatrics and Adolescent Medicine 159, 619-625.
http://dx.doi.org/10.1001/archpedi.159.7.619
PMid:15996993