Chapter 3: Methodology
3.2. Measurements
The major variables in the current study included Internet self efficacy, parental Internet mediation, risk perception, and media usage. Factor analysis was first conducted for each scale of the aforementioned variables. It was confirmed that all the items
generated only one dimension prior to reliability tests.
Internet self efficacy
The measurement for Internet self efficacy is adapted and carefully chosen from the study by Eastin & Larose (2000) and Hsu & Chiu (2003). The measurement uses a 5-point Likert scale, ranging from strongly disagree to strongly agree. It generally contains statements regarding their confidence on completing Internet related tasks as well as their familiarity with the Internet (e.g., I feel confident installing an application or software, I feel confident understanding terms or words related to Internet). The reliability test indicates Cronbach’s α = .880 (N=8)
Parental Internet mediation
The measurement for Internet mediation is adapted from the measurement utilized by Livingstone and Helsper (2008; 2010). The measurement uses a 5-point likert scale, ranging from strongly agree to strongly disagree. The statements generally contains efforts
of the parents to mediate children’s Internet activities, that include statements indicating restrictive mediation (e.g., I set up filters / monitoring software on websites), statements indicating active mediation (e.g., I talk to child about Internet use), and statements
indicating co – viewing (I stay nearby when child is online). For items indicating active/co use and interaction restrictions, Cronbach’s α = .749 (N =9). For items indicating
monitoring and technical restrictions, Cronbach’s α =.868 (N=5) Factor analysis was conducted on all thirteen items of parental mediation, and it can be concluded that it yields four components, though for the purposes of this paper, interaction and active – co use strategies will be studied together, and Monitoring and technical restrictions will be studied together.
Table 2
Factor loadings for the parental mediation items
Item Interaction Active/ co-use Monitoring Technical 1. I stay nearby when children is online. .582
2. I set rules for my child about time .542 spent online.
3. I put computer separate to my
child’s bedroom .515
4. I talk to child about Internet use. .495 5. I set up filters/ monitoring software
on websites. .537
6. I check websites that child visited. .755
7. I check child’s SNS. .774
8. I check child’s email messages. .778
9. I check who my child has been .741
chatting with online.
10. I forbid child to use SNS. .427 11. I forbid child to download
things on the Internet. .565 12. I forbid child to give out personal
info online. .577
13. I forbid child to buy things online. .698
Risk perception
The measurement for risk perception is a self – designed measurement that refers to the classifications of Internet risk by Staksrud and Livingstone (2009). Specifically, it asks parents to assess the likelihood and severity of each type of Internet risk (contact, content, conduct) in relation to their children. The measurement uses 5-point likert scales, ranging from strongly agree to strongly disagree. Cronbach’s α = .925 (N=6).
Media usage
The measurement for media usage is a self – designed measurement that intends to investigate their acquisition of knowledge regarding Internet risk. As such, the
measurement asks participants how many days do they watch / see the news from each of the commonly used media (television, newspaper, Internet).
Media exposure to Internet issue
The measurement then asks how often do they see the news regarding Internet risk on the medias mentioned above using a 5-point likert scale, ranging from common to never).
In the last part, the measurement asks the approximate age of the participants, the age of the children who are in the elementary schools, and the household monthly income for reference. The initially English - written questionnaire had been translated into Chinese language by Chinese instructor in National Chengchi University. Both of the questionnaires are attached in this proposal for reference. A pre – test on the measurement was conducted before the execution of the study with a total of ten parents, which brought to light some formatting error on the questionnaire.
Chapter 4 Results
4.1. Demographics of sample
Out of the total 207 participants, females made up for the majority with 146 respondents (70.5%), while 59 respondents (28.5%) were males. In terms of the parents’
age group, the age between 41 – 45 years old made up the highest percentage with 41.7%
(N= 85), followed by 36-40 years old age group with 23.5% (N= 48), and Over 45 years old with 17.6% (N= 36). As for the children’s average age, the peak can be found in 8 years
old with 22.4% (N= 46), followed closely by 10 years old with 2.0 % (N= 41). The median of Children’s age was 9. The data on average household income indicated there are no significant differences between top three variables: NTD 60,001 – 80,000 with 23% (N=
47), NTD 80,001 – 110,000 with 20.6% (n=42), and NTD 110,001 – 140,000 with 19.6%
(N=40). Median score fell under NTD 80,001 – 110,000.
Table 3
Gender and age of the respondents
Item Variable N %
Gender Male 59 28.8%
Female 146 71.2%
Age Group Under 25 years old 1 0.5%
25-30 years old 8 3.9%
31-35 years old 26 12.7%
36-40 years old 48 23.5%
41-45 years old 85 41.7%
Over 45 years old 36 17.6%
4.2 General findings
The following is a report of descriptive results of the major variables in the current study.
Internet self efficacy
The survey on Internet self efficacy items indicated that the large majority of participants were able to find information online with 87.5 % answered or agree or strongly agree. When it comes to understanding terms related to Internet and explaining why a task will not run on the Internet, participants were less confident with 54.6 % and 50.7%
responded agree or strongly agree respectively. The highest self efficacy was found on “I feel confident using an e-mail” item, with 96.6% answered agree or strongly agree. The results also showed that the majority of participants are able to navigate social media (82.1% answered agree or strongly agree). Other items being asked included “I feel confident chatting with other people online“ (82.1% agree or strongly agree), “I feel confident installing an application or software” (74.4% agree or strongly agree), and “I feel confident filtering or blocking websites with inappropriate contents (59.4%).”
Parental mediation
In regards to interaction restrictions, 59.4% of parents would stay nearby when the child is online, with 25% were neutral, and 15.5% disagreed. 84.5% of parents would not put computer in child’s bedroom, while 87.4% would set rules about time child spent online. 76.2% of parents also found it helpful to talk to their child about Internet use. When it comes to technical restrictions and monitoring strategies, 58 % of parents would set up filters on the website, while more than half the parents would monitor child’s visited websites (58.3%), check child’s SNS (52.2%), and check child’s online chat partners (55.1%), however only 43% agree to check child’s email messages. Another thing worth of note is that about 44% of parents disagree on forbidding child to use SNS, with 37.2% were neutral about it and only 18.3% agreed to do so. Other results include most parents
agreeing to forbid child’s online shopping (70%), and agreeing to forbid child to give out personal info online (67.2%).
Risk perception
The majority of the parents agreed or strongly agreed that their child may be exposed to inappropriate contents (61.9%), while 60.1% of parents agree that these
contents are likely to severely affect their child. Slightly less parents agreed that their child is likely to experience risky online interaction (54.6%), but slightly more parents agreed that their child would be severely affected by risky online interaction (61.9%). Last but not least, when asked whether their child is likely to initiate or produce risky contents or interactions online, more parents disagreeing (40.1%) than agreeing (33.3%) with 26.6%
were neutral. In regards to the severity of the conduct risk, about 40.1% agree that it will severely affect their child.
Media usage
In terms of which media outlet do parents hear more about Internet risk news, television (54.2% answered often or common) and Internet (56.3%) are the two major sources. Newspapers were rated lower as the source of Internet risk news, with only 26.1%
of parents answered common or often. In terms of the media that parents normally use to read or watch news, Internet had a slight edge with 74% of parents answered 5 to 7 days, followed by television (66.8% answered 5 to 7 days), and newspapers (55.4% answered 5 to 7 days).
4.3. Tests of hypotheses
In this section, the previously stated hypothesis will be tested in relation with the data acquired. The Statistical Package for the Social Sciences (SPSS) program was utilized to validate, compute, and describe the relations between variables that shape the
hypothesis.
H1: Parents’ exposure to Internet risk news would increase their risk perception of Internet use for their children.
In order to test the first hypothesis, Pearson’s correlation was used to test the relationship between Parents’ exposure to Internet risk news and parents’ risk perception towards Internet. Partial correlation was used for the analysis, taking into account the participants’ age and gender. The result showed a positive correlation (r = .20) with a significance of p<.01 level therefore supporting the hypothesis one.
RQ1: How do different media (i.e. television, newspapers, and Internet) influence parents’ perceptions and attitude towards Internet risk?
A more comprehensive look on Pearson’s correlations among parents’ exposure to risk news and parents’ risk perception showed there was a moderate positive correlation between exposure to Internet risk news via Internet and parents’ risk perceptions towards Internet (r = .303) with a significance of p<.001 level. No significant results were found between exposure to Internet risk news via television, via newspapers, and parents’ risk perception. The number of days parents see news online also correlated (r=.320, p<.001) with how often they see the Internet risk news online. No such results were found with television and with newspapers.
H2: Parents who perceive Internet as a high risk to their children’s safety would be more likely to mediate children’s Internet use.
For testing the second hypothesis (and third), regression analysis was used to examine different variables that might explain or predict parental mediation behavior. As shown in Table 3, the hierarchical regression model that included self efficacy, risk perception, media usage and exposure overall explained for about 43% of variance in predicting parental mediation behavior (R square=.427). In particular, when risk perception was entered at the
second level after demographic variables were controlled, the variance explained increased drastically (β= .429, p< .001, R square change = .236). Such finding indicated risk
perception as an important positive predictor for parental mediation behavior. The variable resulted in it can be concluded that parents’ risk perception towards Internet has a moderate to strong positive correlation with Parental Mediation, therefore supporting H2.
Table 4
Hierarchical multiple regression analysis of parental mediation
Level Variables β t p-value
1 Sex .068 .945 .346
Parent’s age -.140 .-1.637 .103
Children’s age
.003 .038 .970
Household Income .192 2.549 .012*
R square =.043
2 Risk Perception .498 7.788 .000***
R square change =.235
3 Self Efficacy .322 4.227 .000***
R square change = .095
4 Days Watching TV News -.098 -1.570 .118
Days Reading Newspapers -.001 -.019 .985
Days Seeing Internet News -.007 -.096 .924 R square change = .009
5 Issue Exposure on TV 227 2.997 .003**
Issue Exposure on Newspapers 060 .872 .385 Issue Exposure on Internet -.080 -.960 .338 R square change = .045
R square total = .427 Note: ***=p <.001, **=p<.01, *=p<.05
RQ2: How do different types of risk perceptions (i.e. content, contact, or conduct) correlated with different use of parental mediation strategies?
To answer RQ2, Pearson correlations were conducted between different types of risk perception (Content, Contact, Conduct) and parental mediation. Partial correlation method was used to control participants’ sex and age variables. Results showed that all three types of risk perception have a moderate to strong positive correlations with parental mediation, with Content reported as the highest (r = .478), followed by Contact (r=.410) and Conduct (r
=.393). The results were all significant at p <.001.
H3: Parents’ Internet self efficacy is positively correlated with use of technical restrictions and monitoring strategies.
As the regression table suggested, self efficacy had some effect on parental mediation (R square change =.094), ( R square =.115) and further regression analysis was conducted on interaction restrictions, active co-use strategies of parental mediation and the one with technical restrictions, monitoring strategies of parental mediation. The results indicated that internet self efficacy had a slightly bigger impact on technical restrictions and monitoring strategies (R square =.115) than on interaction restrictions and active co-use strategies (R square =.053). Partial Pearson’s correlations were also conducted to test the H3 while
controlling the participants’ sex and age variables. The results showed that parents’ Internet self efficacy correlated positively with technical restrictions and monitoring strategies of parental mediation (r = .348, p<.001), therefore both the correlations and the regression analysis supported H3. In regards to two other types of parental mediation strategies, results also showed a positive correlation between interaction restrictions, active & co-use strategies and parental mediation (r=.248, p<.001) albeit slightly weaker than the other two strategies.
Chapter 5 Discussion
5.1. Major findings and implications
As expected, the participants from Taipei sample showed less proof of digital divide when it comes to household income and Internet access, based on their general competence in operating Internet technology. A large majority of sample knew how to find information online, chat with others online, and navigate social media. Results from the media usage section also revealed that people read more Internet news and read fewer newspapers, with television news ranked in between. According to the survey, participants read news related to Internet risk the least amount in the newspapers. While television and Internet fare pretty similar with each other in this regards, television was found to be one medium that significantly affects parental mediation, while Internet was found to significantly affect parents’ risk perception. This confirmed that parents’ fear and behavior on Internet mediation had been influenced or aggravated by media portrayals of Internet risk. Clark’s (2009)
argument that such apprehension is to be documented more so among parents from lower income families was also not found in this study, as there was no negative correlation between parents’ household income and either parents’ risk perception or their mediation behavior.
While Clark (2009) previously stated that parents who have less experience with communication technology may express more concerns on the matter, the current study revealed that parents’ risk perception actually escalated among parents who are high on Internet self efficacy. Savolainen (2002) also suggested that parents’ Internet literacy would make parents supervise less, which was also not supported by the current study. The possible explanations to all these is that the familiarity to communications technology may make parents become more aware of the issues related to the Internet and therefore encouraging parents to take a more active role in managing it. Such awareness might also explain parents’
relative unwillingness to forbid children to use social media, as they might have perceived it as a beneficial tool for children’s development and social interaction.
Despite the fact that most literature on parental mediation focused on television as medium, the risks are greater on the Internet, and the parents are well aware of it. This is evinced by parents’ relatively equal apprehension between content risk and contact risk, which is a variable that did not previously exist in media studies with television. Parents’
anxiety in (especially) content risk and contact risk would then translate to how parents mediate children’s Internet, as risk perception was indeed found to be the biggest predictor of parental mediation. The study showed parents employed different kind of strategies to get a control of the situation. While the majority of parents felt communication was key and decided to talk about Internet usage to their children, some found it helpful to monitor some
of the child’s activities on the Internet. The results from this section were for the most part comparable to those of the study conducted by Livingston and Helsper in 2008.
With parent’s Internet self efficacy was found to affect parental mediation, how and to what extent does the process work? Using Livingston and Helsper (2008)‘s categorization of four different strategies of parental mediation, it was originally hypothesized that parents with high Internet self efficacy would be more likely to use mediation strategies that indeed required more Internet skills, namely technical restrictions and monitoring strategies. While the positive correlation was found between them, the positive correlation was also found between parent’s Internet self efficacy and two other strategies that did not require high Internet skills, namely active and co-use and interaction restrictions. The results suggested that what Internet self efficacy might have brought to the parents were not a mere certain technical advantages in parenting, but instead, an overall willingness and motivation to take an active role in mediating children’s Internet.
5.2. Limitations and future research
One major limitation to the study would be the use of convenient sampling, as parents who were available in the afternoon to pick up their child may have shared certain traits or profiles. The question of whether the parent who picked up the child at school would be the same one who set up the rules at home also could only be answered with a mere guess at the moment. That is, the survey was done by one half of the parents and therefore may not have told the whole story. It would also be insightful to do an in depth interview with some of the parents to find the motivational factor behind their behavior, as it would explain why certain parents choose to adopt certain methods and to find out
whether some parents actually learn the Internet skills to better for a purpose of mediating children’s Internet behavior better.
In regards to questionnaire items, certain questions in this study were more difficult to answer among parents who had younger child as some of the situations may not have applied to the child as yet and therefore could only be answered hypothetically, including questions such as banning the use of e-mail, social media, and online shopping. With this in mind, future research may benefit from narrowing and focusing on particular age group (e.g. 6-9 years old or 10-12 years old) to prevent such inconsistencies. Some questions also did not take into account Internet access through mobile phone / Smartphone. Survey item such as “I would not put computer in my children’s bedroom” may not be sufficient to explain parents’ mediation behavior and children’s Internet usage in current situation.
While Smartphone may not be owned by all of the children, especially considering some of their ages, its inclusion into the variables will still be helpful in explaining the whole picture. As Smartphone starts booming in a highly digital city nowadays such as that of Taipei, children’s uses of Internet are consequently more sporadic and even more difficult to track down. As such, Internet’s risks for children are more likely to increase, and the dynamics of parents’ Internet mediation may have been slightly altered, which in turn, demand future research in this area.
References
Ajzen, I. (1991). The theory of planned behavior. Organizational behavior and human decision processes, 50(2), 179-211.
Bandura, A. (1977). Self efficacy: toward a unifying theory of behavioral change.
Psychological review, 84(2), 191.
Bawden, D. (2008). Origins and concepts of digital literacy. Digital literacies: Concepts, policies and practices, 17-32.
Byron, T. (2008). Safer children in a digital world: the report of the Byron Review: be safe, be aware, have fun. London: Department for Children, Schools and Families, and the Department for Culture, Media and Sport.
Cassell, J., & Cramer, M. (2007). High tech or high risk: Moral panics about girls online.
Digital Youth, Innovation, and the Unexpected, 53-75.
Cassidy, W., Brown, K., & Jackson, M. (2012). " Making Kind Cool": Parents' Suggestions for Preventing Cyber Bullying and Fostering Cyber Kindness.
Journal of Educational Computing Research, 46(4), 415-436.
Clark, L. S. (2009). Digital media and the generation gap: Qualitative research on US teens and their parents. Information, Communication & Society, 12(3), 388-407.
Clark, L. S. (2011). Parental mediation theory for the digital age. Communication Theory, 1(4), 323-343.
DeHue, F., Bolman, C., & Völlink, T. (2008). Cyberbullying: Youngsters' experiences and parental perception. CyberPsychology & Behavior, 11(2), 217-223.
Eastin, M. S., & LaRose, R. (2000). Internet self‐efficacy and the psychology of the
digital divide. Journal of Computer‐Mediated Communication, 6(1).
Elena, A., Laouris, Y., & Taraszow, T. (2010). Identifying and Ranking Internet Dangers.
2010 Social Applications for Life Long Learning, 68.
Henke, L. L. (1999). Children, advertising, and the Internet: An exploratory study.
Advertising and the World Wide Web, 73-80.
Kerr, M., & Stattin, H. (2000). What parents know, how they know it, and several forms
Kerr, M., & Stattin, H. (2000). What parents know, how they know it, and several forms