«Jl. of Interactive Learning Research (2005) 16(3), 259-272 How Low-Income Children Use the Internet at Home LINDA A. JACKSON, ALEXANDER VON EYE, ...»
Jl. of Interactive Learning Research (2005) 16(3), 259-272
How Low-Income Children Use the Internet at Home
LINDA A. JACKSON, ALEXANDER VON EYE, FRANK BIOCCA,
GRETCHEN BARBATSIS, YONG ZHAO, AND HIRAM E. FITZGERALD
Michigan State University, Lansing, MI USA
HomeNetToo is a longitudinal field study designed to examine home Internet use by low-income families in the United States. Participants were 140 children, mostly African American, whose Internet use was continuously and automatically recorded for one year. This article focuses on relationships between children’s main computer activities, academic interests, career aspirations, social engagement, and their Internet use. Results indicated that children used their computers mainly to play games and search the Web. Using the computer to listen to music or to e-mail was related to greater Internet use whereas using it for schoolwork was related to less Internet use. Children whose academic interest was social science used the Internet more than did children whose interest was mathematics or science. Children who aspired to careers in the professions or computing used the Internet more than did children who aspired to careers in sports, entertainment, or human services. Internet use was unrelated to social engagement. Academic performance could not explain relationships between main computer activity, academic interest, career aspirations, and Internet use. Implications for research on children's beliefs about the Internet and their influence on Internet and technology use are discussed.
HomeNetToo is a longitudinal field study designed to examine the antecedents and consequences of home Internet use in low-income families (NSF-ITR #085348, http://www.msu.edu/user/jackso67/HomeNetToo/).
Participants were 140 children, most of whom were African American 260 Jackson, von Eye, Biocca, Barbatsis, Zhao, and Fitzgerald (83%), male (58%), and living in single-parent families (75%) in which the median household income was less than $15,000 (USD) per year. Internet use was automatically and continuously recorded. Children completed surveys at multiple points during the 16-month project. Among the survey items were questions about how they used their computers, their academic interests, career aspirations, and social engagement. This article focuses on relationships between children's answers to these questions and their Internet use.
Survey research on children and computing indicates that childrenuse home computers for a variety of purposes, including leisure activities (e.g., game playing, web surfing) and schoolwork (Pew Internet and American Life Project, 2000a, 2000b, 2002; Shields & Berman, 2000). However, only a handful of studies have examined how children use the Internet at home.
Some findings suggest that children use the Internet primarily for schoolwork (i.e., searching the Web for information for school projects), followed by communication with peers (i.e., e-mail, instant messaging, and chat;
Kraut, Scherlis, Mukhopadhyay, Manning, & Kiesler, 1996; Pew Internet and American Life Project, 2002; Turow & Nir, 2000; UCLA Internet Reports, 2000, 2001, 2003). More recent findings suggest that the main computer activities for children are playing games and searching the Web (Jackson, Biocca, et al., 2004). Moreover, the extent of children’s Internet use for communication is uncertain (Becker, 2000; Cho, Gil de Zuniga, Rojas & Shah, 2003; Facer & Furlong, 2001; Gorski, 2002). The only study to automatically record Internet use was conducted in 1995-1996. Findings based on the combined sample of adults and children indicated that although participants reported that e-mail was a very important reason for going online, they actually sent less than one e-mail a week (Kraut et al., 1996).
Consistent with these findings, Jackson, Biocca, et al. (2004) found very little e-mail use by HomeNetToo children when automatically recorded measures of Internet use were considered.
Absent from the research are studies of relationships between children’s academic interests, career aspirations, and their Internet use. There is reason to expect such relationships (Jackson, vonEye, Biocca, et. al., 2004; Rocheleau, 1995; Wenglinsky, 1998; Woodard & Gridina, 2000; Yelland & Lloyd, 2001). Children interested in science, mathematics, and computing, and children oriented toward careers in these and related professions, are likely to use the Internet more than are children interested in other subjects and careers. Jackson, vonEye, Barbatsis, et al. (2004) reported preliminary findings contrary to this view. Children whose favorite school subject was social science used the Internet more than did children with other favorite school subjects (e.g., mathematics). Unfortunately these investigators did not examine the relationship between career aspirations and Internet use. Moreover, predicted relationships rest on the assumption that children recognize How Low-Income Children Use the Internet at Home 261 the relevance of Internet use to their academic interests and career aspirations.
Studies of adults suggest a relationship between social engagement and Internet use (Katz & Rice, 2002; Kraut et al., 1998, 2002). However, some studies find a positive relationship – more social engagement is related to greater Internet use (UCLA Internet Reports, 2000, 2001, 2003; Katz & Rice), some find a negative relationship – more social engagement is related to less Internet use (Kraut et al., 1996; Nie & Erbring, 2000) and some find no relationship between the extent of Internet use and social engagement (Jackson, von Eye, Barbatsis, et al., 2004; Kraut et al., 1998; Kraut et al., 2002). Whether there is a relationship between social engagement and Internet use for children has yet to be systematically investigated.
The paucity of previous research suggests an exploratory approach to examining relationships between how children use their computers, their academic interests, career aspirations, social engagement, and Internet use.
The following exploratory hypotheses were formulated:
Hypothesis 1: Children who use their home computers mainly for schoolwork will use the Internet more than will children who use them primarily for leisure activities (e.g., listening to music, communicating with peers).
Hypothesis 2: Children whose academic interests are mathematics or science will use the Internet more than will children whose academic interests are in other disciplines.
Hypothesis 3: Children aspiring to careers in the professions (e.g., doctor) or computing will use the Internet more than will children aspiring to careers in other fields.
Hypothesis 4: Children who are less socially engaged will use the Internet more than will children who are more socially engaged.
Also considered in this research were gender differences in main computer activities, academic interests, career aspirations, social engagement, and Internet use (Jackson, 2004; Jackson, Ervin, Gardner & Schmitt, 2001) and whether academic performance could explain relationships between these factors and Internet use (Jackson et al., in press).
METHODS Participants and Procedures Participants were 140 children residing with 120 adult participants in the HomeNetToo project. Adult participants were recruited at meetings held at the children’s middle school and at the Black Child and Family Institute, Lansing, MI. Requirements for participation were that the child be eligible 262 Jackson, von Eye, Biocca, Barbatsis, Zhao, and Fitzgerald for the federally-subsidized school lunch program, that the family had had a working telephone line for the previous six months, and that the family had never had home Internet access. Participants agreed to have their Internet use automatically and continuously recorded, to complete surveys at multiple points during the project, and to participate in home visits. In exchange they received home computers, Internet access, and in-home technical support during the Internet recording period (i.e., 16 months).
Adult participants were primarily African American (67%) and female (80%). The majority reported having some college education or earning a college degree (62%), making our sample much better educated than the typical low-income sample (e.g., UCLA Internet Report, 2000, 2001, 2003).
Average age of adult participants was 38.6 years old. As noted earlier, child participants were primarily African American (83%), male (58%), and living in single-parent households (75%) in which the median household income was $15,000 (USD) annually (49%). Average age was 13.8 years old.
Measure Internet use. Four measures of Internet use were recorded continuously for the first year of the project: time online (minutes/day), number of sessions (log-ins per day), number of (unique) domains visited (per day) and number of e-mails sent (per day). To examine changes in relationships over time Internet use measures were divided into two time periods: time 1 (1 to 6 months) and time 2 (7 to 12 months).
Main computer activity. Children responded to the question: “What is the main thing you do when you use your home computer? Circle only one number. 1=Play games, 2=E-mail, 3=Search the Internet (Surf the Web), 4=Make documents (word processing, spreadsheets), 5=School work, 6=Listen to music, 7=Other.” Academic interests. Children’s open-ended responses were obtained to the question “What is your favorite school subject?” Career aspirations. Children’s open-ended responses were obtained to the question “What kind of a job would you like when you're grown up?” Social engagement. Children responded to three questions about their afterschool activities: “Are you on a sports team?” (Yes/No). “Do you belong to any groups or clubs?” (Yes/No). “Do you have any hobbies or interests?” (Yes/No).
Academic performance. Children's grade point averages (GPAs) and scores on standardized tests of academic performance (Michigan Educational Achievement Program (MEAP) tests) were obtained from their schools (with parental permission). GPAs were obtained for Fall, 2000 (i.e., pretrial), Spring, 2001 (i.e., six months into the project), Fall, 2001 (i.e., 12 How Low-Income Children Use the Internet at Home 263 months into the project), Spring, 2002 (i.e., posttrial). MEAP scores (reading and math) were obtained at the end of each academic year of the project (i.e., 2001 and 2002).
RESULTS Frequency of Children’s Internet Use Children spent an average of 27 minutes per day online, participating in
0.6 sessions (i.e., they did not log on daily) and visiting 10 domains. They sent a negligible number of e-mails – less than one per week. Paired contrasts to examine changes in Internet use over time indicated very little change. Time online and number of e-mails sent remained the same throughout the project. Number of sessions decreased slightly but number of domains visited remained the same. Because the distributions of all 4 Internet use measures were highly skewed log transformations were used in all the analyses.
Children’s Main Computer Activity and Their Internet Use The main reasons that children used their computers were to play games (34%) and search for information on the Web (33%) followed, in roughly equal frequencies, by school work (11%), listening to music (11%), and emailing (8%). Making documents (3%) and other uses (0%) were rarely mentioned and were excluded from subsequent analyses.
Means for the four measures of Internet use for each main computer activity group are presented in Table 1. Analyses of variance (ANOVAs) and posthoc contrasts (LSD test) were used to examine differences in Internet use among the five main computer activity groups. According to Hypothesis 1, children who use their home computers mainly for schoolwork will use the Internet more than will children who use them primarily for leisure activities. Contrary to this hypothesis, children who used their computers mainly to listen to music spent more time online, F(4, 120)=4.22, p.05, participated in more sessions, F(4, 120)=3.98, p.05, and visited more domains, F(4, 120)=4.41, p.05, than did children in any of the other main computer activity groups. In fact, children who used their computers mainly for schoolwork used the Internet less than did children in any other group. In addition, children who used their computers mainly for e-mail sent somewhat more e-mails than did other children, F(4, 88)=2.05, p.1.
Relationships between main computer activity and Internet use were different during the second half of the year. Again contrary to Hypothesis 1, children who reported that e-mail was their main computer activity spent more time online, F(4, 117)=3.61, p.05, engaged in more sessions, F(4, 117)=3.09, p.05, visited more domains, F(4, 113)=3.60, p.05, and sent more e-mails (LSD test) than did children in any of the other computer activJackson, von Eye, Biocca, Barbatsis, Zhao, and Fitzgerald
Note: T1 is Time 1: 1 to 6 months. T2 is Time 2: 7 to 12 months.
ity groups. Once again, children who used their computers mainly for schoolwork were least likely to use the Internet.
Chi-square analyses revealed gender differences in main computer activity (Π 2(4)=16.83, p.001). Boys were more than twice as likely as girls to report that playing games was their main computer activity.
There were no differences among the main computer activity groups in GPAs or MEAP scores. Thus, differences in academic performance cannot explain differences in Internet use among the main computer activity groups. Of the 12 measures of academic performance considered in these analyses, gender differences were obtained for one measure only. Girls’ GPAs in Spring 2001 were higher (2.28) than boys’ GPAs at that time (1.86, F(1, 103)=5.78, p.05).
Children’s Academic Interests and Their Internet Use Children's responses to the question of what was their favorite school subject were coded into six categories. Mathematics was by far the most popular school subject (31%), followed in roughly equal frequencies by science (17%), literature (17%), and social science (15%). The categories gym (3%) and other (18%) were excluded from subsequent analyses, the former because it was infrequent and the latter because it was too ambiguous.
Means for the four favorite school subject groups are presented in Table