Internet use and Online Literacy Among Middle Grade Students at Risk of Dropping out of School Free

Risk for dropping out of school has been linked to many different factors, including SES, ethnicity, and school location (NCES, 2000; Smink & Schargel, 2004). African American and Hispanic students are particularly at-risk and often experience higher rates of poverty than White Americans (DeNavas-Walt, Proctor, & Smith, 2006; Smink & Schargel, 2004). School location can also affect the likelihood of students dropping out. Students attending schools in highly urban areas, with increased rates of gang activity, violence, and drug use, as well as those in rural areas, where staffing, extracurricular programs, and money are at a short fall, are at an increased risk of school dropout (Smink & Schargel, 2004). The existence of two or more of these risk factors increases the probability of a student dropping out (Croninger & Lee, 2001; Farmer, et al., 2004). When multiple risk factors are present, students often experience disengagement and a lack of motivation related to school (Sun et al., 2004).

Not only is engagement important for dropout prevention, but research also indicates that student engagement in school is likely to lead to higher achievement. Finn and Rock (1997) report that students being disengaged and emotionally disconnected from school compounds existing obstacles to school success. Unfortunately, those most likely to become disengaged are those who are already among the population with the greatest risk of dropout, namely minority students and students from low-income homes (Finn, 1993; Rumberger & Larson, 1998).

Because at-risk students who are engaged in some aspect of school, whether they are connected to a particular teacher, academics, or an extracurricular activity, are more likely than nonengaged students to achieve academic success (Finn & Rock, 1997), the topic of reading engagement becomes an important one. Reinking (2001) argues that the Internet offers students new opportunities for increased engagement with text and that electronic texts are inherently more engaging than traditional printed texts. Thus, making use of the Internet and its tools may increase students’ academic engagement (Becker, 2000; Means & Olson, 1995) and, perhaps, increase the likelihood that students will stay in school.

Not only is technology potentially useful in the classroom to engage students, but it may also prepare students for the changing literacy practices needed to succeed in higher education and in the workplace (Leu, 2002; The New London Group, 2000). Leu et al. (2004) have argued that unique skills and strategies are required for reading, writing, and communicating online, and that reading on the Internet involves the following: (a) formulating questions, (b) locating information, (c) critically evaluating information, (d) synthesizing information, and (e) communicating information. Because reading and navigating the Internet requires unique literacy skills and higher-level reading comprehension skills (International Reading Association [IRA], 2002; Leu et al., 2004; RAND Reading Study Group, 2002), the use of complex, multimodal texts on the Internet may further marginalize an already at risk group of students, especially if they are not learning the new literacy skills required for reading and writing in online, digital environments

Additionally, recent research indicates that the requirements of No Child Left Behind legislation (U.S. Department of Education, 2002) may negatively impact the integration of twenty-first century literacies within classrooms located in poorer communities (Henry, 2007; Leu et al., 2008). This fact may also place minority students at an even greater risk for unemployment, as they are unable to develop the digital literacies required for the workplace of the twenty-first century. Research also shows increased Internet use to have a positive correlation with higher levels of SES (Sun et al., 2005); therefore, students who are at-risk of dropping out of school may also be those who are least likely to use the Internet.

The term digital divide refers to the inequality of access to the Internet and the skills to successfully utilize it (DiMaggio, Hargittai, Newman, & Robinson, 2001; Norris, 2001). Because many at risk students do not have Internet access outside school, it is important to look at the availability of Internet access inside school. Teachers often report the lack of access as the most inhibiting factor for using technology (Henry, 2005). Perhaps more importantly, we need to look at how technologies are used among economically advantaged and disadvantaged learners. Hargittai (2002) refers to this as a second-level digital divide. Her definition includes differences in how the Internet is used and the skills necessary to use the Internet effectively. Research indicates a second level digital divide related to Internet use and skill level exists (Hargittai, 2002; Henry, 2005, 2007). Students in low SES school districts are shown to use computers more frequently for remedial or vocational purposes than those in high SES districts (Henry, 2007; Warschauer, Stone, & Knobel, 2004). The consequences of this situation may result in a widening gap between economically advantaged and disadvantaged learners in ways that have considerable import for their engagement in school, their likelihood to stay in school, their opportunities for social participation, and economic success after they leave school.

The Internet holds promise for increasing engagement in academic pursuits that may keep at risk students engaged in school. Additionally, a lack of access to the Internet and opportunities to develop the skills necessary to use it effectively may inhibit disadvantaged students’ competitive stance in seeking upward social mobility (DiMaggio et al., 2001; Norris, 2001). If this is the case, educators need to frame instruction to foster and develop the new literacies of the Internet among these at risk students for two main reasons: (a) To keep these students engaged in school and reduce their risk of dropping out; and (b) To provide these students with the requisite skills to be competitive in today’s work force and global economy.

The first step in addressing these issues, the researchers believe, is to know more precisely the extent to which students, particularly those most at-risk of dropping out, during the critical middle-school years are using the Internet and to what extent they have acquired the twenty-first century skills and dispositions needed to be literate in digital environments. Yet, to our knowledge, there are no data in the literature about how often economically disadvantaged or at risk students use the Internet or which strategies they have developed specific to reading and writing on the Internet. The present study addresses this limitation.

As part of a larger study aimed at developing an instructional model and activities for teaching Internet reading comprehension (Leu & Reinking, 2005), the present study sought to investigate the Internet use and online literacy skills of students at risk of dropping out of school. Specifically we addressed the following research questions:

1. Where do adolescents at risk of dropping out of school obtain Internet access and how frequently do they use it? How does Internet access and frequency of use differ based on race, gender, and locale of at risk students?

2. Considering their experiences inside school and outside school, in what types of Internet activities do adolescents at risk of dropping out of school engage in, what technology-related tools do they use, and how frequently do they engage in technology-related activities in general?

3. How skilled are adolescents at risk of dropping out of school in the new literacies of Internet-based reading, writing, and communicating?

The research sites included 12 middle schools located in nine economically challenged school districts in the northeast and southeast regions of the United States. The schools were selected because their student populations exhibited characteristics associated with high dropout rates, such as low SES, large minority populations, high participation rates in free and reduced-price lunch programs (Nowicki, Duke, Sisney, Strickner, & Tyler, 2004; Richman & Bowen, 1997) and because the school districts have a history of high student dropout. Five of the six schools in the northeast region were located in urban districts and one was located in a rural area. All six schools in the southeast region were located in rural areas or small cities. Access to computers varied by site as listed in Table 1. 

A sample population of seventh-grade students (n = 1,025) was selected because research indicates that Internet use surges at the seventh-grade level (Lenhart, Madden, & Hitlin, 2005). The researchers recognize that it is unlikely that every student in the present study is at risk for dropping out of school. However, all students attend a school with risk factors associated with high dropout rates and receive the same instruction and resources as those students who are at risk. Table 2 presents student data from participating schools. The pseudonyms in the table indicate whether the school was located in an urban or rural setting as well as the geographic location as either being in the northeast or southeast region.

All students in each school who returned signed parental consent forms and gave assent completed the survey and assessment measure either on the day it was administered or, to accommodate absences, on a day within a week of the original administration. The students used desktop computers in a computer lab setting. Administration took place during one 45-minute class session during the instructional school day in January and February of 2006. At least one member of the research team was present during administration to ensure consistency across sites. The researchers followed a set protocol in order to reduce bias and increase credibility and integrity of the results (Hammersley & Gomm, 1997).

Paricipants completed a researcher-designed survey and assessment instrument (Carter & Henry, 2006; Henry, Mills, Rogers, & Witte, 2006; Hutchison, 2008). Scaling procedures for survey creation and evaluation were used to guide the instrument development (Netemeyer, Bearden, & Sharma, 2003). To begin, a survey blueprint was created to measure the following three constructs related to new literacies:

• Use of Internet Tools: This construct sought to identify the types of online tools adolescents use to locate information, interact/communicate with others, and to access/engage in multimedia or share documents via the Internet.

• Online Reading Material: This construct sought to identify the types of online reading materials that adolescents access for academic assignments or to learn more about a personal interest or hobby. This factor addressed the types of information accessed as well as the frequency of access.

• Internet-Based Literacy Skills: This construct sought to identify the locating, evaluating, synthesizing, and communicating behaviors that adolescents employ to utilize sources of information they access via the Internet.

An initial item pool was created with 96 items that were developed based on the researchers’ knowledge of the constructs relevant to the new literacies framework used to guide this work. Seven experts in the field of new literacies research were recruited to establish content and face validity through a content validation procedure (Netemeyer et al., 2003). These experts determined how well each item in the pool reflected the constructs to be measured. Fifteen items were flagged as problematic and were either revised or removed from the item pool. An online format that could be administered over the Internet using OnQ, an online survey creation tool (Henry et al., 2006), was created. The instrument was then piloted with a sample of sixth through eighth- grade students (n = 386) to further refine the items. Revisions involving clarity of items and format were made based on the pilot data. The final instrument contained 6 demographic variables, 70 Likert-style items, 5 forced response questions, and 4 open-ended questions for a total of 85 items. The instrument can be viewed at http://camss.clemson.edu/READING.

Several statistical procedures were used to determine the factor structure of the instrument and test for internal reliability, including (a) principal axis factoring (PAF) procedure using an oblimin rotation to determine the structure of the factors that the Likert-style items measured (Pett, Lackey, & Sullivan, 2003; Thompson, 2004), which showed factor loadings between .410 and .791, (b) an item analysis (p values ranged from .31 to .86) to determine if the item difficulty of the forced response questions was adequate (Haladyna, 1999), and (c) a Cohen’s Kappa analysis (K = .87) to determine the degree of reliability among the four raters of the open-ended questions. The validation procedures that were used showed that the instrument developed for this study was a sufficient measure of the three identified factors previously noted.

Research Question 1: Where do adolescents at risk of dropping out of school obtain Internet access and how frequently do they use it? How does Internet access and frequency of use differ based on race, gender and locale of at risk students?

Internet access and frequency of use. Eighty-one percent of the students reported that they have a computer in their home. Of those students, 47% reported having more than one computer. Most students (90%) with computers at home have at least one computer connected to the Internet, and the majority of students (63.6%) also reported that they use the Internet most often at home. Of the students who reported that they do not have a computer at home, a small percentage (4.8%) accesses the Internet at a public library or at a relative’s house (17%). Many more students (45%) reported accessing the Internet at an Internet café, community center, or friends’ homes (83%). Despite having access to the Internet outside their own home, an independent-samples t test, t(1,012) = -7.71, p < .01, revealed that students without a computer at home have a lower frequency rate of Internet use outside school than students with access to a computer at home.

Among urban and rural users, access to the Internet outside school is similar. For example, four percent of urban users and three percent of rural users access the Internet at a public library. About the same number of urban users (39%) and rural users (40%) report using the Internet at an Internet café or community center. Similarly, the majority of urban users (70%) and rural users (73%) access the Internet at a relative’s home or at a friend’s home with rates of 82% and 88% respectively. There were no significant differences in Internet access found when comparing students in urban versus rural settings.

When asked about the amount of time they spend on the Internet, 55.4% of the students reported that their time online has increased from the previous year. Despite this increase, only 45% of the students reported using the Internet a few times each week or more inside school and even fewer students (43%) reported using the Internet with the same frequency outside school.

Comparisons by race, gender, and locale. Students were assigned an “inside school use” score and an “outside school use” score based on how frequently they reported participating in various Internet activities (see item descriptions in Table 3 for a complete list of activities). One to five points were allocated for each Internet activity a student reported based on how frequently the student participated in each activity. These points were used to calculate composite scores for inside and outside school use of the Internet, which were used to compare differences in frequency of use among racial, gender, and geographic groups.

A one-way analysis of variance (ANOVA) was conducted to compare the frequency of Internet use inside and outside school among racial groups (African American, Asian, Hispanic, Caucasian, and Other). The main effect for race/ethnicity for inside school frequency of use was statistically significant F(4, 1,013) = 5.10, p = <.001, η² = .02. A post hoc analysis using the Bonferroni procedure indicated that African Americans used the Internet at statistically higher levels than Caucasians inside school. No other significant effect was found. The main effect for outside school frequency of use was also statistically significant F(4, 1,013) = 4.36, p = .002, η²=.02. A post hoc analysis using the Bonferroni procedure indicated that African Americans, Asians, and Caucasians used the Internet more frequently outside school than Hispanics. No other significant effect was found.

Independent samples t tests were used to test for differences between inside school and outside school frequency of use among males and females and among rural and urban users. The tests revealed that females have a significantly higher frequency of use outside school than do males, t(1,012) = 3.77, p < .01. There was no significance for inside school frequency of use among males and females. Students in rural settings showed a significantly higher frequency of use inside school than do students in urban settings, t(1,022) = 8.43, p < .01. There was no significant difference in frequency of use outside school based on students’ locale.

Research Question 2: Considering their experiences inside school and outside school, what types of Internet activities do adolescents at risk of dropping out of school engage in, what technology-related tools do they use, and how frequently do they engage in technology-related activities in general?

Internet activities inside and outside school. Using sets of parallel questions, students reported how often they engage in the same activities inside and outside school. For example, students were asked: (a) How often do you use the Internet inside school?; and (b) How often do you use the Internet outside school? Table 3 lists the percentage of students engaging in each of the different Internet activities and reading environments on a regular basis. For comparison purposes, activities in which students reported they engaged in at least a few times each week or more frequently were identified as being used “on a regular basis.”

As can be seen from Table 3, the frequency of use for Internet activities varies greatly when comparing inside and outside school settings. The most popular inside school activity identified by nearly one third of the students (32.1%) was the use of search engines, whereas outside school the most popular activity was to read about entertainment topics (40.4%). Overall, these results show that students engage in tasks related to online communication at much higher rates outside school than inside school.

Research Question 3: How skilled are adolescents at risk of dropping out of school in the new literacies of Internet-based reading, writing, and communicating?

To address this question, students were given a set of questions and tasks that measured specific skills related to online literacy, including (a) formulating questions, (b) searching and locating, (c) reading and evaluating, and (d) communicating. A rubric was created and each item was scored from 0-2 with a score of 0 representing “no response” or “incorrect response,” a score of 1 representing a “moderately skilled response,” and a score of 2 representing a “highly skilled response.” Table 4 provides examples of various skill items, a description of scoring procedures, and sample student responses. Scores from these questions were combined to create a total score for each of the skill areas being measured. Then, the scores for each skill area were combined to give students a total skill score with a range of 0-16 points.

A one-way analysis of variance (ANOVA) was conducted to examine differences on individual skills and total skill among racial groups. The results revealed significant main effects for all skill areas, including formulating questions F(4, 1,013) = 7.29, p = <.01, η² =. 03, searching F(4, 1,013) = 11.67, p = <.01, η² = .04, locating F(4, 1,013) = 6.63, p = <.01, η² = .03, reading and evaluating F(4, 1,013) = 16.88, p = <.01, η²=.06, and communicating F(4, 1,013) = 12.32, p = <.01, η² = .05, as well as the total skill score F(4, 1,013) = 25.05, p = <.01, η² = .09. Post hoc analyses using the Bonferroni procedure were conducted to evaluate pairwise differences among the mean scores for each racial group. Table 5 reports the mean scores for each skill area by racial group and the results of these analyses.

Additional post-hoc analyses using independent-samples t tests were conducted to test for differences in total Internet reading skill based on home computer access, gender, and locale. The tests revealed: (a) students with a computer in their home have significantly higher total skill scores than students without a computer at home, t(1,020) = 6.82,p < .01; (b) female students are significantly more skilled than male students, t(1,020) = 3.61, p < .01; and (c) rural users have significantly higher total skill scores than urban users, t(1,022) = 4.03,p < .01.

Results of this study show that there are clear differences in Internet use and skill level among racial groups and when comparing Internet-based activities inside and outside school. In addition, students in this population have not developed the skills that are necessary for reading, writing, and communicating on the Internet as defined by Leu and colleagues (2004).

There is a large research base that looks at issues of inequality in relation to computer and Internet access, which is widely referred to as the digital divide (Attewell, 2001; Compaine, 2001; Norris, 2001; Warschauer et al., 2004) and is principally associated with economic standing (Fairlie, 2005; Norris, 2001; Lenhart et al., 2005; Reuters Limited, 2003). The results of the current study echo this trend. For instance, a Reuter’s study (2003) indicated that approximately two thirds of low-income homes had Internet access versus 98% of high-income homes, and the Pew Internet and American Life Project ([PIP], Lenhart et al., 2005) reported that 13% of teens that do not use the Internet are among those from homes with lower levels of income. Our results were similar among students at risk of school dropout; only about 73% indicated that they had Internet access at home. These findings have important implications for school success and reducing dropout rates. Research by Fairlie (2005) suggests that access to a home computer increased the likelihood that an individual would graduate from high school. As such, it is important to pay attention to these results and support programs (e.g., Computers for Learning) that provide computers to students in low-income households.

Although the percentage of students who reportedly use the Internet on a regular basis was similar inside and outside school, Internet use patterns for outside school settings were found to be different than those for inside school. Students regularly use the Internet for a greater variety of activities outside of school than they do inside school. For example, students communicate on the Internet much more frequently outside school than they do when inside school, using tools such as email, blogs, discussion boards, chat rooms, and instant messenger. The largest difference was noted with the use of email, the most popular mode of communication reported by participants in this study. Over 35% of students use email on a regular basis outside school compared to only 8.8% using it inside school. It could be argued that one of the most essential skills for nearly any workplace today is the efficient and appropriate use of email (Extejt, 1998), yet over 80% of students in this study indicated that they never read or send email while at school. These results are similar to those of a study that highlighted the “digital disconnect” between teenagers’ use of the Internet while in school compared to outside school (Levin & Arafeh, 2002). That study showed that students’ use of the Internet for educational purposes most often occurred outside of the school day in locations away from the school building. Levin and Arafeh (2002) also reported that the quality of Internet access and restrictions placed on the use of the Internet were among the greatest barriers preventing students from using the Internet at school. Although each of the schools in this study had access to computers, it was only through a central computer lab in most cases. Honan (2008) argues that the placement of computers in a central location, such as a computer laboratory, can act as a barrier because there are typically fewer computers available in classrooms for use throughout the school day. Her observation provides an argument for the implementation of one-to-one laptops in schools, especially since research has documented that the effective use of technology in classrooms can result in better grades, increased standardized test scores, higher levels of school attendance, and overall improvements with student behaviors (Lazarus, Wainer, & Lipper, 2005).

Aside from issues of access and frequency of Internet use, it is also important to pay attention to what Hargittai (2002) refers to as a second-level digital divide, or differences in individual’s skill level with using the Internet. The results of the current study show that students are not highly skilled when it comes to the aptitudes that are necessary for successful online reading. A finding of great concern is the discrepancy in the frequency of school Internet use and Internet skill among the African American population. Students in the African American group had a significantly higher frequency of Internet use in school than Caucasian students, yet were significantly less skilled than both Caucasian and Asian students. Additionally, although Hispanic students did not differ from other groups in their school frequency of use, they scored significantly lower than Caucasian students in all areas on the measure of online reading comprehension. In this case, these discrepancies in skill cannot be attributed to differences in access to computers inside schools. It is more likely that discrepancies occur in the amount and quality of instruction students receive, thus contributing to a second-level digital divide (Hargittai, 2002).

A concerning discrepancy also exists among males and females. Females use the Internet at significantly higher rates outside of school than do males and also have significantly higher online reading comprehension skill scores. This indicates a need to promote Internet use and the acquisition of online reading skills among males in this population. Because males drop out of school at higher rates than females (Greene & Winters, 2006), it becomes especially important to engage them by any means possible. The Internet may be a valuable tool for engaging these students (Becker, 2000) and lowering their likelihood for dropping out (Finn & Rock, 1997).

Overall, students in this population do not use the Internet regularly at school, and they are not skilled Internet users. Findings indicate that students in this study were more proficient with tasks for formulating questions than they were with tasks for locating, evaluating, and communicating information. These results point to a strong need for instruction in the new literacies of the Internet. As the Internet becomes an increasingly prevalent source of information for students both inside and outside school, it is essential that students become proficient users of Internet-based texts.

There are several limitations to this study. First, because technology changes so rapidly, this set of data may quickly lose its currency. Since the time this survey was conducted, the popularity of many Internet applications has grown and the popularity of others may have diminished. Second, all of the data regarding Internet usage is self-report data, which is subject to potential error (Rockwood, Sangster, & Dillman, 1997). Additionally, some schools restrict Internet applications at school. For example, some teachers may wish to teach their students how to communicate using email, but the school may not allow email use by students. This survey did not distinguish between ICTs that were allowed or disallowed by schools in any way.

The Internet may have the potential to narrow the gap between economically advantaged and disadvantaged learners and to foster the engagement that may help keep students in school. Despite this potential, educators are not taking advantage of the Internet as much as they could. Nearly seven percent of students in this study still report never using the Internet while at school. This issue needs to be addressed if we are to combat the problem of school dropout and prepare students for a competitive work place. Additionally, these students should be learning the skills necessary to compete in a posttypographic world (Reinking, 1995). Reading and completing low-level or remedial tasks on the Internet, even if done often, does not provide students with the higher-level reading skills necessary for successful online reading (Leu et al., 2004).

These results support the concept of a second-level digital divide (Hargittai, 2002) among adolescent Internet users both inside and outside of school. The students in this study are shown to have access to computers and the Internet, but their Internet reading skills are not sufficient. With increasing numbers of Internet connected computers in schools, our classrooms are the best place for students to acquire the new literacy skills they will need to compete in the information-driven workplaces of the twenty-first century. The results of this study show that only 15.3% of students send email at school, yet email is the most frequent online activity that adults engage in at the workplace (Taylor, 2001). In today’s technological society, sending email is an essential part of most jobs, yet students are not learning this very basic communication skill at school. Now that we understand how students who are at an increased risk for dropping out of school use the Internet, it is time to develop and implement strategies to increase their Internet use and online reading, writing, and communicating skills.

When it comes to new literacies (Leu et al., 2004), educators cannot assume that students will “figure it out on their own” at home. Students may not have access to the Internet at home, as shown in this and many previous studies. Even if all students have Internet access at home, not all students have a means of learning new and advanced skills with technology, especially those skills needed to navigate the multimodal texts of the Internet. Thus, we need to pay attention to whether or not we are preparing our students for this changing landscape of literacy by including “a broadened view of text and the multiliteracies made possible in today’s new information communication technologies” (Alvermann, 2005, pp. 10-11).Incorporating Internet-based learning into the curriculum has the potential to engage students who are at an increased risk for dropping out of school and to decrease the gap between advantaged and disadvantaged students and among racial groups by providing scaffolded opportunities to acquire the new literacies that will define their future. A second-level digital divide will continue to grow unless there is a mechanism to ensure that all students are taught higher-level Internet reading comprehension skills, rather than just how to perform basic tasks. As educators, we hold the key to a promising future for our students. It is unfair to leave them unplugged, disengaged, and unmotivated to learn.

This research was supported by Grant #R305G050154 from the United States Department of Education, Institute of Education Sciences, which was conducted in collaboration with members of the TICA Project Team at Clemson University and the University of Connecticut.