Re-thinking scientific literacy out-of-school: Arguing science issues in a niche Facebook application

Computers in Human Behavior xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Computers in Human Behavior journal homepage: www.elsevier.com/locate/comphumbeh

Re-thinking scientific literacy out-of-school: Arguing science issues in a niche Facebook application Christine Greenhow a,⇑, Thor Gibbins b, Melissa M. Menzer c a Counseling, Educational Psychology and Special Education, College of Education, Michigan State University, 513F Erickson Hall, 620 Farm Lane, East Lansing, MI 48824, United States b Department of Teaching and Learning, Policy and Leadership, College of Education, University of Maryland, 2311 Benjamin Building, College Park, MD 20742, United States c Human Development, College of Education, University of Maryland, 3304 Benjamin Bldg, College Park, MD 20742, United States

a r t i c l e

i n f o

Article history: Received 29 December 2014 Revised 3 June 2015 Accepted 8 June 2015 Available online xxxx Keywords: Scientific literacy Social media Facebook Computer-supported collaborative learning Informal learning Argumentation

a b s t r a c t Social network sites (SNSs) like Facebook.com, are the dominant technology-mediated leisure activity among teenagers in different countries, prompting researchers to explore their suitability as learning tools, largely in formal higher education settings, and with mixed results. In contrast, this paper examines whether an open-source social networking application implemented outside of the school context engaged young people (ages 16–25) in debating socio-scientific issues. A multi-dimensional approach to analyzing argumentative knowledge construction in a designed Facebook.com application yielded insights about the presence and nature of young people’s socio-scientific issue argumentation along four process dimensions (participation, argumentative, epistemic, social co-construction). We discuss the implications of these findings for computer-supported collaborative learning (CSCL) theory and the design of similar applications that attempt to supplement formal learning or bridge formal-informal learning settings. Ó 2015 Elsevier Ltd. All rights reserved.

1. Science literacy and argumentation in social network sites Improving adolescents’ scientific literacy, preparation for the 21st century workplace, and engagement in current affairs are critical problems facing U.S. education (Bureau of Labor Statistics, 2007; Collins & Halverson, 2009; National Center for Education Statistics (NCES), 2005). A 2012 Program for International Student Assessment (PISA) revealed that U.S. students ranked 23rd in science (Kelly et al., 2013), and the U.S. had lower rates of American undergraduates earning STEM degrees compared to other countries despite higher per-pupil spending (National Science Board, 2010). Furthermore, research in the U.S. continues to document young people’s disengagement from school (Levin & Arafeh, 2002) and public life (Putnam, 1995): more than one-third of people under age 25 do not get any news on a daily basis (Pew Research Center for the People, 2008). Within the broad field of science education, conceptualizations of scientific literacy or science literacy has typically emphasized learners’ coming to know science as practicing experts do (e.g., developing content knowledge and authentic inquiry processes). In recent years, the definition of scientific literacy has emphasized ⇑ Corresponding author. E-mail addresses: [email protected] (C. Greenhow), [email protected] (T. Gibbins), [email protected] (M.M. Menzer).

reading, writing and communicating about current science topics— herein referred to as socio-scientific issues (SSIs)—for civic, cultural, and personal understanding as these topics relate to everyday life and policymaking (Polman et al., 2010). Argumentation or negotiation competencies, such as participation, epistemic skills, argument skills, and social co-construction skills (Sadler & Fowler, 2006; Weinberger & Fischer, 2006) are important to engaging in socio-scientific issues. Argumentation of SSIs have been argued to be powerful facilitators for teaching and learning science (Barab, Sadler, Heiselt, Hickey, & Zuiker, 2007; Cavagnetto, 2010; Sadler, Barab, & Scott, 2006), particularly in formal learning settings, such as online (Yeh & She, 2010), offline (Berland & Hammer, 2012; Nussbaum & Edwards, 2011; Yoon, 2011), and in museums (Gutwill & Allen, 2012). Such studies have also identified the challenges of inserting SSIs into already time-strapped classrooms and test-driven curriculum. Yet, socio-cultural learning theory and theories of computer-supported collaborative learning have suggested that informal environments may also be particularly supportive of learning and have argued that a considerable amount of learning occurs through informal interactions with others, reading, and observation (Brown, Collins, & Duguid, 1989; Greeno, 1989; Lave & Wenger, 1991; Vygotsky, 1978). However, little is known about informal learning processes within out-of-school online contexts,

http://dx.doi.org/10.1016/j.chb.2015.06.031 0747-5632/Ó 2015 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Greenhow, C., et al. Re-thinking scientific literacy out-of-school: Arguing science issues in a niche Facebook application. Computers in Human Behavior (2015), http://dx.doi.org/10.1016/j.chb.2015.06.031

2

C. Greenhow et al. / Computers in Human Behavior xxx (2015) xxx–xxx

such as those within online SNS (Lenhart, Purcell, Smith, & Zickuhr, 2010; Rideout, Foehr, & Roberts, 2010). Thus, the focus of the current study was to extend socio-cultural learning theory and computer-supported collaborative learning theory developed in formal learning contexts to informal contexts by (1) examining the nature and extent to which young people engage in a variety of argumentation skills around socio-scientific issues (i.e., climate change issues) within an informal online social networking application (HotDish) in Facebook.com and (2) exploring whether and how argumentation skills are related to each other within this context. Next, we situate our work within previous research and theory in argumentation and computer-supported collaborative learning. We then introduce the socio-technical features of the social networking application, Hot Dish, before providing an explanation of our methods and subsequent presentation and discussion of results. In the conclusions section we suggest the strengths and limitations of this work and suggest areas for future study. 1.1. Socio-scientific issue argumentation Sociocultural learning theories have assumed that learning is derived from participation in joint activities, inextricably tied to social practices, and is mediated by artifacts over time (Greeno, 1989; Lave & Wenger, 1991; Vygotsky, 1978). Such theories have helped researchers conceptualize and study learning as participation in communities of practice (Wenger, 1998). Highlighted by science educators and new science standards for its role in science learning (Yeh & She, 2010), researchers have argued that scientific literacy requires argumentation of SSIs in formal science classrooms (Barab et al., 2007; Sadler et al., 2006; Zeidler, Walker, Ackett, & Simmons, 2002). Furthermore, some have proposed that argumentation of SSIs in science classrooms and beyond is essential to developing modern scientific literacy, especially during activities that prompt certain types of learning behaviors important for increasing students’ use of argumentation skills (Berland & Hammer, 2012; Chin & Osborne, 2010; Nussbaum & Edwards, 2011; Sadler et al., 2006; Yoon, 2011; Zeidler et al., 2002). There are four types of argumentation skills: participation, epistemic skills, argument skills, and social co-construction skills (Sadler et al., 2006; Toulmin; 1958; Weinberger & Fischer, 2006). 1.1.1. Participation Participation refers to how much someone participates in argumentation (Weinberger & Fischer, 2006). Scholars have noted that participation is important in constructing knowledge and initiating collaboration and cooperation. For example, Cohen and Lotan (1995) argued that elementary school classroom contexts that have high interaction among students are more likely to have higher levels of participation at the individual level compared to classrooms with less interaction among students. It has also been speculated that engaging in content with others is associated with academic success and learning. 1.1.2. Epistemic skills Another form of argumentation is epistemic skills, such as complexity, skepticism, and inquiry (Sadler et al., 2006). Complexity refers to the ability to use multiple sources of information and perspectives during problem-solving discussions; using multiple sources and perspectives may complicate the picture regarding the socio-scientific issue and the solution to the issue. For example, when discussing whether using solar and wind power is better than using nuclear or fossil power, the advantages and disadvantages of using one strategy or another may become more complex as additional sources and perspectives are included in the discussion. Sadler et al. (2006) defined skepticism as the ability to

question the validity or reliability of information based on the source (Kolsto, 2000). Inquiry referred to being able to question or gather information in order to begin finding a solution or compromise. Learners who use sophisticated epistemic skills are more successful at understanding theoretical concepts and complex problems and issues (Fischer, Bruhn, Gräsel, & Mandl, 2002; Hogan, Nastasi, & Pressley, 1999; Salomon & Perkins, 1998). Often SSIs are complex problems that do not have clear-cut solutions. Thus the ability to think about multiple perspectives, to be skeptical about multiple views or perspectives based on the source of evidence, and to ask questions that dig deeper into socio-scientific issues would be associated with learning (Sadler et al., 2006). 1.1.3. Argument skills Argument skills refer to the ability to construct a case for potential solutions or perspectives (Toulmin; 1958; Weinberger & Fischer, 2006), though these reasons need not be persuasive or justified. Argument construction can comprise claims, grounds with warrant, and qualifiers. Claims refer to simple statements that present an argument. Qualified claims or qualifiers refer to claims that also provide limitations to the claim. Grounds with warrant refers to claims that do not include a qualifier but do provide supporting evidence for the argument; this type of skill is the most salient form of arguments that learners use. In addition to being able to construct an argument, argument competence also includes the ability to carry out a dyadic or group debate with others, through the following steps: (1) arguments, (2) counterarguments, and (3) integrated replies (Clark & Sampson, 2008; Leitão, 2000; Weinberger & Fischer, 2006). In other words, first an argument or case must be presented, then, a challenge or counterargument to the initial argument is presented. Following this challenge, integrated replies are presented, which synthesize the arguments and counterarguments in order to advance or move forward the discussion. In each of these processes, scholars have argued that learners explore the material and build their argument; other learners may challenge the first learner’s position by bringing thoughtful insight to the interpretation of the material or bringing in prior knowledge; and lastly, both learners work together to reach a consensus (Weinberger & Fischer, 2006). Importantly, argument skills have been shown to be conducive to learning (Andriessen, Baker, & Suthers, 2003), and the ability to engage in debate has been shown to foster perspective-taking skills (Spiro, Feltovich, Jacobson, & Coulson, 1991), which are important for modern notions of scientific literacy. 1.1.4. Social co-construction skills The fourth and final form of argumentation skills is social modes of co-construction, which refers to whether learners collectively acknowledge and draw on the contributions of their learning partners (Brown et al., 1989) through the use of externalization; elicitation; and consensus-building moves (Weinberger & Fischer, 2006). Externalization is defined as statements that describe basic thoughts, or one’s personal feelings or reactions to a group (e.g., this makes me angry). Elicitation is defined as when learners request other sources of information or knowledge by questioning or trying to get another learner to expand on his or her initial statements. For co-construction tasks, learners can use three types of consensus-building techniques to compromise on a course of action or an idea (Weinberger & Fischer, 2006). Researchers have defined quick-consensus building as accepting arguments from other learners, regardless of whether or not he or she agrees or disagrees with the argument (Clark & Brennan, 1991; Fischer et al., 2002; Weinberger & Fischer, 2006). We might think of this as accepting arguments uncritically despite contradictions and

Please cite this article in press as: Greenhow, C., et al. Re-thinking scientific literacy out-of-school: Arguing science issues in a niche Facebook application. Computers in Human Behavior (2015), http://dx.doi.org/10.1016/j.chb.2015.06.031

C. Greenhow et al. / Computers in Human Behavior xxx (2015) xxx–xxx

inconsistencies (Kanuka & Anderson, 1998). Integration-oriented consensus building is defined as being persuaded to change ones position based on the persuasive arguments of others. Within formal learning environments, researchers claim that integration -oriented consensus building is the least used technique of all social co-construction techniques (Weinberger & Fischer, 2006), and research has been mixed regarding the utility and function of integration-oriented consensus building for learning (Fischer et al., 2002). Conflict-oriented consensus building is comprised of disagreements with or modifications of other learners’ perspectives or arguments. According to scholars, conflict-oriented consensus building is the most important and most sophisticated component of social perspective-taking in collaborative learning environments (e.g., Doise & Mugny, 1984; Teasley, 1997). This form of consensus building helps all learners to use multiple perspectives in order to build a stronger, more concise argument (Chan, Burtis, & Bereiter, 1997). However, research has emphasized that rhetoric style matters; disputative forms of argumentation (a focus on undermining others’ claims and winning the debate) rather than deliberative forms (a focus on critically, yet collaboratively exploring different viewpoints) have been shown to negatively affect learning through argumentation (Asterhan & Hever, in press). Lastly, in terms of sophistication in social co-construction techniques, scholars have argued that techniques in order of least sophisticated to most sophisticated are (1) externalization, (2) elicitation, (3) quick consensus building, (4) integration-oriented consensus building, and (5) conflict-oriented consensus building (Teasley, 1997). The four forms of argumentation (participation, epistemic skills, argument skills, and social co-construction skills) may also be related to each other. As previously discussed, the ability to communicate about socio-scientific issues requires skills that overlap in all four dimensions. For example, scholars have noted that participation is important for effective collaboration and cooperation between people (Cohen & Lotan, 1995), some of who may differ in their opinions and perspectives (Sadler et al., 2006). To be able to navigate, negotiate, and integrate differing perspectives require some level of participation in the conversation through the use of arguments, counter-arguments, and integrated replies (Leitão, 2000; Weinberger & Fischer, 2006) and different forms of consensus building (Weinberger & Fischer, 2006). In other words, it was expected that there would be strong linkages between these four forms of argumentation.

1.2. Formal and informal learning environments The literature described thus far has suggested that all forms of argumentation skills are important for developing young people’s SSI negotiation and scientific literacy. However, much of the synthesized research has predominately been conducted within formal learning environments, including computer-supported collaborative learning (CSCL) environments (Weinberger & Fischer, 2006). Within formal CSCL environments, learners frequently use argumentation techniques by engaging in discourse through non face-to-face, text-based interaction, such as via discussion boards or chats. Discussion boards are asynchronous interactions that provide the opportunity to ‘‘think before posting;’’ learners are given ample time to construct and reconstruct their contributions prior to making their viewpoints known to other learners (Stegmann, Wecker, Weinberger, & Fischer, 2012). Under these constraints, learners are generally able to contribute to discussions at their own convenience and pace. Kol and Schcolnik (2008) have noted that these sorts of asynchronous interactions are likely to lead to further participation and discourse.

3

CSCL researchers often include scripts that encourage specific epistemic, argument, or social co-construction skills (Weinberger & Fischer, 2006) but may also hinder other argumentation skills. For example, Weinberger and Fischer (2006) found that epistemic scripts bolstered the use of epistemic skills but decreased the prevalence of argument skills and social co-construction skills. However, some researchers have argued that the intended affect of a prompt may not occur in formal CSCL environments because students may ignore prompts or not use them as frequently as the teacher intends, and when students do use prompts their responses may lack the sophistication a teacher wants or expects (for a review, see Furberg, 2009). Other researchers have found that CSCL activities that did not include a prompt were more likely to lead to sophisticated oral discourse, and subsequently, to learning and knowledge building compared to activities with prompts (Davis, 2003; Davis, 2004; Furberg, 2009; Renkl, 2002; Weinberger & Fischer, 2006); that is, learners in an unstructured but formal learning environment appeared to learn more than did learners in either structured or scripted environment. Studies of formal environments that have appropriated social network sites (SNS) as a medium for teaching argumentation, though few, have yielded promising findings. For instance, Beach and Doerr-Stevens (2011) explored how students engaged in argumentation about their school’s Internet policies within the social network site, Ning. Through collaborative role-play of divergent discursive communities or perspectives (e.g., business management perspective, legal/civil liberties perspective, authoritarian/parental perspective) the students challenged each other’s as well as their own perspectives on the issue, recognized alternative arguments, and ultimately, were able to convince school administrators to unblock previously-blocked sites. Beach and Doerr-Stevens conclude that social network sites can facilitate argumentation when the issues students are addressing have some direct bearing on their lives, rather than simply used as sites to share information. Similarly, Tsovaltzi, Puhl, Judele, and Weinberger (2014) investigated the effects of argumentation scripts in Facebook among German university students who were part of a formal course for teacher trainees. They found that the argumentation script helped learners create arguments of higher quality and structure. Taken together, these studies suggest that social network sites can have affordances for developing students’ abilities to defend arguments, develop counter-arguments, and elaborate upon their own arguments within formal learning environments. Given the findings related to formal learning environments that had minimal constraints or parameters, perhaps online, informal, non-school learning contexts can supplement or bolster the development of scientific literacy and learning (Palmquist & Crowley, 2007; Sefton-Green, 2004). Research has begun to examine scientific literacy within socio-technical spaces that dominate popular culture, such as online social network sites (SNS) (Greenhow, 2011a,b). SNSs, such as Facebook.com or LinkedIn.com, are a form of social media defined by: (1) uniquely identifiable profiles that consist of user-supplied content and/or system-provided data; (2) (semi-) public displays of connections that can be traversed by others; and (3) features that allow users to consume, produce, and/or interact with user-generated content provided by their connections on the site (Ellison & Boyd, 2014, p. 7). As the widespread use of virtual environments like Facebook.com has attested, young people are increasingly invested in their online identities, relationships, and knowledge (Greenhow, Robelia, & Hughes, 2009). However, few studies have explored the link between SNS use and education (Gibbins & Greenhow, 2014; Greenhow & Robelia, 2009a,b; Junco, 2012a,b, 2013; Manca & Ranieri, 2013; Mazer, Murphy, & Simonds, 2007; Ranieri, Manca, & Fini, 2012), and learning sciences research has yet to examine what these online exchanges reveal about learning processes, such as argumentation.

Please cite this article in press as: Greenhow, C., et al. Re-thinking scientific literacy out-of-school: Arguing science issues in a niche Facebook application. Computers in Human Behavior (2015), http://dx.doi.org/10.1016/j.chb.2015.06.031

4

C. Greenhow et al. / Computers in Human Behavior xxx (2015) xxx–xxx

1.3. Overview of the current study and hypotheses To investigate the presence and nature of young people’s SSI argumentation, we extended an argumentation coding system originally designed for formal learning environments (Sadler et al., 2006; Weinberger & Fischer, 2006) to the informal learning context: an open-source Facebook.com application called Hot Dish. Learners who posted content or responses to content within the Hot Dish Facebook application were coded for argumentation skills (participation, epistemic skills, argument skills, and social modes of co-construction skills). We then explored whether the behaviors for each process dimension were part of the same underlying constructs with exploratory factor analyses. Third, we examined relations between these argumentation factors to further understand SSI argumentation in informal, online learning contexts. Based on prior research, we speculated that Weinberger and Fischer’s model of argumentation would occur in the Hot Dish application (for a description of the application, see below) focused on engaging young people in socio-scientific issues. Results from a pilot study suggested that the young people were drawn to the social networking application because they were interested in its focal topic and perceived it as a place to engage with like-minded others (Greenhow, 2010, 2011a). We hypothesized that SNS users would use argument and epistemic skills by debating and arguing issues in Hot Dish, and that we would see evidence of people’s building on each other’s contributions (i.e., social co-construction skills) as we have seen in other more formal online learning, or collaborative knowledge building communities. 1.3.1. Hot dish application Hot Dish is an open-source social networking application within Facebook.com designed to encourage information-sharing, commentary, and problem-solving on environmental science and climate change issues (for a screenshot, see Fig. 1). Launched in February 2009, Hot Dish registered 1157 members over a four-month period. Youth (grades 10–16) were invited to join.

Hot Dish offered multiple channels for sharing knowledge about these issues and meets the definition of a SNS (Ellison & Boyd, 2014). An environmental science organization provided some editorial content. Additional features facilitated user-generated content in the form of posting original article entries or circulating existing articles from online sources. Members could read an article’s overview or read the full article; curate and rank posted articles: vote them up, customize the entry title, or write a short summary; comment on articles: share them within or outside the social network; and tweet and chat about them. Individuals could craft a self-profile in the profile area. The Hot Dish profile, separate from the user’s Facebook profile, was featured within the Hot Dish application. Similar to Facebook.com, members portrayed their background, interests, and ideas through online photos, bio, blog and data-reporting features, which showcased what Hot Dish users had contributed. The application automatically tracked participants’ use of these features so that usage statistics could be generated and analyzed for those who agreed to take part in the research study. In addition, the application sought to promote pro-environmental activism and civic engagement around environmental science issues through a series of Action Team challenges. In their review of the environmental science literature, Heimlich and Ardoin (2008) stressed the importance of providing learners with opportunities to act on attitudes, positing that actions strengthen relevant attitudes. Action Team challenges provided incentives to learn new personal action strategies as well as opportunities to practice them. Challenges were offered both online and offline. For example, users could visit other sites to sign online petitions or learn more about climate change. Users could also complete offline challenges such as recycling old electronics, volunteering for an environmental organization, writing letters to the editor, or attending a town meeting with an environmental issue on the agenda. Offline challenge completion required uploading documentation (text, video, images) for subsequent evaluation by the project partners.

Fig. 1. Screenshot of the Hot Dish Facebook application.

Please cite this article in press as: Greenhow, C., et al. Re-thinking scientific literacy out-of-school: Arguing science issues in a niche Facebook application. Computers in Human Behavior (2015), http://dx.doi.org/10.1016/j.chb.2015.06.031

C. Greenhow et al. / Computers in Human Behavior xxx (2015) xxx–xxx

2. Method 2.1. Participants and procedure The present study was part of a larger investigation of informal learning among high school and college students’ within the Hot Dish application (Robelia, Greenhow, & Burton, 2011). Participants were recruited to use Hot Dish largely through our partnership with Grist environmental magazine, who invited potential users through their listserv of subscribers. We also reached out to environmental science programs at U.S. colleges and universities and to high school science teacher listservs, inviting them to disseminate information about the Hot Dish application to their students. Young students who elected to use the Hot Dish application did so voluntarily and were not part of any formal science, environmental science or related course. When signing up for the application, individuals were asked to participate in the research study and were asked to indicate their level of interest in climate change. At the end of the application pilot, users were asked to take an online survey, which included demographics (e.g., age, gender, ethnicity, level of education, family’s educational experience, which is often used as a proxy for income, technology access) and their context (e.g., city, state, country, and school/college major interest), among other information such as their level of interest in climate change after using Hot Dish. Of the 346 total users (ages 16–25) who opted into this study, 31 users were subsampled because of their participation in posted comment strings in response to news articles that were posted in the application. The focus of the current study was on these 31 users (77% female). 11% of the final sample had a high school diploma or below, 44% had 1-to-3 years of college, 39% had

5

graduated college, and 5% had 1-to-3 years of graduate school. Preliminary chi-square analyses revealed that the group of 31 users were not significantly different from the group of users who did not participate in commenting (n = 315) with regard to college major (v2 = 18.45, p = .14), age (v2 = 14.23, p = .08), gender (v2 = 1.66, p = .20), race (v2 = 1.23, p = .87), and interest in Hot Dish prior to the study (v2 = 4.09, p = .67). 2.2. Data sources and measures 2.2.1. Observed discourse and argumentation User-generated online content, in the form of online comments on discussed articles, were collected and coded for evidence of argumentation about environmental issues. In total, 220 articles and the 971 comment strings (unit of analysis) posted to those articles were included in this analysis of SSI argumentation in the Hot Dish environment. Importantly, all 220 articles included three or more comment strings. We coded the comment strings described above along the aforementioned four dimensions: participation, epistemic, argument, and social co-construction skills (see Table 1 for codes, definitions of codes, and examples). Each skill was coded for its presence or absence, and skills were not mutually exclusive. In other words, a comment string could be coded for the presence of multiple skills. Methodologically, Weinberger and Fischer (2006) have described participation as the quantity of words used during discourse, and the heterogeneity of participation as the proportion of words used by one learner divided by the total number of words used by all learners together. For the purposes of the current study, we operationalized participation based on unique comment strings contributed by users rather than at the word-level in order to be consistent with how we coded epistemic, argument, and social

Table 1 Descriptions of argumentation codes. Dimension/Code

Definition

Example

Argument Claim

A simple statement that presents an argument

‘‘Wow, Shell Oil, we think your commitment to ‘tackl[ing] climate change’ just fell off the back of the truck.’’ ‘‘There will always be kinks with things first start out. . .it’s science people.’’ ‘‘Although, vanilla soy milk is great with cereal in the morning.’’ ‘‘What do you mean by profitable. . .that has a lot of definitions.’’ ‘‘How about buying used? Or restyling your old clothes to make them new again? It’s more fun that way.’’ ‘‘This is wonderful! And the gardener in me is drooling at all of those lovely, diverse varieties.’’

Ground with warrant Qualifying statement Other POV

A claim that does not include a qualifier but does provide supporting evidence for the argument A claim that also provides limitations to the claim

Counter-argument

A claims that challenges a previously stated claim

Integrated reply

A statements that synthesizes arguments and counterarguments in order to advance or move forward the discussion

Epistemic Complexity Inquiry Skepticism Non-epistemic talk Social Elicitation Externalization Conflict oriented consensus building Integration oriented consensus building Quick consensus building

An arguments that includes a reference to another person’s contributions

Using multiple sources and perspectives in understanding the material and engaging in problem-solving discussions Questioning or information gathering in order to begin finding a solution or compromise Questioning the validity or reliability of the information gathered or the person from which the information was gathered Off-task talk, or talk that is unrelated to the material at hand

‘‘True. And even though it’s valuable to encourage people to unplug their electronics when not in use.’’ ‘‘This links may help. [reference to a website]’’

Using other learners as a source of information or knowledge by questioning or trying to get another learner to expand on their initial externalization Proposing a basic thought to a group Using disagreements or modifications of other learners’ perspectives or arguments

‘‘How much more are Americans willing to pay for health care to treat all the illnesses that our dirty energy is causing?’’ ‘‘Right on. Showing the nation how agriculture should be done: small scale, local, pesticide and herbicide free.’’ ‘‘Regardless of what Congress knows, our representatives need to realize that We the People want them.’’

Being persuaded to change ones position based on the persuasive arguments of others

‘‘None of the companies on this list surprised. What does surprise me is that Fiji even makes a pretense of trying.’’

Accepting arguments from other learners, regardless of whether he or she agrees or disagrees with the argument

‘‘Thanks for reminding me to stay positive and optimistic’’

‘‘Really?’’ ‘‘Hahaha, and they’re going to be the worlds ‘‘greenest’’ city!’’

Please cite this article in press as: Greenhow, C., et al. Re-thinking scientific literacy out-of-school: Arguing science issues in a niche Facebook application. Computers in Human Behavior (2015), http://dx.doi.org/10.1016/j.chb.2015.06.031

6

C. Greenhow et al. / Computers in Human Behavior xxx (2015) xxx–xxx

skills of argumentation. Frequency scores for each participant were calculated by summing up the number of comments that the participant posted for each story. Heterogeneity of participation was also calculated by taking the proportion of comments by the participant and dividing that score by the total number of comments by all users. For each comment string, the presence or absence of five argument skills were coded: Claim, GWB (grounds, warrants, backing), Qualifying Statements, Other Point of View, and Integrated Reply (see Table 1). For example, one participant, Jen, responded to an article by describing a hotel’s decision to stop giving newspapers to every occupant in order to reduce its carbon emissions. Jen’s comment reveals the beginning of a constructed argument: ‘‘This is good. A lot of people in hotels don’t even read them [newspapers]!’’ The first clause, ‘‘This is good,’’ states a simple claim that the hotel is making the correct decision; while the second clause, ‘‘A lot of people in hotels don’t even read them!’’ provides grounds. For each comment string, the presence or absence of three epistemic skills were coded: Complexity, Inquiry, Skepticism (see Table 1). The following exchange between Doug and Dwight, stemming from an article posted in Hot Dish on fusion energy, illustrates all three epistemic skills: Doug: I’ve read about the NIF and fusion power in general, and asked a couple scientists about the principles, and while there may be negatives, there are inherently fewer than in either coal (or oil) plants, or fission reactors. Fission power, by its very nature, is designed to be self-sustaining (control rods exist to intentionally disrupt the process), which means that things can get awful (Chernobyl) and stay bad (nuclear waste) for a long time, so fission reactions escalate unless everything goes right. By contrast, fusion reactions collapse unless everything goes right. If anything goes wrong, the only ones who’ll get hurt are the workers of the fusion plant themselves. As for emissions, fusion power produces almost exclusively helium, but only in minuscule amounts, and helium is harmless in any case. The negatives of hydrocarbons are inherent to the myriad of chemicals contained within. Fusion follows Thoreau and simplifies. . . Dwight: The problem here is that fusion is an exotic technology that has been promised to solve the energy problem. . .but never quite gets there. The problem is ‘‘breakeven’’: it takes an enormous amount of energy to start and maintain a fusion reaction. . .so far, more than you get out of it. Fusion research is enormously expensive. So, my question is, are we better off spending billions of dollars on something that may never work out, vs spending it on actually installing renewable generation and getting the coal plants shut down pronto. We’re running out of time here folks, and we have to spend money wisely. Doug’s comment on the possible promises of fusion energy indicate the complexity of the issue. He cites scientists he knows who serve as additional sources of information; they assert that there are ‘‘fewer negatives’’ to fusion energy compared to fossil fuels and fission energy as energy sources. He also considers the issue from public health and safety and environmental perspectives, citing the Chernobyl nuclear meltdown in the 1980s as a disastrous negative consequence of fission reactions that escalate, uncontained. Dwight counters Doug’s initial argument, also demonstrating complexity in his response. His counterargument brings an economic perspective to bear on the issue; he argues that fusion energy is an ‘‘exotic technology’’ that research has not proven energy-efficient or effective and the research needed to do so is expensive. Dwight demonstrates skepticism of the article’s orientation toward adopting fusion energy by highlighting the long-standing, but yet fulfilled ‘‘promise’’ of fusion energy from

the scientific community in light of the carbon neutral problem, with fusion requiring ‘‘an enormous amount of energy.’’ Dwight moves from skepticism into eliciting further discussion by inquiring whether the solution should be to invest a considerable amount of money in a technology that may never work or in renewable resources that can begin to replace the fossil-fuel burning energy sources currently used in the United States. In addition, five codes referred to social co-construction skills: Externalization, Elicitation, Quick Consensus-Building, Integration-Oriented Consensus-Building, and Conflict-Oriented Consensus-Building (see Table 1). Each comment string was coded for the presence or absence of the five codes above. In the initial post that began the discussion thread on fusion energy, Jen posted, ‘‘Well this sure sounds good! But so did oil at first. . . hmm. . . i [sic] wonder what the negatives are, they don’t mention that!’’ Jen’s post externalizes her feelings about the idea of fusion energy sounding ‘‘good,’’ at the same time eliciting other participants for more information on the possible negative consequences. Doug’s reply to Jen, as noted in the preceding paragraph, integrates her invitation to consider the negatives of fusion energy and builds on it. Dwight’s counter-argument, also noted in the previous paragraph, illustrates an example of conflict-oriented consensus building in that he highlights his disagreement with Doug’s post supporting the promise of fusion energy. Dwight’s response also illustrates the sophistication of conflict-oriented consensus building in that Dwight has modified the initial problem brought up by Jen’s elicitation (to explore the negatives of fusion energy) with the immediate need to replace fossil-fuel related energy sources. After several conflict-oriented exchanges between Doug and Dwight where they debate the positive and negative points of nuclear energy sources, Jen closes the discussion thread with a quick-consensus building post: ‘‘wow [sic]. . . it will take me a while to read all of this! but im [sic] glad we got a good discussion going!’’ Jen’s post quickly ends the debate without truly accepting either Doug or Dwight’s position; however, she acknowledges both arguments by pointing out that she will take her time to read them. 2.2.2. Training and reliability One master coder and two trained coders coded all of the comment strings. The master coder (the first author) met with the trainees to first discuss the codes and then they coded a few of the comment strings together. Then the master coder and the trainees independently coded several comment strings and reliability was calculated using Cohen’s Kappa (1960); the master coder and the trainees discussed any discrepancies. This process continued until a satisfactory reliability statistic was achieved (j > .80). Once reliability was attained, the trainees coded the remaining comment strings independently. A random sample of comments posted to stories equaling 20% of the data were double-coded by the master coder and inter-rater reliability was calculated: Argument skills (j = .85), Epistemic skills (j = .89), and Social skills (j = .89). 2.3. Data reduction and analysis First, descriptive means and standard deviations of the comment data were described in terms of specific argumentation skills used. Second, within each dimension of argumentation, we explored the nature of the skills with exploratory factor analysis to determine whether argument skills, epistemic skills, and social co-construction skills emerged as overarching factors or whether there were subsets of skills within each skillset. Exploratory factor analysis (EFA) reveals patterns of correlations among observed argumentation behaviors that are thought to reflect underlying or latent processes within specific dimensions of argumentation (e.g., within epistemic skills, do certain types of skills relate to each other or do they represent differing and unique skills). This type of

Please cite this article in press as: Greenhow, C., et al. Re-thinking scientific literacy out-of-school: Arguing science issues in a niche Facebook application. Computers in Human Behavior (2015), http://dx.doi.org/10.1016/j.chb.2015.06.031

C. Greenhow et al. / Computers in Human Behavior xxx (2015) xxx–xxx

analysis allows researchers to explore the factor analysis structure to see whether the data itself can be reduced into meaningful subscales. For each process dimension, behaviors were entered into an Exploratory Principal Component Factor Analysis (EFA). Given than only one paper to our knowledge has used factor analysis to examine CSCL environments (O’Mahony et al., 2012), we have opted to use a similar Varimax rotation, or orthogonal rotation, in order to decrease the inter-correlation of factors and provide a cleaner, more parsimonious structure. Similar analyses using an oblique rotation were also conducted and these analyses yielded similar factors and results. Factors with eigenvalues above one were retained in the final models. Factor scores were then created by averaging the behaviors that highly loaded on each of the factors. The last analytical process for the current paper included a bivariate correlational analysis of these factor scores to examine the relations between factors. 3. Results and discussion Little work has focused on argumentation in informal learning environments. We aimed to expand current theory of SSI argumentation within formal CSCL environments (e.g., Sadler et al., 2006; Weinberger & Fischer, 2006) by exploring young people’s asynchronous interactions with peers around environmental science and climate change issues within an out-of-school social networking application within, Facebook.com. Frequencies, means, and standard deviations for all epistemic, argument, and social co-construction skills are presented in Table 2. For epistemic skills of argumentation, comment strings most often comprised Complexity skills. For argument skills of argumentation, comment strings most often comprised Claims and Grounds with Warrants (GWB). For the social co-construction skills of argumentation, comment strings most often comprised Externalization and Integration-Oriented Consensus Building skills. 3.1. Participation Of the 346 members of Hot Dish who consented to be in our study, 31 users commented or responded to comments about a posted story. Participation frequency and proportion scores were calculated for these 31 users. On average, users posted 31.32 Table 2 Means and standard deviations of all argumentation skills. Frequency

Mean

Std. deviation

Argument Claims Grounds, Warrants, & Backings Qualifying Statements Other Points of View Counterarguments Integrated Replies Non-Argument Skills

343 345 60 49 128 216 68

.35 .36 .06 .05 .13 .22 .07

.48 .48 .24 .22 .34 .42 .26

Epistemic Complexity Inquiry Skepticism Non-Epistemic Skills

792 86 215 164

.82 .09 .22 .17

.39 .28 .42 .38

144 328 225 270

.15 .34 .23 .28

.36 .47 .42 .45

105

.11

.31

Social Elicitations Externalizations Conflict-Oriented Consensus-Building Integration-Oriented ConsensusBuilding Quick Consensus-Building Note: N = 971 comment strings.

7

comment strings (SD = 69.19). The average participation heterogeneity score, or the likelihood that two comment strings were posted by two different people was .28 (SD = .07). In other words, participation heterogeneity was low, suggesting that the majority of the comment strings were posted by a small number of users.

3.2. Argument skills First, we saw evidence of users demonstrating argumentation in the online commentary on Hot Dish, specifically within the argument dimension. According to Weinberger and Fischer (2006), there are three subsets of skills within the argument dimension that should emerge in formal CSCL environments: arguments, counterarguments, and integrated replies. Exploratory factor analyses, specifying Varimax rotation and eigenvalues above one, revealed a three-factor model for argument skills (see Table 3 for factor loadings). The Counterarguments Factor (Factor 1) comprised the argument behaviors of Counterarguments and Other Point of View; the Arguments Factor (Factor 2) comprised the Claims, GWB, and Qualifying Statements behaviors; and lastly, the Integrated Replies Factor (Factor 3) comprised only the Integrated Replies behavior. In other words, arguments, counterarguments, and integrated replies generally do not occur within the same comment and thus suggests that each of these skills display unique characteristics. The results of this study revealed that these three skill subsets were also evident in informal learning environments at relatively high rates, though arguments were used the most. It is likely that arguments should occur more often than other skills, because the argument processes occur in three sequential steps: (1) presentation of an argument by Person A, (2) presentation of a counterargument by Person B, (3) the integrated reply or reaction by Person A (Leitão, 2000). Thus, by definition, arguments should be more prevalent than counterarguments and integrated replies. There were also some cases where skill-sets within the argument dimension were differentially related. For example, arguments and counterarguments were positively related to each other. This makes sense; as people state their assertions in order to persuade their peers, they logically follow those assertions with evidence for their viewpoint and this evidence can take several forms. Indeed, in order to present a counterargument, there already must be an argument that has been presented (Leitão, 2000; Weinberger & Fischer, 2006). However, integrated reply, or a statement that synthesizes arguments and counterarguments in order to advance or move forward the discussion, was negatively related to both arguments and counterarguments. It should be noted that many of the articles which users posted comments to had three comments posted to them, and that the participant heterogeneity of the posts varied considerably. Thus, perhaps, it is plausible to suggest that in many cases, the initial users who offered arguments did not also post statements that provided a merger or synthesis of two opposing arguments in order to move forward with a discussion within those same comment strings. It is also possible that the users did not read other users’ posts prior to posting their own comments (Kim, Anderson, Nguyen-Jahiel, & Archodidou, 2007), and thus may post arguments that are only connected to the article that they are posting a response to. Importantly, these findings confirmed our hypotheses that theory regarding the nature of argument skills within the formal environment may also be applied to informal environments, thereby extending and building a new theoretical basis regarding the use and utility of argument skills. Situating opportunities for argumentation within issue-oriented niche social networking applications, may complement and extend efforts to develop young people’s scientific literacy in formal learning environments, while helping to

Please cite this article in press as: Greenhow, C., et al. Re-thinking scientific literacy out-of-school: Arguing science issues in a niche Facebook application. Computers in Human Behavior (2015), http://dx.doi.org/10.1016/j.chb.2015.06.031

8

C. Greenhow et al. / Computers in Human Behavior xxx (2015) xxx–xxx

Table 3 Exploratory factor analysis.

Factors Eigenvalue Percent of Variance Explained All Skills Counterarguments Other Points of View Claims GWB Qualifying Statements Integrated Replies Complexity Inquiry Skepticism Conflict-Oriented CB Integration-Oriented CB Quick CB

Counterarguments

Arguments

1 1.69 28.08

2 1.51 25.15

Integrated Replies 3 1.01 16.83

Epistemic Skills 4 1.35 44.94

Conflict-Oriented Consensus Building 5 1.32 43.99

Quick Consensus Building 6 1.16 38.50

.88 .82 .78 .73 .59 .98 .76 .64 .61 .82 -.80 .93

Note: Factor analysis with Varimax rotation and Eigenvalues above one were retained in the final models. Factors 1, 2, and 3 fall within Argument Skills. Factor 4 falls within Epistemic Skills. And Factor 5 and 6 falls within Social Co-Construction Skills. GWB = Grounds, Warrants, and Backings. CB = Consensus Building.

overcome their constraints.

hierarchical,

structural

and

time-bound

3.3. Epistemic skills As expected, we also saw evidence of users demonstrating a variety of epistemic skills (Sadler et al., 2006), most of which were with regard to using complexity. Exploratory factor analysis revealed that a one-factor model fit (see Table 3 for factor loadings). Results from the exploratory factory analysis, specifying Varimax rotation and eigenvalues above one, provided empirical support suggesting that the Sadler et al. (2006) epistemic dimension could not be further divided into subsets of skills and have extended this finding to informal learning contexts, such as Hot Dish. Our findings confirm Sadler and colleagues’ notion that complexity and inquiry are highly and positively related, and also that skepticism is a critical component of argumentation. It appears likely that as participants register the complexity of the issue under discussion, their inquiry and skepticism, or critical evaluation of the information in claims and evidence are also registered. In fact, we hoped to see students’ attention to complexity and a healthy dose of skepticism in their online commentary about socio-scientific issues. 3.4. Social co-construction skills Thus far, two of our hypotheses were confirmed. Turning to social co-construction, we also saw a high rate of social co-construction skills (Weinberger & Fischer, 2006), with the three most used skills being externalization, integration-oriented consensus building, and conflict-oriented consensus building. As expected, given the number of different articles that were posted in Hot Dish, externalization, or stating one’s thinking with no connection to what other participants have expressed, was used by the majority of participants. In fact, the number of externalizations surpassed the number of unique articles posted. The use of externalization in social co-construction skill could be related to when people state their opinion (Leitão, 2000; Weinberger & Fischer, 2006). For example, there may be qualitative differences in posts expressed early on in the comment string of a specific article compared to those posts later in the comment strings for the same article (Weinberger & Fischer, 2006). That is, there is less room to co-construct knowledge when a user is the first person to respond or comment on a news story. As mentioned previously,

there is also the potential that users comment without reading others’ comments first (Kim et al., 2007). These notions of temporal parameters and individual investment in the comments of other users might also explain the high prevalence of arguments, whereby an opening comment string to an article might include both an externalization and an argument to foster a potential discussion. Weinberger and Fischer (2006) argued that conflict-oriented consensus building (i.e., using disagreements or modifications of other learners’ perspectives or arguments) and integration -oriented consensus building (i.e., being persuaded to change ones position based on the persuasive arguments of others) are the two most complex or sophisticated forms of social co-construction. The results of the current study suggest that users who engaged in posting and responding to comments in Hot Dish were using highly advanced co-construction skills, both in the form of conflict-oriented consensus building and integration-oriented consensus building. Importantly, Weinberger (2003) found little evidence of integration-oriented consensus building within formal CSCL environments, whereas we saw the opposite effect within the informal learning environment. This inconsistency in the literature suggests that informal learning environments may provide some users a unique opportunity to use more integration -oriented consensus building. Perhaps the high prevalence of integration-oriented consensus building resulted from the influence of peers within the informal online environment. Some scholars have speculated that peer influence is just as strong or potentially stronger for learning and knowledge building in the online community than in the offline community (Kim et al., 2007). Other scholars have noted that asynchronous interactions, which were present in this SNS, are important for learners because it provides ample opportunities for sharing knowledge, skills, and perspectives (Bonk, Hansen, Grabner-Hagen, Lazon, & Mirabelli, 1998; Heo, Lim, & Kim, 2010; Lapadat, 2002), especially in social media spaces (Mason, 2006). Moreover, several studies have documented the affordances of social network sites for helping young people increase and maintain social connections over time. For instance, longer Facebook use, but not use of other social media, has been linked to higher reported levels of social connectedness in adolescents (12–18) (Alloway, Horton, Alloway, & Dawson, 2013). In several studies, intense social networking has been associated with an increase in social capital among adolescent and undergraduate users, both bonding capital (close ties) and bridging capital (loose ties associated with access to resources and

Please cite this article in press as: Greenhow, C., et al. Re-thinking scientific literacy out-of-school: Arguing science issues in a niche Facebook application. Computers in Human Behavior (2015), http://dx.doi.org/10.1016/j.chb.2015.06.031

9

C. Greenhow et al. / Computers in Human Behavior xxx (2015) xxx–xxx

information) (Ellison, 2009; Greenhow, Burton, & Robelia, 2011; Valenzuela, Park, & Kee, 2009). Thus, it may be in implementing a Facebook application within these young people’s existing Facebook-ing routines, there is an existing context for peer influence, resource-sharing, and connectedness that is more conducive to fostering Hot Dish users’ self-correction based on peers’ contributions than has been found in formal CSCL environments. In addition, all social co-construction skills except externalization and elicitation were entered into an exploratory factor analysis, specifying Varimax rotation and eigenvalues above one, which yielded a two-factor model. The Conflict-Oriented Consensus Building Factor (Factor 5) comprised Conflict-Oriented Consensus Building and Integration-Oriented Consensus Building codes. Importantly, Integration-Oriented Consensus Building had a negative factor loading, indicating that the presence of Conflict-Oriented Consensus Building and Integration-Oriented Consensus Building behaviors were mutually exclusive. In other words, the likelihood of coding a comment string with both types of consensus building was minimal. The Quick Consensus Building Factor (Factor 6) comprised only of the Quick Consensus Building behavior. Thus, a score for the Conflict-Oriented Consensus Building Factor was calculated by subtracting the Integration-Oriented Consensus Building behavior from the Conflict Oriented Consensus Building score, due to the negative loading for the Integration-Oriented Consensus Building behavior (DiStefano, Zhu, & Mindrila, 2009). Externalization was purposely excluded from the final exploratory factor analysis since Externalization generally refers to a behavior that starts a conversation rather than maintaining it or furthering it (Weinberger & Fischer, 2006). Furthermore, when included in the exploratory factor analysis, it highly, but negatively, cross-loaded on three separate factors; scores ranged from .79 to .33, which violates the assumptions for acceptable loading scores (Tabachnick & Fidell, 2007). Elicitation was also purposely excluded from the exploratory factor analysis since it was theoretically argued to be both related and unrelated to co-construction depending on contextual circumstances or parameters (Webb, 1989). 3.5. Relations among argumentation skills Beyond examining the nature and extent to which certain argumentation skills were evident in the comments of Hot Dish users, the second focus of this study was to examine the relations between the factors resulting from the exploratory factor analysis to further understand SSI argumentation in informal learning contexts and whether these associations were similar to those seen in formal learning contexts. According to theory (Sadler et al., 2006; Weinberger & Fischer, 2006), argument skills, epistemic skills, and social co-construction skills were generally expected to be positively related to each other. Correlational results from the current study supported these notions, though there was also some evidence of negative relations as well as nonsignificant relations (see Table 4). First, we found that both arguments and counterarguments were positively related to complexity. These findings are in line with scholars that argue that significant components of epistemic skills are being able to use multiple sources and perspectives in order to engage in discussions (i.e., complicating the issue), questioning the source of information or arguments, and gathering the appropriate sources of information in order to provide additional information to the discussion (Andriessen et al., 2003; Sadler et al., 2006; Weinberger & Fischer, 2006). These behaviors are all demonstrated in arguments and counterarguments, where users draw on a variety of perspectives and information to develop their claims.

Table 4 Bivariate correlations between argumentation factors. 1 1. 2. 3. 4. 5.

Counterarguments Arguments Integrated Replies Epistemic Skills Conflict-Oriented Consensus Building 6. Quick Consensus Building

–

2

3

4 ***

.062 –

.103 .131*** –

5 **

.214 .304*** .030 –

6 ***

.404 .044 .200*** .164*** –

.117*** .272*** .027 .375*** .018 –

Note: N = 971 comment strings. ⁄ p > .05. ** p > .01. *** p > .001.

Second, we found that counterarguments were positively related to conflict-oriented consensus building, whereas both arguments and counterarguments were negatively related to quick consensus building. In the former case, in order to engage in debate and subsequent clash of perspectives or ideas, there must be the demonstration of a counterargument or opposing force (Laursen & Pursell, 2009). A high concentration of arguments and counterarguments further the likelihood of conflicting values and attitudes to occur or continue to occur. Thus, the positive link between counterarguments and conflict-oriented consensus building would be expected. In the latter case, perhaps it may be that presenting arguments and counterarguments promote conflict and discourse, whereas quick consensus building prevents, ameliorates, and resolves conflict, but also prevents more sophisticated discussion (Chinn & Brewer, 1993; Weinberger & Fischer, 2006). In fact, a criticism of engaging argumentation in formal CSCL environments is that the argumentation is often poor, dismissing contradictions and inconsistencies (Kanuka & Anderson, 1998). Also, consider that engaging in conflict involves providing multiple perspectives, which ties the epistemic dimension of complexity to social co-construction skills (Gijlers, Saab, Van Joolingen, De Jong, & Van Hout-Wolters, 2009). Indeed, by definition, complexity refers to using and interpreting multiple perspectives on a similar issue in order to understand the issue (Sadler et al., 2006). In this regard, having multiple users argue and counter-argue (i.e., engaging or soliciting conflict-oriented consensus building) allows users to learn the opinions and positions of each other and take those pieces of information into account, regardless of if they conflict with or confirm their own views (Weinberger & Fischer, 2006). Using arguments and counterarguments also allow users to use information presented by other users in order to further their own position or agenda; and the sharing of such information increases the complexity of the issue and promotes learning (Doise, Mugny, Pérez, & Duveen, 1998; Gijlers et al., 2009; Nastasi & Clements, 1992). Aligned with the notion that discussion that encourages further discourse is associated with sophisticated social co-construction, is the notion that discussion which discourages further discussion or exploration of ideas (quick consensus building) is associated with low complexity (Weinberger & Fischer, 2006). 4. Conclusions It is evident from the results that informal, non-school, social media environments like Hot Dish can provide opportunities for young people to engage in debate about socio-scientific issues, which in turn, serve as powerful facilitators for developing learners’ contemporary scientific literacy. Arguments, counterarguments, integrated replies, epistemic skills, conflict-oriented consensus building, and quick consensus building, all components of argumentation competence, were evident.

Please cite this article in press as: Greenhow, C., et al. Re-thinking scientific literacy out-of-school: Arguing science issues in a niche Facebook application. Computers in Human Behavior (2015), http://dx.doi.org/10.1016/j.chb.2015.06.031

10

C. Greenhow et al. / Computers in Human Behavior xxx (2015) xxx–xxx

Most of the work on argumentation has occurred in formal settings (Weinberger & Fischer, 2006); our work extends the current research and theory regarding the relations of argumentation within informal learning environments. Importantly, there may be unique characteristics that differentiate informal from formal learning environments; and the findings reported in this paper suggest that those unique characteristics may be influential in the nature of and the extent to which users employ argumentations skills. The implications of these findings can further research on the advantages and disadvantages of nontraditional learning (Lave & Wenger, 1991), specifically within SNS contexts. We next discuss strengths and weaknesses of the current study, as well as address the implications of these findings for the design and study of similar social networking applications. 4.1. Strengths, limitations, and future directions One of the strengths of the current study was that it provided support for the notion that sophisticated learning can and does occur within informal learning environments (Lave & Wenger, 1991), specifically those powered by social media. Furthermore, this information can inform how to potentially integrate these niche networks into young people’s overall learning ecology either as complements to formal learning environments (e.g., science classes), or as informal complements to online courses. This integration would foster comparative studies of informal and formal learning environments to address qualitative differences in argumentation as well as whether the link between argumentation and learning is stronger in one context versus another. The current study also provided a snapshot of online social networking practices over a very short timeframe; furthermore, posts that occurred early in the project were not distinguished from posts that occurred later in the project. Thus, little is known about whether online social networking practices vary and change over time within the Hot Dish application’s study period as well as beyond it. In studies with the next generation of Hot Dish or other niche Facebook applications we design, we might also look at students’ development over time or changes in argumentation over time. For example, some work has examined argumentation from a sequential perspective (Leitão, 2000; Weinberger & Fischer, 2006), and next steps might be to further distinguish whether argument skills, epistemic skills, and social co-construction skills vary depending when posts are posted and in what order they are posted. Examining how people’s practices within the social networking application change over time – perhaps how early engagement might look different from later engagement—and mapping this to the technical features people use and their social practices, could help us iterate both the application’s design and our evolving theories of use. A second limitation of our study is our focus on the small percentage of participants from the larger sample who chose to comment. This self-selection bias, a relatively common occurrence in online communities (e.g., the majority of the content is generated by a minority of the participants), suggests our results should be interpreted with caution. It may be that such social media applications will be most effective with a core group of highly interested, niche users rather than broadly applicable to all learners. Our analysis may have yielded different results if we had expanded it to include ‘low’ users or those who posted fewer than three comments on any given story. From this investigation, we suggest a number of additional fruitful areas for research. For instance, future research should attend to individual differences among the students (e.g., gender/sex, age, use of certain skills with certain people versus others, are certain participants engaging in discourse with only certain others, varying levels of engagement) and what distinguishes highly active

engagement with the application from those with medium to low to no activity. The results from the current study also suggest that the prevalence of integration-oriented consensus building may be uniquely tied to the informal learning environment, such as an online Facebook.com application. Thus, it would be important to examine this particular consensus-building skill and other consensus-building skills across a variety of informal and formal learning environments to further understand the function and outcome of integration-oriented consensus building as well as which socio-technical features unique to the social networking environment may bring this about. With respect to individual differences, it might also be important to examine issues such as shyness or developmental disabilities in learning settings (Kim et al., 2007). Scholars have noted that CSCL learning environments are particularly beneficial for youth who are socially anxious or who have deficits or disabilities in speech production (e.g., a lisp or stutter). These non-face-to-face interactions that are purely focused on text communication provide SSI negotiation practice opportunities for these youth. Scholars have documented that in face-to-face environments, individual differences can influence the way youth act with their peers (Rubin, Bukowski, & Parker, 2006), yet much is still unknown regarding peer relationships in online settings such as social network sites. In this study, exploratory factor analysis was used to reveal patterns of correlations among observed argumentation behaviors that were thought to reflect underlying or latent processes within specific dimensions of argumentation. A follow up study may be to use confirmatory factory analysis to further evaluate our ideas in this paper about the different types of argument skills and the relationships between them. Furthermore, the findings of the current study also have development and design implications. For example, only 31 out of 346 users participated in commenting on stories, and the results of the current study only pertain to those 31 users. Perhaps, future renditions of a Hot Dish-like application should take into consideration additional methods or techniques that encourage youth to participate in reading and posting responses to stories. For instance, modifications in the design of the next version of Hot Dish might include expanding the number of articles that can be displayed simultaneously on the screen and incentivizing a the contribution of a wider variety of article genres to post responses to, both on the topic of environmental issues as well as off topic. Other modifications might be to provide additional recognition for posting comments such as indication of expertise within the community or recommendations for one’s expertise, written by Hot Dish members. Taken together, the current findings provide important contributions to our understanding of how young people use argumentation in informal learning environments like Hot Dish toward developing contemporary scientific literacy. This work builds upon and advances recent efforts to document and explore aspects of learning outside of school in youth-initiated online spaces (Peppler & Kafai, 2007; Steinkuehler, 2008) as potential avenues for bridging formal and informal learning experiences.

Ethics in publishing None violated.

Conflict of interest There are no conflicts of interest.

Please cite this article in press as: Greenhow, C., et al. Re-thinking scientific literacy out-of-school: Arguing science issues in a niche Facebook application. Computers in Human Behavior (2015), http://dx.doi.org/10.1016/j.chb.2015.06.031

C. Greenhow et al. / Computers in Human Behavior xxx (2015) xxx–xxx

Acknowledgement This work was supported with a grant from the John S. and James L. Knight Foundation.

References Alloway, T. P., Horton, J., Alloway, R. G., & Dawson, C. (2013). Social networking sites and cognitive abilities: Do they make you smarter? Computers & Education. Andriessen, J. E. B., Baker, M., & Suthers, D. (Eds.). (2003). Arguing to learn. Confronting cognitions in computer-supported collaborative learning environments. Dordrecht: Kluwer. Asterhan, C. S. C., & Hever, R. (in Press). Learning from reading argumentive discussions in Facebook: Rhetoric style matters (again). Computers in Human Behavior. Barab, S. A., Sadler, T., Heiselt, C., Hickey, D., & Zuiker, S. (2007). Relating narrative, inquiry, and inscriptions: A framework for socio-scientific inquiry. Journal of Science Education and Technology, 16, 59–82. Beach, R., & Doerr-Stevens, C. (2011). Using social networking for online role-plays to develop students’ argumentative strategies. Journal of Educational Computing Research, 45(2), 165–181. Berland, L. K., & Hammer, D. (2012). Framing for scientific argumentation. Journal of Research in Science Teaching, 49, 68–94. Bonk, C. J., Hansen, E. J., Grabner-Hagen, M. M., Lazon, S. A., & Mirabelli, C. (1998). Time to ‘connect’: Synchronous and asynchronous case-based dialogue among preservice teachers. In C. J. Bonk & K. S. King (Eds.), Electronic collaborators: Learner-centred technologies for literacy, apprenticeship and discourse (pp. 289–314). Mahwah, NJ: Lawrence Erlbaum Associates. Brown, J. S., Collins, A., & Duguid, P. (1989). Situated cognition and the culture of learning. Educational Researcher, 18, 34–41. Bureau of Labor Statistics (2007). The 30 fastest growing occupations covered in the 2008–2009 Occupational Outlook Handbook. Cavagnetto, A. R. (2010). Argument to foster scientific literacy. A review of argument interventions in K-12 science contexts. Review of Educational Research, 80(3), 336–371. Chan, C., Burtis, J., & Bereiter, C. (1997). Knowledge building as a mediator of conflict in conceptual change. Cognition and Instruction, 15, 1–40. Chin, C., & Osborne, J. (2010). Supporting argumentation through students’ questions: Case studies in science classrooms. Journal of the Learning Sciences, 19, 230–284. Chinn, C. A., & Brewer, W. F. (1993). The role of anomalous data in knowledge acquisition: A theoretical framework and implications for science instruction. Review of Educational Research, 63, 1–49. Clark, H. H., & Brennan, S. A. (1991). Grounding in communication. In L. B. Resnick, J. M. Levine, & S. D. Teasley (Eds.), Perspectives on socially shared cognition. Washington: APA Books. Clark, D. B., & Sampson, V. D. (2008). Assessing dialogic argumentation in online environments to relate structure, grounds and conceptual quality. Journal of Research in Science Teaching, 45, 293–321. Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20, 37–46. Cohen, E. G., & Lotan, R. A. (1995). Producing equal-status interaction in the heterogeneous classroom. American Educational Research Journal, 32, 99–120. Collins, A., & Halverson, B. (2009). Rethinking education in the age of technology. New York: Teachers College Press. Davis, E. A. (2003). Prompting middle school science students for reflection: Generic and directed prompts. Journal of the Learning Sciences, 12, 91–142. Davis, E.A. (2004). Creating critique projects. 215–229. DiStefano, C., Zhu, M., & Mindrila, D. (2009). Understanding and using factor scores: Considerations for the applied researcher. Practical Assessment, Research & Evaluation, 14. Doise, W., & Mugny, G. (1984). The social development of the intellect. Oxford: Pergamon. Doise, W., Mugny, G., Pérez, J. A., & Duveen, G. (1998). The social construction of knowledge: Social marking and socio-cognitive conflict. In U. Flick (Ed.), The psychology of the social (pp. 77–90). New York, NY, US: Cambridge University Press. Ellison, N. B., & Boyd, D. (2014). Sociality through social network sites. In W. H. Dutton (Ed.), The Oxford handbook of internet studies. Oxford: Oxford University Press. Ellison, et al. (2009) – social capital in Facebook study. Fischer, F., Bruhn, J., Gräsel, C., & Mandl, H. (2002). Fostering collaborative knowledge construction with visualization tools. Learning and Instruction, 12, 213–232. Furberg, A. A. (2009). Socio-cultural aspects of prompting student reflection in webbased inquiry learning environments. Journal of Computer Assisted Learning, 25, 397–409. Gibbins, T., & Greenhow, C. (2014). Social networking applications as third spaces for writing development. In R. Anderson & C. Mims (Eds.), Digital tools for writing instruction in K-12 settings: Student perception and experience. New York, NY: Springer. Gijlers, H. H., Saab, N. N., Van Joolingen, W. R., De Jong, T. T., & Van Hout-Wolters, B. M. (2009). Interaction between tool and talk: How instruction and tools support

11

consensus building in collaborative inquiry-learning environments. Journal of Computer Assisted Learning, 25, 252–267. Greenhow, C. (2010). The role of youth as cultural producers in a niche social network site. New Directions in Youth Development: Theory, Research & Practice, 128, 55–64. Greenhow, C. (2011a). Online social networks and learning. On the Horizon, 15(1), 4–12. Greenhow, C. (2011b). Learning and social media: What are the interesting questions for research? International Journal of Cyber Behavior, Psychology and Learning, 1(1), 36–50. Greenhow, C., Burton, L., & Robelia, B. (2011). Help from my ‘‘Friends:’’ Social capital in the social network sites of low-income high school students. Journal of Educational Computing Research, 45(2), 223–245. Greenhow, C., & Robelia, E. (2009a). Old communication, new literacies: Social network sites as social learning resources. Journal of Computer-mediated Communication, 14, 1130–1161. Greenhow, C., & Robelia, E. (2009b). Informal learning and identity formation in online social networks. Learning, Media and Technology, 34(2), 119–140. Greenhow, C., Robelia, E., & Hughes, J. (2009). Web 2.0 and classroom research: What path should we take now? Educational Researcher, 38(4), 246–259. Greeno, J. (1989). The situativity of knowing, learning, and research. American Psychologist, 53, 5–26. Gutwill, J. P., & Allen, S. (2012). Deepening students’ scientific inquiry skills during a science museum field trip. Journal of the Learning Sciences, 21, 130–181. Heimlich, J. E., & Ardoin, N. M. (2008). Understanding behavior to understand behavior change: A literature review. Environmental Education Research, 14, 215–237. Heo, H., Lim, K., & Kim, Y. (2010). Exploratory study on the patterns of online interaction and knowledge co-construction in project-based learning. Computers & Education, 55, 1383–1392. Hogan, K., Nastasi, B. K., & Pressley, M. (1999). Discourse patterns and collaborative scientific reasoning in peer and teacher-guided discussions. Cognition & Instruction, 17, 379–432. Junco, R. (2012a). Too much face and not enough books: The relationship between multiple indices of Facebook use and academic performance. Computers in Human Behavior, 28(1), 187–198. Junco, R. (2012b). The relationship between frequency of Facebook use, participation in Facebook activities, and student engagement. Computers & Education, 58(1), 162–171. Junco, R. (2013). Inequalities in Facebook use. Computers in Human Behavior, 29(6), 2328–2336. Kanuka, H., & Anderson, T. (1998). Online social interchange, discord, and knowledge construction. Journal of Distance Education, 13(1), 1–19. Kelly, D., Xie, H., Nord, C.W., Jenkins, F., Chan, J.Y., & Kastberg, D. (2013). Performance of U.S. 15-Year-Old Students in Mathematics, Science, and Reading Literacy in an International Context: First Look at PISA 2012 (NCES 2014-024). U.S. Department of Education. Washington, DC: National Center for Education Statistics. Kim, I., Anderson, R. C., Nguyen-Jahiel, K., & Archodidou, A. (2007). Discourse patterns during children’s collaborative online discussions. Journal of the Learning Sciences, 16, 333–370. Kol, S., & Schcolnik, M. (2008). Asynchronous forums in EAP: Assessment issues. Language Learning & Technology, 12, 49–70. Kolsto, S. (2000). Consensus projects: Teaching science for citizenship. International Journal of Science Education, 22, 645–664. Lapadat, J. C. (2002). Written interaction: A key component in online learning. Journal of Computer-Mediated Communication, 7. Laursen, B., & Pursell, G. (2009). Conflict in peer relationships. In K. H. Rubin, W. M. Bukowski, & B. Laursen (Eds.), Handbook of peer interactions, relationships, and groups (pp. 267–286). New York, NY, US: Guilford Press. Lave, J., & Wenger, E. (1991). Situated learning: Legitimate peripheral participation. New York: Cambridge University Press. Leitão, S. (2000). The potential of argument in knowledge building. Human Development, 43, 332–360. Lenhart, A., Purcell, K., Smith, A., & Zickuhr, K. (2010). Social media and young adults. Washington, DC: Pew Charitable Trusts. Levin, D., & Arafeh, S. (2002). The digital disconnect: The widening gap between internet-savvy students and their schools. Washington, DC: Pew Charitable Trusts. Manca, S., & Ranieri, M. (2013). Is it a tool suitable for learning? A critical review of the literature on Facebook as a technology-enhanced learning environment. Journal of Computer Assisted Learning, 29(6), 487–504. Mason, R. (2006). Learning technologies for adult continuing education. Studies in Continuing Education, 28, 121–133. Mazer, J. P., Murphy, R. E., & Simonds, C. J. (2007). I’ll see you on ‘‘Facebook:’’ The effects of computer-mediated teacher self-disclosure on student motivation, affective learning, and classroom climate. Communication Education, 56, 1–17. Nastasi, B. K., & Clements, D. H. (1992). Social-cognitive behaviors and higher-order thinking in educational computer environments. Learning and Instruction, 2, 215–238. National Center for Education Statistics (NCES) (2005). The condition of education in brief. NCES 2005095. Washington DC. Author. National Science Board (2010). Science and engineering indicators: 2010. Arlington, VA: National Science Foundation. Nussbaum, E., & Edwards, O. V. (2011). Critical questions and argument stratagems: A framework for enhancing and analyzing students’ reasoning practices. Journal of the Learning Sciences, 20, 443–488.

Please cite this article in press as: Greenhow, C., et al. Re-thinking scientific literacy out-of-school: Arguing science issues in a niche Facebook application. Computers in Human Behavior (2015), http://dx.doi.org/10.1016/j.chb.2015.06.031

12

C. Greenhow et al. / Computers in Human Behavior xxx (2015) xxx–xxx

O’Mahony, T. K., Vye, N. J., Bransford, J. D., Sanders, E. A., Stevens, R., Stephens, R. D., et al. (2012). A comparison of lecture-based and challenge-based learning in a workplace setting: Course designs, patterns of interactivity, and learning outcomes. Journal of the Learning Sciences, 21, 182–206. Palmquist, S., & Crowley, K. (2007). From teachers to testers: How parents talk to novice and expert children in a natural history museum. Science Education, 91, 783–804. Peppler, K. A., & Kafai, Y. B. (2007). From supergoo to scratch: Exploring creative digital media production in informal learning. Learning, Media and Technology, 32, 149–166. Pew Research Center for the People & the Press (2008). Key news audiences now blend online and traditional sources. Washington DC: Pew Charitable Trusts. Polman, J. L., Saul, W., Newman, A., Farrar, C., Singer, N., Turley, E., Pearce, L., Hope, J., & McCarty, G. (2010). A cognitive apprenticeship for science literacy based on journalism. In K. Gomez, Lyons, L., & Radinsky, J. (Ed.), Learning in the disciplines: Proceedings of the 9th international conference of the learning sciences (ICLS 2010) – Vol. 2, short papers, symposia, and selected abstracts (pp. 61–68). Chicago, IL: International Society of the Learning Sciences. Putnam, R. D. (1995). Bowling alone: The collapse and revival of the American community. New York, NY: Simon & Schuster. Ranieri, M., Manca, S., & Fini, A. (2012). Why (and how) do teachers engage in social networks? An exploratory study of professional use of Facebook and its implications for lifelong learning. British Journal of Educational Technology, 43(5), 754–769. Renkl, A. A. (2002). Learning from worked-out examples: Instructional explanations supplement self-explanations. Learning & Instruction, 12, 529–556. Rideout, V.J., Foehr, U.G., & Roberts, D.F. (2010, January 20). Generation M2: Media in the lives of 8- to 18-year-olds (Report No. 8010). Menlo Park, CA: Kaiser Family Foundation. Robelia, B., Greenhow, C., & Burton, L. (2011). Adopting environmentally responsible behaviors: How learning within a social networking application motivated students to act for the environment. Environmental Education Research, 17(4), 553–575. Rubin, K. H., Bukowski, W. M., & Parker, J. G. (2006). Peer interactions, relationships, and groups. In N. Eisenberg, W. Damon, & R. M. Lerner (Eds.), Handbook of child psychology: Vol. 3, social, emotional, and personality development (6th ed., pp. 571–645). Hoboken, NJ, US: John Wiley & Sons Inc. Sadler, T., Barab, S., & Scott, B. (2006). What do students gain by engaging in socioscientific inquiry. Research in Science Education, 37, 371–391. Sadler, T. D., & Fowler, S. R. (2006). A threshold model of content knowledge transfer for socioscientific argumentation. Science Education, 90, 986–1004. Salomon, G., & Perkins, D. N. (1998). Individual and social aspects of learning. Review of Research in Education, 23, 1–24.

Sefton-Green, J. (2004). Report 7: Literature review in informal learning with technology outside school. Bristol, England: Futurelab. Spiro, R. J., Feltovich, P. J., Jacobson, M. J., & Coulson, R. L. (1991). Cognitive flexibility, constructivism, and hypertext: Random access instruction for advanced knowledge acquisition in ill-structured domains. Educational Technology, 31, 24–33. Stegmann, K., Wecker, C., Weinberger, A., & Fischer, F. (2012). Collaborative argumentation and cognitive elaboration in a computer-supported collaborative learning environment. Instructional Science, 40, 297–323. Steinkuehler, C. (2008). Cognition and literacy in massively multiplayer online games. In J. Coiro, M. Knobel, C. Lankshear, & D. Leu (Eds.), Handbook of research on new literacies (pp. 611–613). New York: Lawrence Erlbaum. Tabachnick, B., & Fidell, L. (2007). Using multivariate statistics (5th ed.). New York, NY: Allyn and Bacon. Teasley, S. (1997). Talking about reasoning: How important is the peer in peer collaboration? In L. B. Resnick, R. Säljö, C. Pontecorvo, & B. Burge (Eds.), Discourse, tools and reasoning: Essays on situated cognition (pp. 361–384). Berlin: Springer. Toulmin, S. (1958). The uses of argument. New York, NY: Cambridge University Press. Tsovaltzi, D., Puhl, T., Judele, R., & Weinberger, A. (2014). Group awareness support and argumentation scripts for individual preparation of arguments in Facebook. Computers & Education, 76, 108–118. Valenzuela, S., Park, N., & Kee, K. F. (2009). Is there social capital in a social network site?: Facebook use and college students’ life satisfaction, trust, and participation. Journal of Computer-mediated Communication, 14, 875–901. Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University Press. Webb, N. M. (1989). Peer interaction and learning in small groups. International Journal of Educational Research, 13, 21–39. Weinberger, A., & Fischer, F. (2006). A framework to analyze argumentative knowledge construction in computer-supported collaborative learning. Computers & Education, 46, 71–95. Wenger, E. (1998). Communities of practice: Learning, meaning, and identity. New York, NY: Cambridge University Press. Yeh, K., & She, H. (2010). On-line synchronous scientific argumentation learning: Nurturing students’ argumentation ability and conceptual change in science context. Computers & Education, 55, 586–602. Yoon, S. (2011). Using social network graphs as visualization tools to influence peer selection decision-making strategies to access information about socioscientific issues. Journal of the Learning Sciences, 20, 549–588. Zeidler, D. L., Walker, K. A., Ackett, W. A., & Simmons, M. L. (2002). Tangled up in views: Beliefs in the nature of science and responses to socioscientific dilemmas. Science Education, 86, 343–367.

Please cite this article in press as: Greenhow, C., et al. Re-thinking scientific literacy out-of-school: Arguing science issues in a niche Facebook application. Computers in Human Behavior (2015), http://dx.doi.org/10.1016/j.chb.2015.06.031