A Copernican Turn for the Development of Flexibly Reusable Scaffolds and Collaboration Scripts: the S-col Approach

Collaboration scripts are usually implemented as parts of a particular collaborative-learning platform. Therefore, scripts of demonstrated effectiveness are hardly used with learning platforms at other sites, and replication studies are rare. The approach of a platformindependent description language for scripts that allows for easy implementation of the same script on different platforms has not succeeded yet in making the transfer of scripts feasible. We present an alternative solution that treats the problem as a special case of providing support on top of diverse Web pages: In this case, the challenge is to trigger support based on the recognition of a Web page as belonging to a specific type of functionally equivalent pages such as the search query form or the results page of a search engine. The solution suggested has been implemented by means of a tool called S-COL (Scripting for Collaborative Online Learning) and allows for the sustainable development of scripts and scaffolds that can be used with a broad variety of content and platforms. The tool’s functions are described. In order to demonstrate the feasibility and ease of script reuse with S-COL, we describe the flexible reimplementation of a collaboration script for argumentation in S-COL and its adaptation to different learning platforms. To demonstrate that a collaboration script implemented in SCOL can actually foster learning, an empirical study about the effects of a specific script for collaborative online search on learning activities is presented. The further potentials and the limitations of the S-COL approach are discussed. ha l-0 07 03 89 6, v er si on 1 4 Ju n 20 12 S-COL: Copernican Turn for script development 4 Changing the approach to the development of flexibly reusable collaboration scripts Research on technology-based collaboration scripts has been very successful in terms of the development of a broad range of scripts that effectively foster activities and outcomes of computer-supported collaborative learning (e.g., Baker & Lund, 1997; De Wever, Van Keer, Schellens, & Valcke, 2009; Kollar, Fischer & Slotta, 2007; Rummel, & Spada, 2005; Schellens, Van Keer, De Wever, & Valcke, 2007; Schoonenboom, 2008; Slof, Erkens, Kirschner, Jaspers & Janssen, 2010; Stegmann, Weinberger & Fischer, 2007; Weinberger, Ertl, Fischer & Mandl, 2005; Weinberger, Stegmann & Fischer, 2010). The growing importance of this field of research is evidenced by—among other things—the announcement of “Scripting in CSCL” as a “flash theme” in the International Journal of ComputerSupported Collaborative Learning (Stahl & Hesse, 2007). However, technology-based collaboration scripts are usually developed exclusively for one specific, often experimental, learning platform. Neither the transfer to other experimental platforms nor the transfer into practice has been managed systematically so far. Among the current approaches to overcome these problems, the most prominent one is the attempt to develop a universal formal language (e.g., an extension of IMS-LD) for the specification of scripts to be “read in” and implemented by different collaborative-learning platforms (Weinberger et al., 2007). In this article, we propose a different and by far simpler solution: Instead of trying to get different platforms to display functionally equivalent but platform-specific versions of the “same” script, we suggest using a pre-implemented script that is embedded in the learner’s Web browser. This requires that specific components of the script be invoked whenever the browser recognizes pages displayed by a learning platform as being of the corresponding types of functionally equivalent pages. We call this the S-COL (Scripting for Collaborative Online Learning) approach to the development of flexibly reusable collaboration scripts for diverse Web content. Because of the shift mentioned, the S-COL approach can be regarded as ha l-0 07 03 89 6, v er si on 1 4 Ju n 20 12 S-COL: Copernican Turn for script development 5 a kind of two-fold “Copernican Turn” in script development: First, instead of the learning platform, the browser of the learner is moved into the centre of script development by making it the source of the support displayed to the learners. Second, the burden of creating flexibility is shifted from the idea of a universal formal description of a collaboration script to be generated by any learning platform to the task of triggering the appropriate components from a pre-implemented script. We need to clarify right from the start that our claim in this article is not that the S-COL approach leads to superior learning compared to other approaches to the implementation of collaboration scripts or unstructured collaboration. S-COL simply provides a technical frame for the implementation of collaboration scripts. Accordingly, it can be used to implement a broad variety of collaboration scripts, including ineffective and even detrimental ones. What we do claim, however, is that diverse types of collaboration scripts can be implemented in SCOL with no more effort than implementing a script as part of a specific learning platform, yet with the advantage of flexible reusability within different learning platforms. This is not a claim about the psychology or instructional design of computer-supported collaborative learning. It is a claim about the power and generality of a framework for the implementation of support for computer-supported collaborative learning, which we think advances an ongoing discussion in this journal and in the CSCL community (e.g., Dillenbourg, & Tchounikine, 2007; Harrer & Malzahn, 2006; Kobbe et al., 2007; Miao, Harrer, Hoeksema, & Hoppe, 2007; Stegmann et al., 2009; Tchounikine, 2008; Weinberger et al., 2007). The focus of our claim has consequences for the evidence required to support this claim, which we will elaborate shortly. The genesis of our approach provides some further insights into a more general problem of which the development of flexibly reusable scripts may be regarded as a special case, and into further applications of the approach: Interestingly, we hit on the solution described above when we were working on an apparently quite unrelated problem. We were ha l-0 07 03 89 6, v er si on 1 4 Ju n 20 12 S-COL: Copernican Turn for script development 6 looking for a way to provide context-specific support on top of varying Web pages. At some point, we came to view the problem of developing collaboration scripts for flexible reuse with different learning platforms as a special case of developing context-specific support on top of varying Web pages: From this perspective, different Web-based learning platforms are regarded simply as varying Web pages. Therefore, in order to illustrate the general idea, in the following section, we describe both our initial general problem (of providing support on top of varying Web pages) and the more specific problem (of developing scripts for flexible reuse with different learning platforms), including a review of current attempts at solutions. The third section provides a characterization of the general idea behind the comprehensive solution for the problems on both levels. Based on the insights gained, a tool was designed to solve the general problem of support on top of Web pages, and, hence, also the more specific problem of developing scripts for flexible reuse with different learning platforms. In the fourth section, we describe the main features of the S-COL tool, in particular, its graphical user interface, its functions to provide support for learning and its administration features. The fifth section provides a short description of the technical implementation of the S-COL tool directed at a more technically oriented audience. The sixth section uses two cases of collaboration scripts to provide the evidence required for our main claim: The first case shows that collaboration scripts can be implemented in S-COL with no more effort compared to an implementation as part of a specific learning platform. It deals with the re-implementation of a collaboration script for the construction of single arguments, which was originally implemented as an embedded part of a specific learning platform and was effective with respect to activities and outcomes of collaborative learning in prior studies. In order to allow for an evaluation of the claim that implementation in S-COL requires no more effort than in a learning platform, the process of implementing the script as part of the specific learning platform is compared to the process of implementing it in S-COL, and the effort required to reuse it with other learning platforms is described. The second case shows both how S-COL ha l-0 07 03 89 6, v er si on 1 4 Ju n 20 12 S-COL: Copernican Turn for script development 7 can be used to provide support on top of diverse Web content, and that collaboration scripts implemented in S-COL can foster specific learning activities. A specific script for such an exemplary activity, collaborative online search, is described in detail. Findings from an empirical study about its effects on learning activities during collaborative online search are reported. The final section discusses further potentials as well as limitations of the tool and indicates unresolved problems associated with the approach. Support for learning on top of diverse Web content and reusable collaboration scripts for different learning platforms: Two unrelated problems? As indicated in the introduction, our approach to the development of flexibly reusable collaboration scripts was developed from a rather general perspective on the problem which we adopted while looking for a way to provide support for ninth-grade high school students during collaborative online search. In an interdisciplinary collaboration involving educational psychologists and computer scientists, our goal was to develop a tool that enables learners working on different computers to conduct collaborative online searches and provides contentand role-specific support for this collaborative task to each participant. Because SCOL was designed to solve both problems, we first describe this other setting. Support for learning on top of diverse Web content With the rapid development of information technology and its role in work and everyday life, online search competence is becoming more and more important as a crucial prerequisite for participation in society (e.g., Bilal, 2002; Ikpeze & Boyd, 2007). When using a search engine, specific cognitive processes are required to conduct a successful online search (e.g., Pirolli, 2005): For example, while a user is on the search query form of a search engine, he or she ha l-0 07 03 89 6, v er si on 1 4 Ju n 20 12 S-COL: Copernican Turn for script development 8 needs to generate a set of search terms. This set of search terms should, on the one hand, yield results containing the information needed and, on the other hand, preclude results that are irrelevant to his or her demands. While he or she is at the results page, hits need to be selected based on an evaluation of the information provided along with them (link title, text excerpt, and URL). At the pages reached from there, search strategies have to be applied to locate relevant information on the website. Typically, novices struggle with these cognitive processes: They are less inclined to strive for an overview of information available about a specific topic and to make a plan on how to proceed in an online search (Luconi & Tabatabai, 1999; Rogers & Swan, 2004). Furthermore, they often choose suboptimal search terms (Tomaiuolo & Packer, 1996) and are disoriented in their navigation behavior (Ikpeze & Boyd, 2007). Most importantly, novices need to learn how to evaluate information with respect to its trustworthiness and its relevance for their personal informational needs (Luconi & Tabatabai, 1999; Walton & Archer, 2004). In order to support learners to master these cognitive processes, it is recommended as one among several components that learners receive “just in time” assistance (Van Merrienboer, Clark & de Crock, 2002). The only way of delivering just–in-time assistance for online search activities that we found in the literature was by a teacher (Ikpeze & Boyd, 2007). However, a teacher cannot provide just-in-time assistance to all learners in a classroom. Instead, a feasible approach would rely on computer-based scaffolds that display just-in-time assistance to each individual learner (Pea, 2004; Puntambekar & Hübscher, 2005; Quintana et al., 2006). Furthermore, collaborative online search may be an appropriate setting for fostering the acquisition of online search competence: Research has shown beneficial effects of collaborative online search on the strategies employed, although not yet on learning (Lazonder, 1995). So far, there are few solutions for providing support on top of existing Web pages for individuals or groups of learners. One such solution is Greasemonkey, a browser plug-in that ha l-0 07 03 89 6, v er si on 1 4 Ju n 20 12 S-COL: Copernican Turn for script development 9 allows for changes of the content of a Web page as it is displayed to the user (Greasemonkey, 2009). This functionality could be used, in principle, to incorporate scaffolds into existing Web pages, such as reflection prompts on a Google results page. However, first of all, this approach is fragile to any changes in Web pages that may occur at any point in time. In addition, the Web pages traversed during online search are very diverse, both in terms of content and technical structure. Accordingly, it is rather difficult to develop scaffolds that can be integrated in any Web page encountered during an online search. Therefore, a desideratum for supporting the acquisition of online search competence is a technical solution for implementing scaffolds for individuals or collaboration scripts that can guide collaborating learners strategically during the different stages of an online search depending on where they are in the search process, but apply to any kind of topic as well as to any kind of Web pages encountered. Reusable technology-based collaboration scripts for different learning platforms As we have indicated above, collaboration scripts have typically been implemented as embedded parts of specific, often experimental, learning platforms. In some of our own studies (e.g., Stegmann, Weinberger, & Fischer, 2007; Weinberger et al., 2005), we used a discussion board developed by ourselves because this allowed for the easy implementation of the collaboration scripts under investigation in the learning platform itself: For example, a script for the construction of arguments can easily be implemented by means of prompts and separate textboxes for the parts of an elaborated argument. These textboxes can be embedded in the form for entering messages and their contents can be composed into one continuous message before posting the contribution (see Figure 3, part a). Thus, the collaboration scripts were always part and parcel of the learning platform itself. ha l-0 07 03 89 6, v er si on 1 4 Ju n 20 12 S-COL: Copernican Turn for script development 10 With the accumulation of findings about beneficial effects of collaboration scripts, the question arose how collaboration scripts that have been developed and tested in the context of a specific learning platform can be transferred and reused in the context of other learning platforms. The problem was framed as the task to integrate the original collaboration script into other learning platforms, that is, to get other learning platforms to display the components of the original collaboration script as part of the new learning platform. This approach led the way to the development of a universal formal language for the description of collaboration scripts (Kobbe et al., 2007; cf. also Kollar, Fischer & Hesse, 2006). This language is intended to be used for the specification of collaboration scripts that can be “imported” by different learning platforms and used as a basis to display the components of the original script as an embedded part of their interface. This universal scripting language accommodates a small but still comprehensive number of components and mechanisms of computer-supported collaboration scripts: The components are participants, activities, roles, resources, and groups; the mechanisms comprise task distribution, group formation, and sequencing. On the basis of this universal scripting language, a graphical modelling tool for designing new collaboration scripts has been developed (Harrer & Malzahn, 2006). As an output, the modelling tool produces an IMS-LD file, that is, a file that can be read by all learning platforms that support the IMS Global Learning Consortium Standards (cf. Miao et al., 2007). Based on these ideas, a functional framework for accelerating the implementation of scripts represented in IMS-LD for devices such as tabletop displays or mobile phones has been developed (Stegmann et al., 2009). However, we are not aware of any learning platform that is generally available and can import and implement a description of a collaboration script as an IMS-LD file using this IMS-LD extension. A further promising approach to provide a universal language for the scaffolding of collaborative learning is the “Learning Activity Management System” (LAMS; Dalziel, 2003). LAMS provides a graphical modelling tool for sequencing a variety of predefined ha l-0 07 03 89 6, v er si on 1 4 Ju n 20 12 S-COL: Copernican Turn for script development 11 activities (e.g., a chat tool followed by an individual phase, followed by a plenary discussion). The sequences of activities designed with LAMS can be integrated into several learning platforms such as Moodle, Sakai, or Blackboard. However, the activities that can be sequenced are restricted by the activities available in the graphical authoring tool. Furthermore, the activities cannot be “micro-scripted,” that is, learners can be prompted to discuss, but specific activities during discussion, such as the formulation of arguments, cannot be supported. Another approach is “ManyScripts” (Dillenbourg & Hong, 2008): This tool offers teachers an environment to adapt a set of specific scripts with regard to their own needs, especially their own learning material. At the moment, the Concept Grid, Argue Graph (Dillenbourg & Jermann, 2007), and Ice (Dillenbourg & Hong, 2008) scripts are available (Manyscripts, 2009). For example, the Argue Graph script forms groups of students with divergent opinions with respect to a specific domain (e.g., drug use in sports). To adapt the Argue Graph script, teachers can easily define their own questions that will be used to form these divergent groups. However, the ManyScripts environment is a stand-alone learning platform. A native integration into other learning platforms has not been a goal and is not supported yet. Consequently, currently neither the universal scripting language and graphical modelling approaches, nor the ManyScripts approach are suitable for developing new scripts and implementing them on a broad range of different learning platforms. A framework that effectively supports the reusability of technology-based collaboration scripts is not available so far. The transfer of a collaboration script from one collaborative-learning platform to another is still hampered by the need to adapt and integrate the script into the new learning platform. Therefore, a solution for using scripts developed and tested on one learning platform on other platforms is also still a desideratum. ha l-0 07 03 89 6, v er si on 1 4 Ju n 20 12 S-COL: Copernican Turn for script development 12 The basic idea for a comprehensive solution At first glance, it might seem that the two problems, that is, the development of scaffolds and collaboration scripts for collaborative learning on top of diverse Web content and the reusability of technology-based collaboration scripts, are unrelated. But at a second glance, these two problems are closely related to each other: The reuse of technology-supported collaboration scripts is hampered by the endless variety of possible learning platforms in which scripts should be implemented. If different learning platforms which typically can be accessed via a Web browser are viewed as but one special case of diverse Web content, the problem of transferring collaboration scripts between platforms becomes a special case of the problem of the development of scaffolds and collaboration scripts applicable to diverse Web content. The basic idea to solve this problem is to implement scaffolds and collaboration scripts as part of the browser and trigger them based on the recognition of types of functionally equivalent pages on the Internet or within the learning platform. With respect to the example of support for online search, this approach takes advantage of the fact that any search engine such as Google, Yahoo!, or Bing consists of a form for entering a search query that leads to a series of results pages with a common structure. From here, the user can reach Web pages that may contain the information he or she seeks. Accordingly, there are three types of functionally equivalent pages that users have to traverse during online search whatever Web search engine they may be using: (i) the search query form, (ii) the results page, and (iii) the external Web pages reached from the results page. If a component of the browser manages to recognize these three types of page, it can trigger specific kinds of support embedded in the browser. A search query form, for instance, typically contains one (or sometimes several) text field(s) for entering search terms and a button for starting the query. Such page-specific components in combination with the specific URL of the page can ha l-0 07 03 89 6, v er si on 1 4 Ju n 20 12 S-COL: Copernican Turn for script development 13 be used as a basis for the recognition of the page types. As each of these page types corresponds to a specific phase during an online search, specific support for the cognitive processes associated with each of these phases (e.g., Pirolli, 2005) can be provided. The situation is similar in the case of collaboration scripts for online discussions on collaborative-learning platforms. Many learning platforms such as Moodle, Sakai, and Blackboard contain asynchronous discussion boards. Any discussion board contains functionally equivalent pages such as the form for entering messages. In most learning platforms, the form for entering messages consists of functionally equivalent parts such as separate fields for the message and its title as well as a button for posting the message. Again, if a component of the browser manages to recognize this type of page and the types of its component objects, the components of a collaboration script pre-implemented in the browser can be triggered. The prompts and textboxes constituting the collaboration script can be displayed in a separate area of the browser window, and the contents of the single textboxes can be composed and sent to the message field when posting the message. The advantage of this approach lies in the fact that it allows for the use of a library of already implemented collaboration scripts contained in the browser that can be used with a broad variety of Webbased collaboration tools. Main features of S-COL We now turn to the implementation of these ideas as part of a tool we developed in order to demonstrate that the two interconnected problems described above can be solved in this way and to create a technical frame for providing support for computer-supported collaborative (and individual) learning on the Web. ha l-0 07 03 89 6, v er si on 1 4 Ju n 20 12 S-COL: Copernican Turn for script development 14 The graphical user interface The tool was implemented as a browser plug-in. Accordingly, the main part of its graphical user interface is the browser itself. The area of the browser used for displaying Web pages is broken up in two parts (see Figure 1). The area on the right-hand side is called the “browsing area.” It exhibits exactly the same behaviour as a standard Web browser: It can present any kind of Web page, and the user can navigate by using links and menu elements of the browser such as the home, forward, and backward buttons or entering a URL. The part on the lefthand side is called the “scaffolding area.” Its size is flexibly adaptable both by the user dragging its border as well as by programmed functions (in JavaScript). Furthermore, it can be invoked and hidden by a function key. Its content can be flexibly designed using HTML. The content of the scaffolding area (textboxes, buttons, etc.) can “interact” with objects in the browsing area. For instance, information from the browsing area such as the content of tables and textboxes or the URL of the actually displayed Web page can be read out. Furthermore, the browsing area can be controlled and manipulated by the tool by means of automatically posting text into forms, activating buttons, or even by navigating to an arbitrary URL. The scaffolding area moreover contains a menu bar (right above “Evaluation of the results page” in Figure 1) providing functionalities such as loading collaboration scripts or scaffolds and configuring the navigation behaviour of the tool (see below). ***** Figure 1 here *****