Collaboratively seeking and constructing information is becoming an increasingly important activity in our lives. Yet, many important aspects of such common undertakings are still understudied or unclear, including the effect of spatial configuration in information synthesis or writing process. This article investigates the effects of location-based setups on collaborative writing in the context of information seeking. These investigations are done using a laboratory study and an experimental system for collaboratively seeking and synthesizing information. A total of 30 pairs (60 subjects) were randomly assigned to one of the three experimental conditions: (1) on a single shared workstation; (2) in the same room with individual computers; and, (3) remotely located with text chat. By analyzing the way the collaborative pairs constructed their reports, it was found that those remotely located spent less effort per edit than those in the same room. Separate location discouraged social interactions and boosted the writing efficiency. These differences in various measures across different configurations, as discovered through exploratory analysis — both qualitative and quantitative — provide useful guidelines for developing collaborative systems, specifically presenting trade-offs among communication efforts, social interactions, and productivity. The work reported in this paper is a first known attempt to investigate impacts of spatial configuration on collaborative writing in the context of information seeking. In doing so, it brings together, and benefits to, scholarly inquiries from computer-supported cooperative work (CSCW) and library and information science (LIS) fields.
Results and analysis
Recent trends in the globalization of work-based tasks support an increased demand for collaboration between workers, with clear indications that this trend will only intensify (Denning, 2007; Denning and Yaholkovsky, 2008). Additionally, whether that work is assigned to large, international teams, homogeneous or heterogeneous groups working asynchronously, or even co-located pairs — working synchronously or asynchronously — there is the clear and unspoken requirement that much — if not all — of the work produced be computer-based. Thus, even co-located co-workers will utilize software for the co-creation of product, and also for communication around production. More and more situations expect or require people to work together, and the types of problem they are expected to solve continue to expand.
It is within this context of computer-mediated collaborative work that the present article investigates collaborative synthesis, specifically collaborative writing, focusing on the effects of spatial settings. More precisely, the work reported here uses a laboratory study to explore the result of placing collaborators in different spatial configurations on their collaborative writing as a part of an information seeking task.
The increasing integration of social media and collaborative environments into human interaction outside of work contexts reinforces both expectations and dispositions for collaboration across an inclusive range of information seeking contexts. An important task in many of today’s collaborative projects is seeking information (Foster, 2006) as well as synthesizing an information object, such as a report, a wiki entry, or an assessment. Often people use ad hoc methods and familiar tools that are not designed for easily supporting such collaborative information seeking (CIS) or sense-making activities while working on these projects. Tools for producing user-generated content such as Wikipedia, facilitate dialog and coordination among participants. Similarly, sites such as Delicious allow users to build recommender systems that reference shared preferences in web-based information sources. Increasing attention has been paid to improving precision with respect to the location of web-based resources: ‘social search’ capability is maximized by systems that ‘know’ who a searcher knows and interacts with, and incorporate indexing systems and controlled keyword assignment with social tags in Web-based queries (Evans and Chi, 2010).
Commercially developed tools (many in support of specific domains) have recently joined and challenged these broader spectrum production tools, presenting users with a complex array of potential software venues. There have been several efforts that show what a specially designed tool could do in such situations. For instance, Olson, et al. (1993) developed ShrEdit, a shared text editor. To their surprise, they discovered that the groups working with ShrEdit generated fewer design ideas, but apparently better ones. They believed their tool helped the supported groups keep more focused on the core issues in the emerging design, to waste less time on less important topics, and to capture what was said as they went. Likewise, Paul and Morris (2009) showed how a tool specifically designed for supporting collaborative sense-making in an information seeking task could enhance both information seeking and sense-making.
There remains, however, a lack of research that addresses the relative strengths of these as tools for collaborative synthesis in the context of information seeking: specifically, a lack of understanding of how spatial and temporal configurations affect collaborative work, and why particular tools for collaboration are preferred in specific information seeking contexts. As Grudin  pointed out, “many expensive failures in developing and marketing software that is designed to support groups are not due to technical problems; they result from not understanding the unique demands this class of software imposes on developers and users.” Shah (2010a) also presented a list of challenges for designers of CIS software, and one of them is addressing a real need with the right tools in a collaborative setup.
An extensive literature documenting the nature of human collaboration has been folded into research/experimentation developing software systems capable of facilitating collaborative information seeking (CIS). Such software has addressed the disparate requirements of domains from office work (Hansen and Järvelin, 2005) and design teams (Poltrock, et al., 2003) to health care (Reddy and Spence, 2008). Perhaps one significant indication of the ongoing proliferation of computer-mediated collaborative work is its incursion into earlier educational levels: students in elementary contexts are increasingly required to function as online collaborators (Hyldegård, 2006; Wheeler and Greenstein, 2007). To contextualize the work reported here, further discussion is presented using two aspects: computer-supported cooperative work (CSCW) and library and information science (LIS).
Computer-supported cooperative work (CSCW) aspect
In the history of CSCW, collaborative information synthesis within the context of information seeking is recognized as significantly understudied. One reason for this is the ubiquity and great complexity of human communicative interaction, and, the manner in which that is understood and translated into computer mediated behavior and/or communication. Gutwin and Greenberg (2002) summarize the detailed analysis of face-to-face (FtF) collaboration that fed into the design of collaborative search software. Although a decade old, their work remains valuable for its itemization of the many aspects of communication that must be addressed in translating human intercommunication into electronic affordances. It also serves as an important reminder of the need for grounding work in empirical observations of behavior, and reflects the extent to which we have so accommodated our own behavior to those affordances that we no longer criticize, question, or even notice many of them.
The information gathering tasks and presentation carried out within the kind of CIS situations considered here offer a valuable field for the exploration of a number of models within CSCW and information science fields, situated within an online collaborative context. As assigned tasks, essentially comparable to those routinely imposed in a variety of workplace settings, they can be considered as examples of cognitive work, defined within cognitive work analysis theoretical models in terms of “[t]he work actors do; their information behavior; the context in which they work; and the reasons for their actions” .
Cognitive work analysis (CWA) contends that information system must consider the needs of information actors in the determination of optimal system features. It argues that the needs of information system users should guide “the analysis of human-information interaction in order to inform the design of information systems” . It thus focuses on the information behaviors of people who interact with information: the effects of those features as configured within the problem defined here in terms of relative physical placement of collaborators and the tools of location and communication available to them offer clear affirmation of principles of CWA, as well as a field for the determination of relationships between task completion and the elements of integrated information systems.
Library and information science (LIS) aspect
Addressing CIS and collaborative synthesis activities from information seeking view, or more generally LIS perspective, researchers (e.g., Hyldegård, 2006; Shah and González-Ibáñez, 2010) have looked at Kuhlthau’s (1991) information search process (ISP) model. This model provides a framework within which to assess a range of contextualized information seeking processes in terms of a broadly recognized normative schema. The ISP has been widely accepted as reflective of a number of commonly occurring, process-based information seeking tasks, primarily those assigned within an educational program and carried out in a school or academic library.
The nature of the affective process discovered and described within Kuhlthau’s ISP has been shown to be susceptible to variations in the structure of the task, as well as the extent to which the task is intrinsically motivated, and to which collaboration between participants is required (Hyldegård, 2006; Vakkari, 2003). The truncated nature of the imposed task, and the diminished scale of the process through which the task is enacted suggest that the selection and evaluative synthesis of broadly disseminated information into a coherent information ‘product’ is perhaps a definitional element of any information seeking tasks, and that the nature of the affective process engaged in is significantly modified by considerations of communication and the requirement of embedded collaborative co-construction.
The nature of the specific tasks performed within this research problem, can be framed within Gross’ (1995) Imposed Query Model, as well as the ”problem-treatment process“ proposed by Wersig and Windel (1985), according to which different stages in the problem-solving process require access to different types of information, offering insight into the relationship between the successful resolution of an imposed query and the “availability of resources, quality of available resources, [and] appropriateness of resources to the question” . Imposed queries remain the primary currency of formal education, and are ever-present in the workplace: in both of these it is increasingly contextualized within peer groups rather than the work of individuals.
As evident from the brief reviews of CSCW and LIS views, they lack research that address collaborative synthesis in the context of information seeking. Specifically, CSCW has historically focused on collaborative work, including collaborative construction of information objects, but often ignoring its connection to information seeking. LIS literature, on the other hand, contains a plethora of research on information seeking, sense-making, and information synthesis, but almost always studied for individuals. These shortcomings pave a way for the scholarly inquiry as presented here — investigating collaborative synthesis in the context of information seeking, specifically with varying spatial configurations. It is important to note that in our previous work (González-Ibáñez, et al., 2013; González-Ibáñez, et al., 2012) an investigation regarding differences in information synthesis across conditions was performed where readability scores and human coders were employed in order to rate the quality of reports, however no significant differences were found in the quality of the reports generated by the groups in different conditions. In this article, a closer look to information synthesis is taken by evaluating the collaborative process in the production of the reports.
This paper investigates one of the important factors — space — affecting how people synthesize information together, in the hope that a better understanding of this will enable us to create right (or more suitable) collaborative support to address the real (or more contextualized) needs of collaborators. Rather than addressing specific hypotheses or research questions, the goal of the work discussed here is to investigate in an exploratory manner the differences among the collaborative teams working under different spatial conditions. A laboratory study was designed with three conditions defined using spatial configurations of collaborators — those working at the same desk, those working in the same room using different computers, and those remotely located. They were given the same tools, more importantly a collaborative editor much like Olson, et al.’s (1993) ShrEdit or Google Docs, and asked to perform a task that required them to search, collect, and compile information together.
This study was conducted involving a total of 60 participants in 30 collaborative pairs. Several prior works (e.g., Morris, 2008; Shah, 2010a; Morris, et al., 2010) have found that typical collaborative group size ranges from two to five members, with two being the median. It is expected that group size will have impact on costs and benefits, as well as effectiveness and efficiency of collaboration. To avoid any intervening effects, the group size was fixed at two. Another factor that played a part in this decision was based on previous studies in which it has been stated that as the group size increases, it also increases the number of interactions among group members, thus affecting interpretation of results (Tang, et al., 2010). Limiting group sizes to two allowed studying how people write together while keeping the amounts and the nature of interactions manageable. In this section a description of the rest of the design decisions and the details of the study are presented.
Participants in this study were students recruited from Rutgers University through open calls that were spread through various e-mail lists. The participants were required to sign up in pairs with someone with whom they had previously worked. This design decision was made in order to ensure that participants had common ground (Clark and Brennanm, 1991) and make the collaborative task more realistic. Such requirements have been reported in some of the earlier studies by various authors (e.g., González-Ibáñez, et al., 2012; Shah and González-Ibáñez, 2011; Shah and Marchionini, 2010).
From the 60 participants that were recruited, their ages ranged between 18 and 30. Most of the participants (70 percent) reported using Windows Operating System. Moreover, 84 percent of the participants indicated having intermediate to advanced search skills.
The participants were paid US$10 (per person) for participating in a 60–70 minutes experiment session. In addition, they were informed that the three best performing teams would receive additional compensation per participant at the end of the study. This consisted of US$50 for first place, US$25 for second place, and US$15 for third place. The instructions about what constitutes as the best performance to qualify for these prizes were intentionally kept vague since some of the pilot runs of this study revealed that a clear criteria (e.g., more information, better performance) often led to participants trying to optimize that single aspect of the whole task and ignoring the rest. The feedback from pilot tests, also verified from the actual study, informed the researchers that such indefinite instruction about performance was appropriate to ensure the participants take the task seriously and approach each of its aspects sincerely. Finally, it was also ensured that subjects participated only once irrespective of their group membership.
Collaborative search system
A modified version of Coagmento (Shah, 2010b), a Web-based system and an open source plug-in for the Firefox Web browser, was used. This version of Coagmento provided a set of tools for supporting information seeking, sharing, synthesis, as well as communication for teams (pairs).
Figure 1: A snapshot of the experimental system with parts of it shown in details. Note: Larger version of figure available here.
As depicted in Figure 1, the Coagmento plug-in consists of two major components: toolbar and sidebar. The toolbar contains four buttons: (1) Home, which redirect users to pages containing instructions or questionnaires at different stages of the study; (2) Bookmark that enables users to save, comment, rate, and share any Web page that users find to be relevant for the task being addressed; (3) Snip for saving and sharing portions of texts of a given Web page, along with comments and rating; and, (4) Editor, which opens a collaborative editor for writing the report required in the task (see description below).
The sidebar serves two primary functions: provide awareness of team’s actions and histories relating to bookmarks, snippets, and searches (see work by Shah and Marchionini (2010) for how such awareness functionality has been used in CIS), and allow the team members to communicate using a text chat channel.
A centerpiece of Coagmento system, and certainly one of the most important tools for the study reported here, is a collaborative editor based on Etherpad (http://www.etherpad.org). This editor allows the collaborators to type and see in real time what another team member is typing. As shown in Figure 2, the editor assigns a different color to each collaborator’s text to clearly distinguish individual contributions. The editor also saves the typed text automatically, creating different versions.
Figure 2: A snapshot of the shared editor. Note: Larger version of figure available here.
Beyond the features for supporting collaboration among team members, Coagmento also provides powerful logging functionality capable of recording users’ actions within the Firefox Web browser. This includes pages visited, queries, text messages, bookmarks, and snippets, among other useful data. All of these data are recorded with timestamps. Coagmento was used to record the Web browsing activity of users, messages exchanged during the collaboration process, and data from questionnaires that were introduced as part of the system. Additionally desktop activity was recorded using Camtasia Studio 7 (http://www.techsmith.com/camtasia.html).
The collaborative editor automatically saved not only the final reports that each team generated, but also incremental versions with timestamps.
Coagmento was adapted to guide users through various stages in the session workflow of this study. In each session participants were instructed by the researcher conducting the study to follow the steps in Table 1, using the text chat channel provided by Coagmento.
Table 1: Summary of session stages. Stage Description Time (minutes) 1 Participants were introduced to the study and asked to sign a consent form. 3 2 Participants watched a brief tutorial in order to learn the basic functionalities required during the task. 3 3 Participants individually filled out a set of pre-task questionnaires. 4 4 Participants read the task description (presented later). 3 5 Each team worked for approximately 30 minutes on the given task that included searching and collecting relevant information, and using it to compose a report. 30 6 Participants filled out post-task questionnaires. 6
The participants were asked to collect relevant information in an exploratory search designed to be a realistic task (Borlund and Ingwersen, 1999). The topic of “Gulf oil spill” was selected for the task, which according to few pilot runs was found to be appropriate in terms of the amount of material available on this topic and how engaging it was for the participants. The task description was presented as follows:
“A leading newspaper has hired your team to create a comprehensive report on the causes, effects, and consequences of the recent Gulf oil spill. As a part of your contract, you are required to collect all the relevant information from any available online sources that you can find. To prepare this report, search and visit any Web site that you want, and look for specific aspects as given in the guideline below. As you find useful information, highlight and save relevant snippets. Make sure you also rate a snippet to help you in ranking them based on their quality and usefulness. Later, you can use these snippets to compile your report, no longer than 200 lines, as instructed. Your report on this topic should address the following issues: description of how the oil spill took place, reactions by BP as well as various government and other agencies, impact on economy and life (people and animals) in the Gulf, attempts to fix the leaking well and to clean the waters, long-term implications and lessons learned.”
The tasks carried out within Coagmento do not require participants to engage with a process of task initiation as described by Kuhlthau (1991), during which the information seeker recognizes the need for new information to complete an assignment, with its associated feelings of apprehension and uncertainty. The seeker is also relieved of the responsibility to select a topic, the ISP’s second stage, and the presentation of work is relatively perfunctory and circumscribed. There remain, however, the requirements that participants locate relevant information — a process requiring evaluative judgments — and that the information selected be integrated — synthesized — into a product measured against imposed specifications. This design allows for a more focused investigation of collaborative information seeking and synthesis processes performed in different spatial conditions described below.
The participating groups were randomly divided among one of the three conditions.
C1Together: Two participants working together at the same computer (Figure 3).
C2Colocated: Two participants co-located with different computers and face-to-face (FtF) communication (Figure 4).
C3Remote: Two participants in different rooms with text-based communication (Figure 5).
Figure 3: C1Together: Condition where both the participants shared the same space and computer.
Figure 4: C2Colocated: Condition where both the participants shared the same space, but had individual computers.
Figure 5: C3Remote: Condition where both the participants had their individual space and computers.
In condition C3Remote, participants were located in different rooms, without direct visual or auditory contact, able to communicate only through the text chat provided through Coagmento.
The study was intended to atomize — detail — the elements of interaction involved in Web-based collaborative search, synthesis, and writing, in order to observe the nature of computer-mediated coordination within the context of an assigned task. As noted by Fidel, et al. (2004), the number of possible dimensions present in a collaborative task (independent of context) is significant: these include task definition and source-location, as well as cognitive, organizational, and the affective, among others. Collaborative groups — or pairs — operate within interrelational as well as cognitive space. For the purposes of the study reported here, participants were provided with an assigned task that was relatively simple, in order to free them from the need to develop a task definition, compose queries, or conform to assigned roles (writer, gatherer, etc.). The subject of the “focused observations”  in this study was the spontaneous coordination of activities related to searching, collating, and writing between participants.
Results and analysis
Several forms of data were collected as evident from the previous section. The analyses of search-related data have been reported before (Shah and González-Ibáñez, 2011). Given that the focus of the current paper is on the collaborative writing process and products, the results and analyses here are primarily concerning the report writing activities.
Both qualitative and quantitative analyses were run on the collected data. For the qualitative part, an interface that was available as a part of Coagmento system was used, which is very similar to the collaborative editor. As shown in Figure 6, this interface integrates the collaborative editor and provides additional tools for analysis. One of such tools is a slider at the top that one can use to skim through various versions generated automatically during the process of writing the report (see a video of this at http://coagmento.org/videos/collab_editor.mov, which shows how this is done). The purpose of qualitative analyses was to validate finding from quantitative analyses as well as to obtain more details about them. In the remaining of this section, both quantitative and qualitative analyses are intertwined in the narrative.
Figure 6: Interface for analyzing collaborative report writing. Note: Larger version of figure available here.
First, analysis was done to understand the level of activities as people in different conditions collaborate while writing their reports. For each condition, the number of words and number of versions present at the end of the task were recorded, as well as the number and types of edits applied during the process of writing. Table 2 presents various statistics relating to the length of the reports and the number of versions it took the participants to finish them.
Table 2: Descriptive statistics about the reports per condition. Condition Average length of the report
(characters w/o spaces)
Average number of words Average number of sentences Average number of versions C1Together 2836 (S.D.=1253.53) 445 (S.D.=199.45) 23.20 (S.D.=8.97) 479 (S.D.=195.22) C2Colocated 2472 (S.D.=1351.09) 481 (S.D.=259.86) 19.30 (S.D.=6.93) 1126 (S.D.=423.65) C3Remote 3590 (S.D.=1908.74) 680 (S.D.=364.30) 38.64 (S.D.=23.43) 830 (S.D.=543.79)
The quantitative investigation began by first analyzing all the variables for normality using Shapiro-Wilk test. Based on this testing, versions-per-characters and versions-per-words were the only variables normally distributed. Table 3 provides summary statistics of these two variables.
Table 3: Condition-wise effort a measured by number of versions per production unit (characters/words). Condition Versions per character Versions per word C1Together 0.32 (S.D.=0.27) 1.53 (S.D.=1.24) C2Colocated 0.51 (S.D.=0.21) 2.61 (S.D.=1.05) C3Remote 0.28 (S.D.=0.18) 1.45 (S.D.=0.90)
Due to the normality of the distributions, one-way ANOVA was performed on these variables and versions-per-words was found to have significant difference (p<0.05) among the distributions of the three conditions. Using post hoc test of Games-Howell (equality of variance not assumed), C2Colocated was found to have more versions per words produced than C3Remote. The variable versions-per-words can be seen as an indication of the relationship between the amount of text added, and the extent to which it was subsequently edited. The higher the ratio, the greater the effort for the text added with respect to the changes subsequently made. In other words, the more extensive the copy pasting of the collected pieces of information (snippets) and/or the less extensive the editing done of those snippets, the less the number of versions and thus, less effort. Note that the versions were automatically created by the system, and are indicative of the amount of editing activities within the editor. Thus, the results indicate that when the collaborators with their own computers were remotely located (C3Remote), they were able to accomplish the same amount of writing with less effort compared to the condition where they were in the same space (C2Colocated).
For the remaining of the variables of interest that were not normally distributed, Kruskal-Wallis test was performed to check for the equivalence of variance. Since the results showed versions variable significantly varied among the three conditions (p<0.01), further analysis was warranted. Using post hoc testing with Wilcoxon Rank Sum, it was found that C2Colocated had a higher number of versions than the other two conditions (p<0.05). In addition, C3Remote was found to have more number of versions than that of C1Together (p<0.05). This shows that both C2Colocated and C3Remote did more editing activities compared to C1Together. This can be attributed to the design of the spatial configurations, in which those in C1Together shared a computer, thus limiting their editing abilities. What is interesting, and once again confirming, is that the remotely located collaborators (C3Remote) created fewer versions than those in C2Colocated while maintaining the same level of output quantitatively (since no differences in report lengths were found) and qualitatively (see our previous work in González-Ibáñez, et al., 2013; González-Ibáñez, et al., 2012).
Going beyond what happened during collaborative writing part of the collaborative project, the report writing process was investigated to understand why the participants did what they did.
The figure of 445 words and 479 versions per team for C1Together — the lowest — perhaps reflects the limitation of a shared workstation: only one stream of snippets and edits created the final text. However, C1Together had more characters per version than C2Colocated, indicating that with respect to C2Colocated, larger bodies of “snippet” text were added at a time: that more copy-pasting was done. C2Colocated, interestingly, produced comparable number of words and more than twice the number of versions with respect to C1Together, yielding the highest ratios of versions to character and versions to words, indicating the least amount of copy-pasting in terms of the editing subsequently performed. C3Remote yielded the highest number of words and characters in 830 versions (approximately twice as many as C1Together but just under three-fourth of C2Colocated): its low scores for versions to character and versions to word ratios reflect a greater degree of copy-pasting of snippets (in terms of number of words produced) and considerably less editing of those snippets. This was also verified by examining screen videos and playing the editing process through the interface shown in Figure 6 for each team.
It is perhaps intuitive to account for the word production of C1Together with respect to the others as a function of the single, shared computer. Limitations of time and space correlate with the low word count, although the number of characters with respect to the number of words seems to indicate that larger words may have characterized final versions. In C1Together, the fact that participants focused upon a single point — the shared terminal screen — suggests that distractions were likely to be focused on the shared task and shared content, whereas, for example, in C2Colocated, distractions as well as computer input might be separately focused, reducing productivity.
Perhaps most striking is the production of C2Colocated, with respect to C1Together and C3Remote. Although the availability of a second workstation (with its added pair of actively typing hands) suggests that C2Colocated has an absolute/hypothetical capacity to produce twice the amount of text of C1Together. However, measured using number of words, C2Colocated is actually just slightly higher than C1Together’s single stream, and significantly lower than that of C3Remote. In other words, the production of these collocated (FtF) teams with two streams of input computer is similar to that of the single computer teams, rather than the other two-computer teams. C2Colocated also produced a number of versions significantly larger than that of C1Together or C3Remote, indicating that more operations — additions, deletions, and edits — were performed on that body of text, reducing their efficiency.
How might this phenomenon described in the previous section be explained? Analyzing audio and video data collected from these sessions, it is interesting to note that the FtF teams of C2Colocated did in fact communicate — verbally — on a number of extraneous subjects, leaving less time for the addition of text. The increased number of edits in C2Colocated implies, however, that in spite of their expenditure of increased ‘social’ energy in FtF pleasantries, more communication was actually task-related than that of participants in C3Remote (see Shah and González-Ibáñez, 2011 for more details). The discrepancy can possibly be correlated with the increased energy required of FtF interaction with respect to remote — if synchronous — communication; or to the effect of FtF positioning on any work that is shared, such that having a collaborator physically present perhaps enhances particular sorts of shared attention to a task. Collocation requires participants to expend energy on — and shift focus between — very different sorts of cognitive work: that of FtF communication and that of scanning, selecting, and manipulating text on a screen, thus increasing the amount of cognitive work that must be required to complete the task.
Teams within C3Remote produced more words/characters in fewer versions than those in C2Colocated. This suggests that remote location of collaborators may permit the focus of increased — and undivided — physical and cognitive energy within a shared task space, within which both text generation and task-related communication are framed, cognitively and situationally. Remote location relieves collaborators of the necessity to expend gestural, affective energy (González-Ibáñez, et al., 2012), and may facilitate an increased ability to ignore a collaborator’s physical presence to focus in the task.
To summarize, it was found that for the tasks that are clearly dividable and where the collaborators have symmetric roles, independence and a distraction-free communication channel while working together helps the collaborators accomplish more amount of work (C3Remote). On the other hand, if we care for social interactions and/or if a task was not easily dividable, the collaborators may benefit by being in the same space (C1Together or C2Colocated). The latter part of this statement is a hypothesis that is worth testing in future work.
Collaborative information task completion has been most typically addressed within CSCW, with a number of studies exploring the relationship between time-space and communicative conditions on one hand (Birnholtz, et al., 2013; Twidale and Nichols, 1996) and the performance of collaborative information tasks on the other hand (González-Ibáñez, et al., 2013; Zhang, 2002). However, the specific interrelationships and interactions between communication affordances and information behavior are understudied or often ignored. These two factors manipulated in the study described here have implications for theoretical considerations of information seeking overall. The tasks completed by study participants comprised a number of components integral to most information seeking behaviors: the location, collation, and coordination of information in response to an information need — intrinsic or imposed — are the key elements in what is arguably a growing range of information seeking tasks required in schools and the workplace.
The reported work is not without its limitations. For instance, the study described here is limited in its scope with respect to the task (exploratory search and report writing in time-bound fashion), collaborative configuration (synchronous, dyads), as well as outcome measurements. The reported work was also focused on spatial configuration as the primary variable of interest. This paves a way for future work that includes investigating a variety of collaborative conditions including time dimension, group size, and the nature of the task. Finally, collaborative synthesis (writing) here is studied in the context of CIS. There are several works in the literature that study collaborative synthesis without such association to information seeking (e.g., Blake and Pratt, 2006; Tao and Tombros, 2013). While this can be seen as a limitation, it also adds to the uniqueness of the work reported here.
Although it has long been recognized that variations in time-space and communicative conditions, like those imposed upon study participants, do in fact directly influence overall productivity and information synthesis (González-Ibáñez, et al., 2012; Shah and González-Ibáñez, 2012), the nature of specific effects and the dynamic between affective and situational elements have been much less frequently addressed (Birnholtz, et al., 2013). If, as this work seems to indicate, the configuration of specific means of information location, communication, and co-construction between collaborating information actors is of significance in the determination of the information search product, work that explores this nexus of relationships and effects is as broadly relevant to information retrieval/seeking as it is to the field of LIS.
About the authors
Dr. Chirag Shah is an assistant professor in the Department of Library and Information Science (LIS) within the School of Communication and Information (SC&I) at Rutgers University. He received his Ph.D. from School of Information and Library Science (SILS) at University of North Carolina, Chapel Hill, and M.S. in Computer Science from University of Massachusetts, Amherst. His research interests include studies of interactive information seeking, especially in the context of online social networks and collaborations, contextual information mining, and applications of social media services for exploring critical socio-political issues.
E-mail: chirags [at] rutgers [dot] edu
Dr. Roberto González-Ibáñez is an assistant profesor in Departamento de Ingeniería Informática within Universidad de Santiago de Chile. He received his Ph.D. from School of Communication and Information (SC&I) at Rutgers University. His research interests include studies of collaborative information seeking and affective dimension in information retrieval processes.
E-mail: roberto [dot] gonzalez [dot] i [at] usach [dot] cl
Pamela Read is a doctoral student in the Department of Library and Information Science (LIS) within the School of Communication and Information (SC&I) at Rutgers University. Her research interests include school library media and information seeking in adolescences.
E-mail: PRead [at] somersschools [dot] org
This work was supported by the U.S. Institute of Museum and Library Services (IMLS) Early Career Development grant #RE–04–12–0105–12.
1. Grudin, 1994, p. 93.
2. Fisher, et al., 2005, p. 89.
3. Fisher, et al., 2005, p. 88.
4. Fisher, et al., 2005, p. 165.
5. Foster, 2006, p. 334.
J. Birnholtz, S. Steinhardt, and A. Pavese, 2013. “Write here, write now! An experimental study of group maintenance in collaborative writing,” CHI '13: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, pp. 961–970.
doi: http://dx.doi.org/10.1145/2470654.2466123, accessed 24 November 2015.
C. Blake and W. Pratt, 2006. “Collaborative information synthesis I: A model of information behaviors of scientists in medicine and public health,” Journal of the American Society for Information Science and Technology, volume 57, number 13, pp. 1,740–1,749.
doi: http://dx.doi.org/10.1002/asi.20487, accessed 24 November 2015.
P. Borlund and P. Ingwersen, 1999. “The application of work tasks in connection with the evaluation of interactive information retrieval systems: Empirical results,” In: M. Dunlop, S. Draper, K. van Rijsbergen, and I. Ruthven (editors). Proceedings of MIRA: Electronic workshops in computing. London: British Computing Society (BCS), pp. 29–46, and at http://bcs.org/upload/pdf/ewic_mi99_paper1.pdf, accessed 24 November 2015.
H.H. Clark and S.E. Brennanm 1991. “Grounding in communication,” In: L.B. Resnick, R.M. Levine, and S.D. Teasley (editors). Perspectives on socially shared cognition. Washington, D.C.: American Psychological Association, pp. 127–149.
doi: http://dx.doi.org/10.1037/10096-006, accessed 24 November 2015.
P.J. Denning, 2007. “Mastering the mess,” Communications of the ACM, volume 50, number 4, pp. 21–25.
doi: http://dx.doi.org/10.1145/1232743.1232763, accessed 24 November 2015.
P.J. Denning and P. Yaholkovsky, 2008. “Getting to ‘we’,” Communications of the ACM, volume 51, number 4, 19–24.
doi: http://dx.doi.org/10.1145/1330311.1330316, accessed 24 November 2015.
B.M. Evans and E.H. Chi, 2010. “An elaborated model of social search,” Information Processing & Management, volume 46, number 6, pp. 656–678.
doi: http://dx.doi.org/10.1016/j.ipm.2009.10.012, accessed 24 November 2015.
R. Fidel, A.M. Pejtersen, B. Cleal, and H. Bruce, 2004. “A multidimensional approach to the study of human-information interaction: A case study of collaborative information retrieval,” Journal of the American Society for Information Science and Technology, volume 55, number 11, pp. 939–953.
doi: http://dx.doi.org/10.1002/asi.20041, accessed 24 November 2015.
K.E. Fisher, S. Erdelez, and L. McKechnie (editors), 2005. Theories of information behavior. Medford, N.J.: Information Today.
J. Foster, 2006. “Collaborative information seeking and retrieval,” Annual Review of Information Science and Technology, volume 40, pp. 329–356.
doi: http://dx.doi.org/10.1002/aris.1440400115, accessed 24 November 2015.
R. González-Ibáñez, M. Haseki, and C. Shah, 2013. “Let’s search together, but not too close! An analysis of communication and performance in collaborative information seeking,” Information Processing & Management, volume 49, number 5, pp. 1,165–1,179.
doi: http://dx.doi.org/10.1016/j.ipm.2012.12.008, accessed 24 November 2015.
R. González-Ibáñez, M. Haseki, and C. Shah, 2012. “Time and space in collaborative information seeking: The clash of effectiveness and uniqueness,” Proceedings of the American Society for Information Science and Technology, volume 49, number 1, pp. 1–10.
doi: http://dx.doi.org/10.1002/meet.14504901080, accessed 24 November 2015.
M. Gross, 1995. “The imposed query,” RQ, volume 35, number 2, pp. 236–243.
J. Grudin, 1994. “Computer-supported cooperative work: History and focus,” Computer, volume 27, number 5, pp. 19–26.
doi: http://dx.doi.org/10.1109/2.291294, accessed 24 November 2015.
C. Gutwin and S. Greenberg, 2002. “A descriptive framework of workspace awareness for real-time groupware,” Computer Supported Cooperative Work, volume 11, numbers 3–4, pp 411–446.
doi: http://dx.doi.org/10.1023/A:1021271517844, accessed 24 November 2015.
P. Hansen and K. Järvelin, 2005. “Collaborative information retrieval in an information-intensive domain,” Information Processing & Management, volume 41, number 5, pp. 1,101–1,119.
doi: http://dx.doi.org/10.1016/j.ipm.2004.04.016, accessed 24 November 2015.
J. Hyldegård, 2006. “Collaborative information behaviour — exploring Kuhlthau’s Information Search Process model in a group-based educational setting,” Information Processing & Management, volume 42, number 1, pp. 276–298.
doi: http://dx.doi.org/10.1016/j.ipm.2004.06.013, accessed 24 November 2015.
C.C. Kuhlthau, 1991. “Inside the search process: Information seeking from the user’s perspective,” Journal of the American Society for Information Science, volume 42, number 5, pp. 361–371.
M.R. Morris, 2008. “A survey of collaborative Web search practices,” CHI ’08: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, pp. 1,657–1,660.
doi: http://dx.doi.org/10.1145/1357054.1357312, accessed 24 November 2015.
M.R. Morris, D. Fisher, and D. Wigdor, 2010. “Search on surfaces: Exploring the potential of interactive tabletops for collaborative search tasks,” Information Processing & Management, volume 46, number 6, pp. 703–717.
doi: http://dx.doi.org/10.1016/j.ipm.2009.10.004, accessed 24 November 2015.
J.S. Olson, G.M. Olson, M. Storrøsten, and M. Carter, 1993. “Groupwork close up: A comparison of the group design process with and without a simple group editor,” ACM Transactions on Information Systems, volume 11, number 4, pp. 321–348.
doi: http://dx.doi.org/10.1145/159764.159763, accessed 24 November 2015.
S.A. Paul and M.R. Morris, 2009. “CoSense: Enhancing sensemaking for collaborative Web search,” CHI ’09: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, pp. 1,771–1,780.
doi: http://dx.doi.org/10.1145/1518701.1518974, accessed 24 November 2015.
S. Poltrock, J. Grudin, S. Dumais, R. Fidel, H. Bruce, and A.M. Pejtersen, 2003. “Information seeking and sharing in design teams,” GROUP ’03: Proceedings of the 2003 International ACM SIGGROUP Conference on Supporting Group Work, pp. 239–247.
doi: http://dx.doi.org/10.1145/958160.958198, accessed 24 November 2015.
M.C. Reddy and P.R. Spence, 2008. “Collaborative information seeking: A field study of a multidisciplinary patient care team,” Information Processing & Management, volume 44, number 1, pp. 242–255.
doi: http://dx.doi.org/10.1016/j.ipm.2006.12.003, accessed 24 November 2015.
C. Shah, 2010a. “Designs for systems to support collaborative information behavior,” In: J. Foster (editor). Collaborative information behavior: User engagement and communication sharing. Hershey Pa.: Information Science Reference, pp. 141–158.
doi: http://dx.doi.org/10.4018/978-1-61520-797-8.ch009, accessed 24 November 2015.
C. Shah, 2010b. “Coagmento — A collaborative information seeking, synthesis and sense-making framework,” CSCW 2010: Proceedings of the ACM CSCW Conference, pp. 527–528; version at http://research.microsoft.com/en-us/um/redmond/groups/connect/CSCW_10/docs/p527.pdf, accessed 24 November 2015.
C. Shah and R. González-Ibáñez, 2012. “Spatial context in collaborative information seeking,” Journal of Information Science, volume 38, number 4, pp. 333–349.
doi: http://dx.doi.org/10.1177/0165551512438356, accessed 24 November 2015.
C. Shah and R. González-Ibáñez, 2011. “Evaluating the synergic effect of collaboration in information seeking,” SIGIR ’11: Proceedings of the 34th International ACM SIGIR Conference on Research and Development in Information Retrieval, pp. 913–922.
doi: http://dx.doi.org/10.1145/2009916.2010038, accessed 24 November 2015.
C. Shah and R. González-Ibáñez, 2010. “Exploring information seeking processes in collaborative search tasks,” ASIS&T ’10: Proceedings of the 73rd ASIS&T Annual Meeting on Navigating Streams in an Information Ecosystem, volume 47, article number 60.
C. Shah and G. Marchionini, 2010. “Awareness in collaborative information seeking,” Journal of American Society of Information Science and Technology, volume 61, number 10, pp. 1,970–1,986.
doi: http://dx.doi.org/10.1002/asi.21379, accessed 24 November 2015.
A. Tang, M. Pahud, K. Inkpen, H. Benko, J.C. Tang, and B. Buxton, 2010. “Three’s company: Understanding communication channels in three-way distributed collaboration,” CSCW ’10: Proceedings of the 2010 ACM Conference on Computer Supported Cooperative Work, pp. 271–280.
doi: http://dx.doi.org/10.1145/1718918.1718969, accessed 24 November 2015.
Y. Tao and A. Tombros, 2013. “An exploratory study of sensemaking in collaborative information seeking,” In: P. Serdyukov, P. Braslavski, S.O. Kuznetsov, J. Kamps, S. Rüger, E. Agichtein, I. Segalovich, and E. Yilmaz (editors). Advances in information retrieval. Lecture Notes in Computer Science, volume 7814. Berlin: Springer, pp. 26–37.
doi: http://dx.doi.org/10.1007/978-3-642-36973-5_3, accessed 24 November 2015.
M.B. Twidale and D.M. Nichols, 1996. “Collaborative browsing and visualisation of the search process,” Lancaster University, Cooperative Systems Engineering Group Technical Report CSEG/3/1996; version at http://www.cs.waikato.ac.nz/~daven/docs/CSEG.3.96.pdf, accessed 24 November 2015.
P. Vakkari, 2003. “Task–based information searching,” Annual Review of Information Science and Technology, volume 37, pp. 413–464.
doi: http://dx.doi.org/10.1002/aris.1440370110, accessed 24 November 2015.
G. Wersig and G. Windel, 1985. “Information science needs a theory of ‘information action’,” Social Science Information Studies, volume 5, number 1, pp. 11–23.
doi: http://dx.doi.org/10.1016/0143-6236(85)90003-1, accessed 24 November 2015.
B. Wheeler and D. Greenstein, 2007. “Open source collaboration in higher education: Guidelines and report of the licensing and policy framework summit for software sharing in higher education,” at http://hdl.handle.net/2022/3076, accessed 24 November 2015.
X. Zhang, 2002. “Collaborative relevance judgment: A group consensus method for evaluating user search performance,” Journal of the American Society for Information Science and Technology, volume 53, number 3, pp. 220–231.
doi: http://dx.doi.org/10.1002/asi.10036, accessed 24 November 2015.
Received 21 February 2015; accepted 16 October 2015.
This paper is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Investigating impacts of spatial configurations on collaborative writing
by Chirag Shah, Roberto González-Ibáñez, and Pam Read.
First Monday, Volume 20, Number 12 - 7 December 2015