Examining value creation in a community of learning practice: Methodological reflections on story-telling and story-reading

Despite the abundant research on communities in various shapes and settings, examination of what community members gained from their participation remains a thorny issue. For this purpose, we adopted and refined the value creation framework developed by Wenger, Trayner and De Laat (2011) to divulge experienced values by community members through “scaffolded narratives” and categorization of the values reported through their stories. However, in doing so two methodological issues emerged – in particular in relation to “values”. This paper reports on our methodological reflection on the challenging process of capturing community members’ value creation within a community of learning practice. More specifically, we reflect on the following questions: (1) To what extent can the values that the participants originally intended to report be identified as such by the researchers/analysts’ without bias due to the researchers/analysts’ own perspectives? and (2) To what extent does a theoretically-driven pre-defined typology of values confine or enrich the range of possible values that can be identified? What adds to this challenging research endeavour is the concept of value in theoretical terms and its associated typologies. Hence, these methodological questions need to be discussed in order to comprehend both the phenomenon of value (creation) per se as well as how it is examined – as close to the participants’ reality as possible – since value creation is the driving force for the sustainability of a community.


The "value" of value
The concept of value is rich, complex and appealing, but often causes conceptual fuzziness among philosophers, sociologists, psychologists and anthropologists due to its either narrow or broad treatment -which in turn resulted in the development of several typologies of theories of value over the last three centuries (Kluckhohn, 1951;Rescher, 1969;Rokeach, 1973Rokeach, , 1979Schwartz, 1992;Williams, 1968). The concept of value might be interpreted as (a) the value of an object (associated with objectivism in value theory) or (b) the process of a subject attributing value to an object based on a set of criteria or standards (associated with subjectivism in value theory; i.e. valuation). However, this axiological division seems contradictory since the criteria or standards lead to the assignment of value to an object and the value to an object requires the existence of standards or criteria set by the subject (Pauls, 1990). Fronzini (1971) adopts a critical view towards the objectivism-subjectivism division and claims that value is a relational construct with the existence of both the subject and the object being prerequisites. The relational nature of value is reflected in Rokeach's (1973Rokeach's ( , 1979 definition of value as "(…) an enduring belief that a specific mode of conduct or end-state of existence is personally or socially preferable to an opposite or converse mode of conduct or end-state of existence" (Rokeach, 1973, p. 5). Rokeach (1979) further emphasized that values imply the presence of criteria or standards of preference for any selective orientation and constitute both guiding factors of expected and goal-oriented behaviour and justification/explanatory factors of past behaviour.
In the context of organizational management, but with a specific focus on communities of practice, Wenger, McDermott and Snyder (2002), proposed a typology of values for community members and organizations, including (a) short-term and long term values, (b) tangible and intangible values, and (c) strategy-implementing and strategy-making values. Their typology was further developed by integrating the idea of the value creation process, a notion which has been mostly associated with financial, organizational and strategic management (see Seth, 1990).
From a community and network perspective, Wenger et al. (2011) conceptualize value creation as "(…) the value of learning enabled by community involvement and networking" (p. 7) with communities or networks to serve as social settings for social learning activities (e.g., sharing ideas, coconstructing knowledge, exchanging experiences). The spectrum of value creation consists of five cycles of value, which do not necessarily need to be all covered or follow a linear sequence: (a) Immediate Value, (indicated by meaningful activities), (b) Potential Value (indicated by robust resources), (c) Applied Value (indicated by implementation of practices), (d) Realised Value (indicated by return on investment), and (e) Reframing Value (indicated by reconsidering ideas and frameworks). Within a community setting members might be involved in sharing of expertise, learning from each other's experiences, and helping each other with challenges. These activities might be related to the values individuals attribute to a community or derive from it (Wenger et al., 2011). The value of learning in a community derives from members' ability to develop a shared intention to enhance learning in a common domain. The shared domain of interest, shared practice (developed through a joint history of learning) and the shared repertoire (consisting of shared perspectives, strategies, and stories), all constitute learning resources for the community members (Wenger et al., 2011, p. 10).
Despite the abundant research on communities in various shapes and settings, examination of what community members gained from their participation remains a thorny issue. Notwithstanding the conceptual advancement by Wenger et al.'s (2011) value creation framework as "a means to appreciate value created in communities and networks" (De Laat, Schreurs, & Nijland, 2015, p. 254), methodological approaches to appropriately, precisely and authentically capture any reported value creation did not emerge concurrently. This gap between theory and method motivated our quest for an analytical discussion of methodological tensions when examining experienced values of participation as reported by members of community.
The main questions that we discuss in this paper are: (1) To what extent can the values that the participants originally intended to report be identified as such by the researchers/analysts' without bias due to the researchers/analysts' own perspectives? and (2) To what extent does a theoretically-driven predefined typology of values confine or enrich the range of possible values that can be identified? By addressing these questions we aim to depict the complexity of the methodological endeavour of capturing members' selfreported experienced values of community participation (i.e., story-telling) by researchers "from outside the box" and "labelling" participants' experienced values with pre-defined typologies (i.e., story-reading). Unravelling this methodological complexity can act as both "warning" and "support" for future researchers of communities and value creation therein.

Where we looked at: Communities of learning practice
Our methodological reflection emerged from our aim to examine the phenomenon of value creation in a Community of Learning Practice (CoLP). CoLP refers to a recombinant community model positioning itself in-between and beyond the existing community models of Communities of Practice (CoP) and Communities of Learners (CoL) (Dingyloudi & Strijbos, 2014). CoLPs are extra-curricular entities that derive from and operate in educational settings, in parallel with the curriculum -but not integrated into it -and with no predetermined pedagogical objectives. CoLPs emerge from students' common needs and are not used as an instructional approach by educators, researchers or stakeholders to enhance curricular learning objectives. Participants of CoLPs are fellow students who gather together as peers to address commonly identified needs that derive from the broader educational setting (e.g., academic challenges) through the sharing mechanism of peer feedback. Peers voluntarily participate in the CoLP and are free to withdraw whenever they sense that their participation is no longer of value to them. Any student can join the community as a plain peer willing to share, negotiate and co-construct learning experiences (Dingyloudi & Strijbos, 2014).
While taking into consideration the sociocultural claim that the individual, the group and the surrounding context cannot be studied in isolation due to their interrelation (Hatano & Wertsch, 2000;Sawyer, 2002), we identify the relevance and importance of examining personal, social, skill-related, studyrelated and contextual values within different cycles of value creation. Thus we aim to provide a variegated picture of the values that are associated with this interrelation among the individual, the immediate social (group) setting and the surrounding educational context.

Participants
The participants were eighteen international graduate students (Mage = 25.90, SD = 2.37, Agerange: 23-31) enrolled in a two-year research oriented Learning Sciences master's program. The 18 students were part of the same cohort of students (31 students) and part of one CoLP (22 students) in parallel to the master's program. Participation in the community was voluntary and participants were free to join or withdraw from any community meeting.

Design
The students voluntarily participated in seven community events (i.e., informal face-to-face community meetings), lasting approximately three hours each and taking place upon community members' request (weekly or biweekly basis). The community events were co-organized and co-structured among the community members and a participatory non-peer facilitator (who was also the lead researcher of the present study. The facilitator was present at every community event in order to support the members' interactions and community activities. Face-to-face peer feedback on "work-in progress for future delivery within the study programme" was one of the main sharing mechanisms in the community events and coordinated by the facilitator. The lifespan of the community was one semester at the beginning of their study programme. All community events were video-recorded with consent by the participants. The values experienced by the participants were collected with the help of narratives, more specifically their "value creation stories".

What we looked for: A situated multilevel typology of values
We employed a Situated Multilevel Typology of Values (SMTV) to study value creation in the CoLP. The premises of a situated perspective to values are (a) Fronzini's (1971) value contextualism theory which implies that the existence and meaning of values is situation-defined and situation-dependent, and (b) Lave and Wenger's (1991) situated approach to learning which implies that learning is a socially situated practice. Although the employed typology relies heavily on the idea of cycles of value creation as developed by Wenger et al. (2011), we extended their framework by including the pre-formation cycle of Expected value (i.e., reasons for participation, needs and expectations) which corresponds to the values that prompt the participants to partake in community formation in the first place (see Figure 1). This extension is based on the following premises (a) needs can act as prerequisites for the development of values (Pauls, 1990, p. 26) and (b) values themselves constitute guiding factors of future behaviour (Rokeach, 1979). Along with this theoretical alignment, the integration of Expected values as a pre-formation cycle is also in alignment with the first stage of community development as described by Wenger et al. (2002), during which a network of people identify the potential for a community to emerge through the identification of common interests and needs. These five types of value have been included in the SMTV due to their relevance to the social setting of CoLPs being examined (Dingyloudi & Strijbos, 2014), which in turn adds the situated nature of this typology. Although we developed the SMTV to study value creation from a situated perspective in the specific context of CoLPs, it also contributes to the theoretical and analytical development of Wenger et al.'s (2011) value creation framework for (online) communities in general -for example CoPs and CoLs often serve members' needs to develop a certain skill in relation to a practice, organization or study programme. However, even if the SMTV can be applied in structure (for the most part), the situatedness of participants' expression of their experienced value will necessitate that the typology is recalibrated (in terms of the description of codes and examples) to the observed setting.

How we looked for it: Narratives and value creation stories
People are storytellers who, individually and collectively, engage in experiencing, imagining, telling, retelling, re-experiencing, and re-imagining stories of their lived-in worlds (Conelly & Clandinin, 1990;Denscombe, 2010;Riessman, 2005). A story can be expressed through different media, such as written text and/or interviews (Denscombe, 2010). Stories can be treated as narratives when written or told with a particular purpose in mind (e.g., an account of personal experience), when drawing a link between the past and the present to reveal any developments or changes over time, or when including feelings and experiences emerging from social activities and interaction (Denscombe, 2010). From a narrative point of view, stories can be analyzed in terms of how individuals construct their personal or surrounding world.
Narrative inquiry or narrative analysis is increasingly used in educational research with the claim that "(…) education is the construction and reconstruction of personal and social stories; learners, teachers and researchers are storytellers and characters in their own and other's stories" (Conelly & Clandinin, 1990, p. 2). From a social research perspective, Riessman (2005) argues that narratives are not self-evident non-analyzable data, but they require interpretation. In the present study, although influenced by narrative analysis, we employed a systematic thematic analysis following the principles of content analysis, in order to identify common thematic elements across participants' experiences (Riessman, 2005). According to Riessman (2005, p. 6), "narratives do not mirror, they refract the past", meaning that storytellers do not just reproduce a past experience but they rethink their experience based on their current interpretation, interests and strategies of sense-making to themselves and others, while drawing parallels among past, present and future. Voice is a fundamental element of stories, since it is through voice that individuals are enabled to participate in a community and convey their meanings to others (Britzman, 2003). Voice aimed at the social process of understanding relationships between the individual, her/his experience and the other (Britzman, 2003).
While considering the importance and richness of participants' stories and voices as devices of capturing in-depth, non-observable participants' experiences of value creation in a CoLP, we invited the participants to write their own value creation stories after their participation in the community events. Therefore, these stories have a retrospective orientation with a direct focus on linking expected, experienced and realized values of the past, with applied values of the present, and potential and reframing values for the future. The following sections describe in more detail how participants' value creation stories were collected. Wenger et al. (2002) provide a framework for collecting community members' value creation stories within organizations and their importance to the members, the community and the organization within which the communities arise. In line with Denscombe (2010) and Riessman (2005), Wenger et al. (2002) state that the realization of values cannot derive from mere identifiable static measurements, but from stories that depict the complex relations between activities, resources and outcomes, while revealing the contextual aspects that frame those relations. The stories themselves -apart from providing evidence of community members' co-construction, exchange and application of gained knowledge -also foster a sharing culture through the visibility of one's practice within their context. According to Wenger et al. (2002), three main components should be incorporated in one's story to foster its systematicity in describing how community resources actually emerged and applied into practice creating value: (1) the initial activity, (2) the knowledge resource generated by this activity, and (3) the way the resource was applied to create value.
The first scaffold aims to capture the overall experience of participation and suggests various ways of talking about it (see Figure 3). It includes several stages of the experienced participation (rows) and several aspects of the participant's experienced values (columns). A variety of types and cycles of values can be identified from the overall personal value narrative, including the Expected values.

Typical cycles
Your story:

Activity:
Describe a meaningful activity you participated in and your experience of it.

Output:
Describe a specific resource this activity produced for you and why you thought it might be useful.

What we found: Analysis and results
Out of 22 students that were members of the CoLP, 18 wrote a value creation story. We conducted content analysis of these eighteen stories to identify their experienced values of community participation. A coding scheme was developed on the basis of the SMTV. The coding scheme thus included six cycles of values (expected, immediate, potential, applied, realized, and reframing), each with five types of values (personal, social, skill-related, studyrelated and context-related); in all 30 codes. We extracted 361 segments from the 18 stories, out of which 340 were codable and 21 were non-codable. Based on the segmentation procedure, a segment was considered as meaningful by the coders when indicative of members' attribution of positive, neutral, or negative oriented values to any aspect of his/her CoLP participation. Any statement whose meaning was not clear to the coders, or was not explicitly related to the CoLP participation, was considered non-codable (e.g., When I was asked what motivates me in general, I answered improving something with meaning for a system / society / community). The coding was performed by two independent coders (i.e., an external observer and the participatory researcher) who identified 21 out of 30 possible values with satisfactory reliability (Cohen's kappa = .72). The content analysis of the value creation stories (see Table 3) shows that a wide range of values could be identified by the coders (21 out of 30 values) with the most prominent ones being Realized skill-related values (24.1%), followed by Realized social values (15.3%), Realized context-related (14.1%), and Realized personal values (7.2%).
None of the coders identified nine of the theoretical value categories which were included in the coding scheme. Four codes were not identified despite the fact that they could have been identified and the remaining five were nonidentifiable due to the nature of the data source which had a retrospective perspective (i.e., immediate values were not possible to be identified). However, the immediate values were retained in the coding scheme for the analysis of the video data, which enables capturing the actual participation of members in the community events.
Although the SMTV enabled us to examine value creation in a CoLPnecessary for understanding of the phenomenon -it also fostered our reflection on the steps and approaches taken to unravel the value creation phenomenon. The following section addresses the main questions as part of methodological reflection and details our lessons learned.

What we have learned: Discussion
In order to examine value creation in a community setting we developed a Situated Multilevel Typology of Values (SMVT), with the concept of value and Wenger et al.'s (2011) value creation framework as our theoretical basis. The SMVT and its application to the analysis of self-reported value creation of members of a Community of Learning Practice (CoLP) prompted our reflection on the following questions: (1) To what extent can the values that the participants originally intended to report actually be identified as such by the researchers/analysts' without bias due to the researchers/analysts' own perspectives? and (2) To what extent does a pre-defined typology of values, based on a theoretical framework, confine or enrich the range of possible values that could be identified? Both questions are associated with the axiological issues of objectivism (e.g., Hartmann, 1967), subjectivism (e.g., Perry, 1954), and value contextualism (Frondizi, 1971), along with the ethnographic issues of realism and relativism with a main focus on the extent to which an outsiders' perspective can depict an insiders' perspective when the former use their own conceptual tools for discovering the latter (Descombe, 2010).
By adopting the contextualism approach to values as described by Frondizi (1971) -i.e. the uniqueness and high situatedness of participants' stories and therefore values reported and described through the stories -a researchers' re-production and interpretation of others' stories is a challenging process. Values are generated within specific situations and circumstances and do not necessarily need to be stable entities across situations (Frondizi, 1971). They can be intangible and therefore non-observable, need time to grow, are not always immediately realized by the participants within a specific situation, and even if they are realized it does not mean that these values can be automatically communicated to others because the personal value of an experience might be of relevance only to oneself. All these elements make the identification of values within the participants' stories even more challenging. While reading the story as told by a participant, researchers might partially reproduce or mis-produce participants' value creation stories due to the researchers' own perspectives, theoretical frameworks applied for interpretation, the situatedness of their interpretation and the constraints of what can be told and what cannot be told (Britzman, 2003, p. 35).
While considering the situated nature of value creation, inviting researchers to observe the context and participants' interactional patterns and value generation through video observation -prior to the analysis of value creation stories -might allow the researchers to understand better the participants' perspectives towards values and the ways they would convey them to others. The closer to the phenomenon under study the less filters -that act as obstacles to the observation and analysis of participants' experienced realityare imposed by the researchers.
EO V PR P Figure 5. Pandora's box as a metaphor for the "filters" that might hinder or direct the identification of experienced values in communities (EO = external observer, PR = participatory researcher, P = participant, and V = value as experienced).
The metaphor of Pandora's box (see Figure 5) illustrates that the seemingly simple action of analyzing values experienced by participants in a community is instead a highly complex process that may lead to an endless complication of unravelling the "real". Analyzing or reporting experiences related to constructs such as values, which are highly dependent on people's own set of criteria (Rokeach, 1979), includes multiple "filters" that may bias the participatory researcher's and external observer's interpretation of participants' experienced reality. More specifically, such a bias of the values told or read in the stories is effected by story tellers' (i.e., participants) and story readers' (i.e., participatory researchers or external observers) own value systems, and in the case of story readers also their understanding of pre-defined typologies of value creation. In sum, the filters represent added layers of subjective interpretation. As such, any idea of an "absolute" or non-selective truth" reported by the storytellers or understood by the story readers should be discarded. A participant's (P) own value system filters the reported values (V), which in turn are even further filtered by the participatory researcher's (PR) value system and any biases deriving from the PR's community participation. The PR's lens is reflected in and shaped the development of the SMTV, which further filters what can be observed by an external observer (EO). In turn, the EO's understanding of the SMTV may further filter the reported "reality". Our reflection with the help of the metaphor of Pandora's box emphasizes the complex interplay between what the story was, what the story tells, how the story is read and what any story reader further "story tells" that can lead to bias or misinterpretation of the experienced phenomenon.
In our study, we involved two coders in the content analysis of the value creation stories. Although these coders achieved satisfactory reliability, they simultaneously differed in their personal perspectives towards values and their interpretations of the stories as told by the participants, which in turn affects the identification and interpretation of values. In the present study the first coder was the participatory researcher (i.e., the non-peer facilitator) and the second coder was an external observer. Although external, the observer was invited to watch the video data of the actual community experience before being involved in the analysis of the value creation stories in order to familiarize herself with the participants, their observed attitudes towards their participation, and their role in the community. Doing so might have brought the observer closer to the phenomenon.
Regarding the theoretical framework by Wenger et al. (2011) that informed the template and both storytelling scaffolds, we think that the template (and scaffolds) both confined and enriched the possible range of values reported by the participants. On the one hand the scaffolds facilitated the narrative storytelling process by serving as stepping stones, but on the other hand they might have directed the participants to attribute values to aspects that they normally would not have attributed value to. However, when invited to tell/ write their experienced stories retrospectively, participants might not be able to recall important aspects if scaffolds are not available. Yet, with respect to the predefined SMTV there is a danger of classifying values too broadly or too holistically and thus and thus losing the particularities of the experienced value. However, the typology provides researchers a framework to identify actual written elements that imply attributed value by the participants. Nevertheless, it should be highlighted that it is still unknown whether the participants would have perceived the classification of their reported values in similar ways to the participatory researcher and the external observer. Hence, further analysis could be done by involving the participants themselves, for example with the help of interviews around the analyzed stories and/or cued retrospective recall of video-fragments to foster participants' reflection.
Our reflections on the examination of the under-researched phenomenon of value creation in communities signify that the analysis of value is highly situated and that any analysis also needs to consider the degree to which the outcome of such an analysis (story-reading) adequately represents participants' experiences (story-telling) in relation to the filters that colour the researchers' interpretation of the value creation stories. It also serves to inform researchers who study the value creation process in face-to-face or online communities (whether CoLP, CoP or CoL) about the complexity of its analysis, but that the SMTV is a first methodological stepping stone to address this thorny issue.