The Affordances of the Medium: A Comparative Mixed-Method Study of an On-Site and Online Shared Reading Group for People with Cancer

Organisation of the reading groups

The online and on-site SR group (with different participants in the groups) met weekly for two hours for 12 weeks, which was extended with an additional four weeks for the participants who wished to continue. In the extension period, three new participants joined the on-site group. They met from September 2021 to January 2022, partly during the COVID-19 pandemic lockdown in Norway. The physical meetings took place at a local cancer organisation. The online meetings were hosted by a hospital library using the secured video platform Whereby. The researcher took part in both groups as an observing participant.

The reading groups were organised in line with the program developed by The Reader (The Reader, n.d.), in which participants read and discuss one prose text and one poem together. In the on-site group, the participants received a paper copy in hand upon arrival at the venue, whereas in the online group, the RL shared the text on the screen, and the text was read on the spot. None of the texts were known to the participants before the session. Hence, there was a low threshold for participation since they did not require any preparation at home. In addition, the participants were not obligated to speak or engage in the dialogue if they did not choose to; in line with the SR method, they could just come and listen. During the reading, the RL paused at particular places in the text for exploration and invited an open discussion of the section read, focusing on the participants’ experiences. The online RL encouraged the online participants to be unmuted and have their cameras on but also communicated that it was up to them to decide. Participants mainly had chosen to have their cameras on and were unmuted when possible. However, technical issues with echo (when multiple people were speaking) made it sometimes necessary to mute. Additionally, some participants had their cameras off while listening to the reading aloud and turned them on again during the discussion.

Participants

A total of 12 women from 22 to 69 years old (M = 51) consented to participate in the study (on-site group: N = 8, online group: N = 4, see Table 1). Over the period the reading group met, they worked part-time, were on sick leave, or had retired. The participants found themselves in various stages of cancer treatment: some were undergoing chemotherapy or other types of treatment; others had just completed their treatment and were coping with the associated side effects, and some were reliant on life-prolonging medication. Notably, the participants in the online group faced more challenging prognoses or had been recently diagnosed. This circumstance underscored the significance of the online format, as it accommodated their inability to attend physical reading groups due to their health conditions, limited energy and thus compromised mobility.

Participants were recruited via flyers and social media in collaboration with cancer organisations, coordinators, and cancer nurses in Norway. The selection criteria were adults (18+) with a cancer diagnosis who could understand Norwegian. We aimed for broad recruitment to be inclusive in our sample, avoiding specific diagnosis requirements due to SR’s novelty and the pandemic, and these circumstances made the recruitment more challenging. Although the participants did not necessarily share a common diagnosis, there was a mutuality when it came to the experience of treatment procedures, hospital meetings, and the experience and emotional impact of receiving a cancer diagnosis, regardless of its specific nature. Two participants withdrew during the study, one from each group (see Appendix 2 with background information on participants and table of session attendance rate), because of limited time and energy.

Reader Leaders

Both groups were facilitated by librarians with RL certification. There were in total three RL’s: on-site RL facilitating session 1–12, another on-site RL from session 12–16, and one online RL facilitating session 1–16. Since the two groups had different RLs, a key limitation of the study is the added difficulty in comparing them and isolating the influence of medium affordances. In addition, their prior experience with SR also varied: the online RL had been running groups for several years and training other RL’s in Norway. The two others were novice RL’s and had one year experience with one or two prior groups. Since the online RL had extensive experience with on-site SR sessions, it enabled her to share insights of how the medium might have impacted her facilitation. She had, for example, previously facilitated on-site SR groups for patients at the hospital library.

Texts

The texts read in SR are usually one short story, or an extract from a novel, followed by one poem. The RL selected texts for the sessions that personally engaged them since, in their experience, personal engagement was necessary for the RLs to facilitate engagement in the sessions. The texts were not about cancer, as the researcher and the RLs wanted to give the participants some distance from their illness. However, it did not hinder participants from reading their illness experiences into the texts we read, and importantly, they made these links themselves. The texts read in the two groups were not the same (see Appendix 1 for the text selection).

Data collection

The data consists of 31 audio-recorded sessions, where 17 sessions got transcribed (online: 8, on-site: 9), field notes from participant observation, focus group interviews with both groups, an individual interview with the participant from the on-site group who also participated in four sessions of the online group, and an interview with the online RL.

The first author conducted participant observation during 15 of 16 sessions in both groups to gain firsthand and embodied insight into the group dynamics and the SR practice. This continued presence was essential to establish trust between the participants and the researcher. The aim of including focus groups was to know more about how the participants experienced the different components of the reading group (the RL, the group discussions, the group, reading aloud, and the texts). In the online group, some of the questions addressed their experience of the online setting. However, since the participants in the online group did not have any experience with on-site SR from before, they could not compare their experience. The participants’ lack of awareness of the potential impact of the medium might be a ‘blind spot’ in the data as they, from the beginning, may have become used to, for instance, reading the text on the screen. Therefore, we decided to invite the participants in the on-site group to join the online group for the last two sessions, resulting in one participant from the on-site group, Elena (fictious name), attending two online sessions that were recorded and two additional sessions after the study had ended since the group continued. Elena is the only participant in our sample with experience from on-site and online SR sessions. However, the researcher had observations from both groups and could compare these. In addition, the online RL had extensive experience facilitating on-site SR groups which she could compare to her experience in the online group.

Focus groups versus individual interviews

An advantage of focus group interviews is that the group setting enables participants to express a wide range of views, creating a stimulating discussion where ideas are built on or challenged (Litosseliti, 2003). However, participants may also be influenced by the opinions of others and, as a consequence, suppress conflicting ideas in the wish to stay aligned with the dominant opinions. Moreover, the group setting does not leave room for in-depth exploration of individual perspectives. These dynamics might have influenced the focus group interview data. Nevertheless, familiarity among participants from prior sessions may have encouraged freer sharing of experiences. In contrast, individual interviews allow for more personal experiences and in-depth answers. While only one individual participant interview informed the role of the medium, a participant from the online group interviewed for another purpose briefly noted the online format made her feel “modern” (an elderly participant) and expressed gratitude for the opportunity to participate. Her other responses, focused on reading experiences, were detailed and personal, suggesting openness in the interview context. The limited response regarding the medium might reflect a lack of participants’ prior consideration of its influence.

Transcriptions of reading sessions

The sessions were transcribed verbatim by a university student who was a native Norwegian speaker. They were transcribed systematically using a transcription key (see Table 2). Since there was a large amount of data, and we aimed to look for overall structures to compare the transcripts, the transcription key mainly focused on turn-taking, emotional responses, interruptions, and silences in the dialogues and did not include a detailed conversation analysis with information on, for example, emphasis, pronunciation, speed, volume, and pitch.

We were especially interested in identifying longer silences because the first author observed that silences occurring in the group discussions, especially in the on-site group, could become quite protracted (up to 46 sec.). Thus, in the transcripts, silences with a length of over 5 seconds are marked, including the length of the silence. The shorter silences (…) we did not explore in this study since the transcripts did not indicate any noticeable differences in the two groups. However, this could be further explored in future research if the occurrence of unfinished sentences, identified as “hangings” (Davis et al., 2016), is equally present in on-site and online groups.

Data analysis

We extracted quantitative data from a total of 17 session transcripts (see Appendix 3 for a table of the data or Table 3 under “Findings”). These comprised nine session transcripts from the on-site group, including six transcripts which covered both the prose and poetry parts of a session, two transcripts containing only the prose part, and one transcript containing only the poem part. From the online group, there are eight session transcripts, of which five cover both the prose and poetry part, and the remaining three only cover the poem part. The reason for this selection is that the sessions had been transcribed for another study that investigated transformative reading experiences (Andersen & Hakemulder, 2024); therefore, these were the session transcripts we had available.

Second, we analysed the data using descriptive statistics applied to Conversation Analysis (CA) and qualitative methods. CA is a methodological framework and a theoretical perspective that systematically examines the structure and organisation of talk and written communication in everyday social interactions (Sacks & Jefferson, 1992). Third, we further explored differences and similarities indicated by the quantitative data qualitatively to get an idea of recurrent patterns in the data. Thus, the study is a mixed method design specifically using a triangulation design (Creswell & Plano Clark, 2006) with the aim “to obtain different but complementary data on the same topic” (Morse, 1991, p. 122), hence allowing us to combine the strengths of both quantitative and qualitative methods to compare on-site and online SR sessions in a complementary manner. The quantitative and qualitative data converged at different points. When analysing the data, we explored quantitative data qualitatively, and in the results, we brought quantitative and qualitative results together for a more nuanced understanding. An example of convergence in the data analysis would be that the quantitative data on silences did not differ much in duration and number. However, an in-depth, qualitative exploration of the individual silences in their context revealed differences related to, e.g., what caused this or that particular silence, what function it seemed to serve, and who was speaking before and after the particular instance of silence.

The importance of the RL’s facilitation in an SR session (Steenberg et al., 2021) motivated a closer look at the role of the RL and how it may be differently enacted in an on-site and online session. As mentioned above, we recognize the limitation of using different RLs, as their facilitation styles and preferences may vary. For example, during the participatory observation, the researcher noticed that the online RL spoke more in the sessions than the on-site RL did. This motivated a close examination of the pattern of turn-taking and the occurrence of periods of silence in the two types of sessions. Identifying the percentage coverage of each speaker in each session was performed in the coding software NVivo by attributing a code to each speaker in a transcript and creating a chart (see generated charts in appendix 4) to visualise how much of the transcribed dialogues in a session is coded into the various speakers (the participants, the researcher, and the RL). For that reason, the reading-aloud part of the prose text was not transcribed since we were interested in the speaking percentage of only the group discussions. The poems were included in the transcription because having them in the transcripts was practical, and all of them were short.

In general, the quantitative data are estimates. For the interruptions and silences, we have subtracted the ones occurring outside the group discussions (so not including the ones happening during ‘chit-chat’ in the breaks or at the beginning and end) as we were mainly interested in the conversation structures when discussing the text being read. There are limitations to this design since there is not necessarily a clear cut between breaks and group discussions. For emotional and non-verbal responses, we have not subtracted any parts in the transcriptions as we were interested in getting an insight into the overall group dynamics.

Next, the observations from the session transcripts are triangulated with the field notes from the researcher’s observations, interview data from the participants in the online and on-site sessions, and an interview with the online RL. Altogether, the various types of data supplement each other to shed further light on what happened in the sessions and how the medium’s affordances may have impacted aspects of the participants’ and the RL’s experience differently.

Ethics

This study was reviewed and approved by the Norwegian Research Ethics Review Board and the Regional Competence Center for Clinical Research. In an online setting additional support could be necessary, as participants may face increased vulnerability when being alone. In addition, facilitators may encounter difficulties in managing discussions on sensitive topics. Therefore, it was important for the study that the online RL had prior experience with leading groups for patients. To provide extra support for the online participants, she offered to remain available after the sessions in case the participants felt uncomfortable or sad. The names used in this article for referring to participants are fictitious.

Physical vs. virtual venue

A fundamental and obvious difference between meeting other people in real life and meeting them online is that between virtual (or digitally mediated) and physical co-presence. In the on-site group, the participants left their homes and came physically to the location of the cancer organisation. They met in a shared physical space with full access to the physical and multisensory presence of the other people in the room. In the focus groups, the on-site participants appreciated the possibility of getting out of the house, meeting, and socialising with others to get a rhythm and new input in their daily lives. The appreciation stemmed from being on sick leave, which involved spending a lot of time at home, making the days feel long and monotonous. In the online group, they participated in a shared digital space from the comfort of their current location, and the participants appreciated the high flexibility this offered them — the participants connected from their homes, an office, in transport, or at the hospital. Theoretically, they could join from any place with an internet connection and a device for connecting which was beneficial for the online participants as they had fluctuating health, many medical appointments and were afraid of being infected by COVID-19.

The shared digital space was constructed from the participants’ individual physical spaces, often a part of the living room, sofa area, or bed, which the video, if it was on, showed partly. Thus, the video provided a window into the participants’ personal and individual spaces, which adds a sense of intimacy to online SR that might differ from when people meet in a public place. Occasionally, the participants’ individual spaces may interfere with the shared digital space: a plate may fall on the floor, creating noise and catching the others’ attention, or the doorbell or telephone rings, and others, including pets, may enter the room inadvertently, and be caught on camera.

The participants’ surroundings can also influence the atmosphere in the group. Rachel, for example, sometimes connected from the hospital while she was receiving chemotherapy. The view into the participants’ private spaces created a feeling of an intimate presence or moment that also opened for emotional and personal sharings:

Nevertheless, the participants only had two-dimensional, audiovisual access to the shared digital space. Due to muting and the fixed camera angle, the access to the other participants’ voices and body language was restricted. Thus, the difference in co-presence may have implications for each participant’s mode of involvement, as well as for the ways in which the participants are present and available for each other.

Displaying and accessing text on paper vs. on screen: Autonomy and engagement

This section explores the differences between presenting and reading the texts on screen and paper. Most importantly, these differences pertain to the participants’ access to the text during the reading and to the RL’s options for sharing the text with the participants. In the on-site group, the participants had a paper copy of the text. They could physically engage with the text as they wished: turn the pages, write notes, or draw doodles on it, underline sentences, or move it closer to their eyes when looking at a particular word or sentence. During the reading aloud, the participants mainly looked down at their text copies to follow. However, from time to time, they would move their gaze to their surroundings, gaze out of the window, or place the text on the table and close their eyes to enjoy being read to. Hence, having and holding the text on paper may have given the participants a sense of autonomy in the reading so that they could freely navigate the text as they wished. They could more easily skim parts of the text during the group discussion, and often, they would take on the role of a co-facilitator, bringing a passage, sentence, or word in the text to the group’s attention by reading it aloud.

In the online group, the RL shared the text using the share-screen option in the Whereby meeting platform. Hence, the online participants were required to follow the text at the pace presented, as they could not access it themselves. When the participants in the online group looked at the text, their gaze was directed at the display of the text on the screen, appearing next to the display of the other participants’ faces (unless they had turned off the camera), possibly creating a feeling of joint attention (Gustafsson et al., 2022). Rachel said in the focus group that she mostly closed her eyes and listened to the RL’s reading aloud of the text, and if she then fell out of the story, she had the text on a screen as a support.

The main difference between the online and the on-site group was that the participants could not physically engage with the text. Unlike the on-site group, the online RL was totally “in charge”, so to speak, of the pace with which the reading unfolded. If, at some point during the reading, any of the participants felt the need to return to a specific location in the text, to slow down, or to reread sections of the text, they had to ask the RL explicitly to do this. For Elena, the lack of physical engagement triggered a sense of distance to the text. She felt that viewing the text on the screen was confusing and it made it more challenging to navigate:

Elena’s comparative observation of the differences between an online and on-site session reflects a feeling of compromised autonomy that may have been an issue in the online setting. In addition, the differences in access to the text may have implications for the social and emotional dynamic and exchange during the dialogues in SR. The co-presence online is two-dimensional and disembodied, lacking the multisensory and embodied co-presence of other people sharing the same space in real-time.

Facilitation and role of the RL

Overall, looking at how the RLs facilitated the text and the discussions in the groups, some interesting differences emerged. Concretely, the on-site RL took on the role of an initiator of the discussions and co-participant in the group, involving a less pronounced reader leader role. This way of facilitating led to more room for the participants’ thoughts and ideas, which made the reading of the text and the discussions more based on the input from the participants and less explicitly guided by the RL. In comparison, the online RL took on more of the role of a mediator of the text and constant moderator of the discussion, where she was more present in leading the discussion and mediating the text’s meaning to the group. In this group, the participants did not readily engage in the discussions at the beginning, which caused the RL, as she explained in the interview, a more extensive and explicit effort to get the conversation going. In her impression, the medium’s affordances required her to be more present during the discussion. This is reflected in the session transcripts, where we can see she responds to all the participants’ input and helps connect these. Thus, the reading of the text became more explicitly guided. In the following sections, we will look closer at facilitation practices in various stages.

Participation and group dynamics

Beyond potentially affecting facilitation, the online medium may also impact participants’ distribution of turns and the way they engage with each other verbally and non-verbally.

A first general observation is that participation appeared to be more naturally flowing in the on-site group, whereas it seemed more inhibited in the online group. Illustrative of such a difference, Fagerlid and Vannes (2022) have described the conversation in an online setting by using the word ‘lugger’; it may abruptly halt, and it does not flow as smoothly as in physical settings. This can be related to the challenges in managing turn-takings when muting and unmuting; the participants were careful not to interrupt each other, which is probably why there were very few interruptions online compared to in the on-site group. A consequence might be that not all spontaneous responses were shared in the room. Elena, who had participated in both the on-site and online groups, experienced that something was holding her back when participating in the online group compared to the on-site group: “But there’s something about it, you get that distance when you’re online. It’s unavoidable that you lose a bit of… So, in that sense, you mute, and you’re supposed to talk, and then you talk over someone. It becomes a bit different then” (interview, 26:58).

In contrast, consider Alice’s experience in the on-site group, where she emphasised spontaneity as a key dynamic: “Like, suddenly something comes from someone and then it starts a thinking-process in us, or feelings, and then we talk about it…” (on-site focus group, 89:38). The researcher observed that participants did not share these spontaneous reactions to the same degree or in the same way online, and this observation is supported when looking at the structure of dialogues in the session transcripts. Thus, the notion of “liveness” (Longden et al., 2015) might be more restrained in online SR.

Social space

The differences between the physical co-presence on-site and the digitally mediated co-presence online also impact the social dimension in SR. In the on-site group, there were more opportunities for social interactions. At the beginning of the session, there was a bit of ‘chit-chat’ with coffee and a snack before we started. Sometimes, a participant shared if they had read something or googled the author we read last time. In the middle, we had a five-minute coffee break where there was room for more casual talks, or sometimes, they kept discussing the text we were reading. In the end, the participants left the room more or less together.

In the online group, they arrived and left the group individually. There was a coffee break in the middle, but the participants went separately to get their beverages and were muted until the RL resumed the session. Thus, the social space seems to have been ‘narrower’ online. For instance, Elena experienced the online group as a safe but not a social space. Her experience was that the limited options for social interactions led to less sharing, especially outside the readings and discussions. In the on-site group, they talked about their day and doings, frustrations with the health care system, or dilemmas from their relations or family. They also shared various tips, such as book or music recommendations or hiking routes. The social breaks and the detours in the discussions were essential to create a group feeling, possibly leading to a broader discussion that included more diverse perspectives on the texts.

Group cohesion in a small and unstable group

When patients in demanding life situations are the data source, it may lead to inconsistent attendance, affecting data collection stability (Buckwalter et al., 2009). In this study, the online group consisted of only 4 participants, and the group had little stability in attendance (on average, there were 2 participants present apart from the RL and the researcher). The low number of participants in the online group reflects the vulnerable health status of cancer patients and the real-world conditions under which such research is conducted. Therefore, it is also valuable to have a study with a small and unstable group, as this would often be the case, and to explore if the participants still experience it as an important psychosocial support despite these obstacles. Although there were fewer participants than in the on-site group, and the online participants seemed to have interacted less with each other, the participants still highlighted the experience of multiple perspectives on the texts and how they were surprised by other ways of thinking and understanding the text as valuable. In the focus group, Rachel identified, and the other participants agreed with her, the experience of contributing to and being part of the group as the most beneficial experience to her.

However, questions also arise from studying a group of only 1–4 participants each session: How often should they meet, and how many participants should be the minimum for the group factor to be in play or to have a sense of community? When considering the results from the present study, the positive reported experiences related to contribution and being part of something in the online group could suggest that in very small groups and with fluctuating attendance rates, there can still be a group factor and sense of community in play if those meetings are profound and a good experience. However, if there is only one participant in a session, the perspectives presented in the discussions of the texts might be limited to fewer perspectives and consequently the discussion risks becoming less enriched. In these cases, the discussions might have been richer if more participants with different life experiences and points of view led the discussion in diverse directions. On the other hand, it might feel safer for the participants to speak up in small groups, and this feeling could potentially create more group cohesion, especially online, where turn-taking can be a challenge.

In a systematic review of telehealth interventions such as online group therapy, Gentry et al. (2018) found that in 40 reviewed studies, the participants had positive perceptions of the telehealth experience. Moreover, several qualitative studies (O’Connell et al., 2014; Taylor et al., 2012; Zhou et al., 2016) that examined group processes and how they were affected by the online delivery, found that participants had reported experiencing a positive sense of group belonging, wanting to continue the group sessions, feeling valued by group members and less alone (Gentry et al., 2019, p. 13). While some studies indicated slight differences between face-to-face and videoconferencing groups, such as minor decreases in therapeutic alliance (the therapeutic collaboration between therapist and group members) and group cohesion, two studies found that participants preferred videoconferencing over in-person interaction, citing it as less distracting and threatening (Khatri et al., 2014; Tsaousides et al., 2014). It might be that for some cancer patients, as Alice, the more focused atmosphere in the online group would be better as there are fewer social ‘distractions’.

Implications and future research

The comparison of the two groups led, maybe unsurprisingly, to the discovery of similarities and differences. Different pathways led to different though similarly valuable benefits. This seemingly straightforward realisation holds potential significance for our target audience, cancer patients, given their often limited mobility, energy, time and confinement to homes or even hospital beds, and their challenges in focusing on stimuli like literary texts, and therefore restricting their access to experiences that may help them cope with their condition. Furthermore, online SR might also have the potential to be a distraction and stimulus while patients are receiving chemotherapy (which can take hours) or for in-ward patients. It could also be further explored if a SR offer that combines online and on-site sessions would have better outcomes and be more accessible to cancer patients. It could be, for example, that group cohesion would be strengthened if the first session was on-site. However, there is the risk that the online version will be seen as less attractive when the on-site experience is fresh in mind.

Surprisingly, the online participants were not very aware of the medium’s impact on their participation, and they did not report any negative experiences, which may seem in contrast to the online RL’s more critical position of how the online group ‘worked’. This could be due to the focus group setting in which these more critical perspectives may not have been expressed. However, they seemed to have focused on how well it worked and accepted the conditions (muting and changing sound quality, reading on screen). They were grateful for the opportunity to participate and had a strong need for such a low-threshold, flexible and meaningful activity in their illness process. Their engagement in the readings and the feeling of safe space might have led them to ignore these technological problems rather than become frustrated because of them. Hence, we would argue that online SR could be of major importance, considering the vast population of people suffering from cancer. Of course, given the limitations of online SR, that is, the virtual and hence restricted ‘social space’ and the less prominent sense of community, on-site meetings would be preferable for patients who have a strong desire to get out of the house and meet other people. However, online SR may still help reduce isolation by connecting individuals in the same situation.

The study has led us to new questions, especially relating to the sense of community and whether it is experienced differently online than when people meet in person. The findings related to a sense of community indicated that the experience of contribution was more prominent than collaboration in the online group. However, it would be beneficial for the present study to investigate further the participants’ experiences of community online versus on-site and explore whether the online experiences are of a different experiential quality than on-site. We suggest this could be done in a self-report within-subjects study design with two conditions (on-site SR and online SR), and include at least two groups, where one group of participants first experience the on-site and then switch to online, and the other group starts with online SR and then switches to on-site.

The data also indicated a more prominent and guiding role of the RL. Implementing online SR for our target group (and probably also for others), would require thorough considerations in relation to: monitoring the group (because of the lower visibility of individual responses) to adequately respond to participants’ input; selecting effective texts suitable for online use; and being aware of potentially emotional triggers and the need for follow-up support. For the online group, it seemed especially important that the RL responded to the participants’ input in a warm, open, and caring way. Thus, the RL did not only seem to have a significant role in getting the discussion going and mediating the discussion but also in making the participants feel safe by responding positively to every one of their inputs. Hence, they felt acknowledged and important in the group. However, the individual facilitator style of RL’s needs further investigation, as it is not possible to determine in this study if the more guiding role of the online RL is an expression of her facilitation style or the affordances of the medium. The safe space seemed to be there from day one, which opened for intimate moments in both conditions, but especially in the online group. On the one hand, Elena, who participated in both groups, noted a lack of social and dialogical dimensions in the online situation. On the other hand, the intimacy in the online group came from the ‘window’ into their personal spaces, such as Rachel, who logged in from the hospital and needed room to share how she felt, which affected the group atmosphere. These findings inform potential difficulties for the online RL in reading the participants’ nonverbal signals and managing situations with little or no response. These challenging situations for the RL, which are further complicated by the health context, could be further investigated in a study that collects and explores RL’s perceptions and experiences of online SR in a survey study to shed light on other potential obstacles and how to prepare future RLs best possible for online facilitation.

The on-site group had some advantages over the online situation. Some are inevitable. However, small adjustments may mitigate others. For instance, the RL could also stimulate participants to print the text in advance and read it from paper (but without reading it before the session), which we thought was an advantage in the on-site group. Second, the RL could incorporate more ‘chit chat’ in the beginning, during the coffee break and before logging off to stimulate more group cohesion. However, it might be that an extended social space in online SR would not function because of the medium’s restrictions on dialogue and presence. Perhaps fewer social interactions and distractions are a relief in this setting and for this cognitively challenged target group, which allows for a deepened connection between the individual and the text. However, no clear recommendations for online SR with cancer patients can be made, but based on indications from this study we encourage future research to explore further the function and outcomes of online SR. There are several limitations to this study, which have been discussed throughout the article. This study could benefit from future research in on-site and online SR using the same RL reading the same texts with the same participants in both online and on-site groups over some time, in a quasi-experimental setting, with texts and medium counterbalanced. Furthermore, it would be necessary to apply conversation analysis on a larger number of session transcripts to see if the recurrent patterns observed in this small study pertain to a larger text corpus.