education · social · digital
in the room, out of earshot
University classes that students can join in person or remotely leave remote students "present yet peripheral": in a New Zealand study they missed classmates' questions, mixed in-room and remote groups struggled to hear one another, and in the largest observed session no student spoke to another
Source: See, “Together yet apart: Understanding hybrid learning through students…” Open ↗source: primaryLast updated: 2026-11-17
Problem statement
Many universities now run "hybrid" classes, in which some students sit in the room and others join the same class live over video. The format is meant to widen access (for students who are ill, far away, working or caring for someone) without cutting them off from classmates. A 2026 doctoral thesis at a New Zealand research-intensive university (one researcher's PhD, not a journal paper), combining student and educator interviews with network analysis of recorded sessions, finds that remote students are "present yet peripheral." With one microphone at the lecturer's podium they heard the lecturer's answer but not the in-room student's question ("It's just quiet. And then you hear an answer"). When lecturers mixed in-room and remote students in one group, an in-room student without headphones described putting an ear to the speaker, and remote students heard several groups' conversations at once. In five recorded sessions of a hybrid maths workshop, from the third session on every exchange ran through the tutor, and the largest session (7 in the room, 7 remote) showed no student-to-student exchanges at all, among in-room students as well as across modes. The audio half of this is not one thesis's finding: a 2020 systematic review of 47 studies calls audio "the biggest challenge" in synchronous hybrid learning, and a 2026 HyFlex review of 30 studies reports that remote students who "have difficulty hearing in-class discussions" feel "less engaged and more like observers." The unsolved problem is letting remote and in-room students actually talk to each other, not just to the teacher, in rooms built for one kind of class, at the moment the class splits into several groups talking at once.
Why this matters
Hybrid attendance is often the route by which students with illness, jobs or long commutes stay in a class at all, and the thesis ties remote students' engagement to their visibility, chances to interact and sense of being near peers, and so to their sense of belonging. The students most in need of the remote option are the ones left at the edge of the room's social life. The thesis also finds that when in-room students speak softly to a nearby tutor who does not repeat them, those contributions become inaudible to the remote half.
What’s been tried and why it hasn’t worked
(1) Hybrid rooms with a lectern microphone and camera. Students reported lecturers walking out of microphone and camera range, side-on views, and in-room questions not captured; the thesis records a lecturer "constantly reminded to be in one place so that we can hear him." (2) Breakout rooms and chat groups. Some lecturers put students into online groups; students said that with strangers "no one will talk," cameras and microphones stay off, and the outcome depends on who you are matched with. A shared online document worked better for one student. (3) Mixed in-room and remote groups. Without planning, in-room students lacked headphones or their own laptops, and remote students could not pick out their own group's talk from the room; one student said "when you're attending remotely, it just doesn't work" for this kind of group. (4) Better audio around the room. The thesis's literature review reports educators seeing drop microphones placed around halls as the way to capture each group's discussion, and its model names visibility, audibility and connectivity as the conditions remote participation needs. (5) Whole-room microphone systems sold for hybrid rooms. Ceiling beamforming arrays (Sennheiser TeamConnect Ceiling 2, Shure MXA910), throwable microphones (Catchbox) and 360-degree camera-speakerphones (Meeting Owl) are sold to let remote students hear in-room discussion. Singapore Management University fitted 110 ceiling arrays in 35 rooms because "online students were having trouble hearing in-class discussions" (vendor case study, no outcome data). The 2020 review reports that one omnidirectional speakerphone serves eight or fewer on-site students and that "larger class sizes will necessitate different approaches." These systems send one room mix to the call: they suit one conversation at a time, and a throwable mic carries one speaker. A competing vendor (Catchbox, not an independent source) says ceiling arrays struggle to separate talkers, so "the audio can get messy when many students talk at once." No independent test of any of them during simultaneous group work was found; the nearest is a test in a company's classroom lab (Tudela et al. 2025; three of its five authors are employees of Quietam Studio, the company that opened the lab), which does not name its devices and had 10 people in all, with each two-group set-up run once. When two hands-free calls ran at once in one room, the authors report that overall sound quality as rated in that room fell from "excellent" to "fair" (an average of 5 down to 3.78 on a five-point scale), and they write that the loss "seemed to always be more acute for the distant person". ANSI/ASA S12.60, the US classroom-acoustics standard, sets in-room noise and reverberation limits; the Acoustical Society of America's companion booklet to Parts 1 and 2 of the standard says the standard "does not apply to" teleconferencing classrooms, among other room types, and a revised Part 1 approved in 2026 was not read. (6) A telepresence robot for the remote student (Finland). In a university language class where one student joined through a remote-controlled telepresence robot, initial trouble operating it led the class to stop using its remote mobility and carry it from place to place instead (Jakonen and Rusk 2025). Positive findings stand: one student said turning the camera on made them feel "truly in the class," tutors who deliberately and consistently drew remote students in reduced the gaps in interaction, and remote students who spoke were often clearer to everyone than in-room students. The 2026 HyFlex review finds engagement can be high where instructors repeat in-class questions and run shared activities across modes, and the thesis author's own 2026 systematic review (76 articles) concludes that students' "dispositions significantly impact student engagement more than peer interactions." Caveat: the network analysis covers five 50-minute sessions of one voluntary workshop, the thesis reports that the mix of remote and in-room students "appears to have little influence" on interactions there (though the same paragraph ends by saying the proportion "plays a role in how interactions take place"), and in-room students in the largest session did not talk to each other either, so the zero figure shows a tutor-centred session, not remote exclusion alone.
What would unlock progress
Audio and seating set up for groups, not for a lecturer: each mixed group needs a shared way to hear and be heard that does not pick up the next table, and in-room students need it without bringing their own headphones. This is the part the sold systems leave open: ceiling arrays, throwable mics and room speakerphones serve one whole-room conversation, not several mixed groups talking at once. The thesis's three conditions (visibility, audibility, connectivity) give a checklist a design can be tested against. Because every exchange in the last three observed sessions ran through the tutor, a design that lets students talk across the room without routing through the tutor would test whether the problem is equipment or habit.
How we’d know it’s being solved
(a) Student-to-student exchanges per session between remote and in-room students, counted from recordings as the thesis did (zero in the largest observed session); (b) share of in-room questions intelligible on the remote feed; (c) remote students' reported sense of belonging, asked before and after; (d) set-up time before a mixed-group activity. A class that "ran hybrid" is a proxy; remote students talking with classmates is the outcome.
Where to start
Facility-free door (audit a room): with two laptops, record a mock tutorial in any campus seminar room, one person attending remotely, and score what the remote attendee can hear and see against the thesis's visibility, audibility and connectivity conditions: in-room questions, group talk at each table, the board. Repeat with the speaker moved and with the "repeat the question" rule. Team members play both sides, so no one outside the team is recorded. No partner or special equipment is needed. Make door: build a table-top group pod for one mixed group: a speaker and microphone pointed at one table, a small screen for the remote member at seated eye level, and a way to keep the next table's talk out. This is not a new idea: educators in the thesis's own educator case study named a microphone in each group pod, set up so that it "only captures that pod," as their ideal set-up. The thesis reports this as a wish; it does not report such a pod being built or tested, or say how the microphone would keep out the other pods. That is the open part the team tests. Test it with student volunteers in a busy campus room against the shared-speaker set-up the thesis describes, measuring how many words the remote member catches and how often each side speaks. What exists: lectern microphones and the drop microphones around halls described in the thesis's review, the per-pod microphone that educators in the thesis named as their ideal (described, not built), and remote-controlled telepresence robots that one Finnish class ended up carrying, plus whole-room ceiling arrays, throwable microphones and 360-degree camera-speakerphones, all built to send one room conversation to the call. The pod differs from the sold systems by serving one group at one table while the next table talks; the team should first test a stock USB conference speakerphone at the table as the cheapest baseline, since the pod must beat it on keeping out the neighbouring table. Users are students and tutors in hybrid classes. A team reaches them through its own department's hybrid tutorials, a tutor who runs mixed groups, or the campus office that supports hybrid teaching; volunteers taking the remote seat can join from wherever they are. Volunteers who are recorded or counted sign up for it first: tell them what is recorded and why, and let anyone opt out. A mock session with volunteers is an informal design test; recording a real class needs ethics review first. Design-the-test door: with a teaching unit that runs hybrid tutorials, design (do not run) a comparison of the pod and current set-ups using recorded-session network counts. Recording a class needs ethics review and each student's consent; the thesis had ethics approval, told students at registration that sessions would be recorded, and had the tutor ask consent at the start of each session, with a way to withdraw. Relevant skills: acoustics, product and interaction design, education.
Genes
Related problems
Stuck the same way, in other fields.
field: water & energyinvisible water, no treaty468 aquifers cross international borders and countries report only 13 aquifer-specific cooperation arrangements — five of which do not actually operatebuilt without the people who'd live with itsource: primaryfield: health & medicinethe pack left open on the washerPackaging rules for laundry detergent pods were followed by fewer child poisonings, but eye injuries from detergent packets rose as a share of child eye-injury emergency visits at Canadian sentinel hospitals (about 17% a year, 2012–2022); a latched lid added in Italy showed no detectable effect, and an in-home study written and funded by Procter & Gamble found adults often did not reclose child-impeding packsbuilt without the people who'd live with itsource: primaryfield: digital & spaceten times the data, same computing budgetThe high-luminosity LHC will produce ten times the collision data, and current computing cannot scale to itbuilt without the people who'd live with itsource: primaryfield: digital & spacea bubble that keeps every stranger outSocial VR platforms give players mute, block, report and personal-space bubbles, but in a 100-player survey harassment stayed common and the VR-specific tools were little used; for disabled players, reactive tools act too late, default boundaries also keep out welcome strangers, and visible tools carry a social costbuilt without the people who'd live with itsource: primarySources
- See PJ. "Together yet apart: Understanding hybrid learning through students, educators and classroom observations." Doctoral thesis, University of Otago, 2026, doi:10.82348/our-archive.00399, accessed 2026-11-17 (a single-author doctoral thesis, examined but not peer-reviewed as a journal article) Link ↗
- Raes A, Detienne L, Windey I, Depaepe F. "A systematic literature review on synchronous hybrid learning: gaps identified." Learning Environments Research 23(3):269-290, 2020, doi:10.1007/s10984-019-09303-z (author manuscript read; Internet Archive copy), accessed 2026-11-17. Link ↗ Link 2 ↗
- Mahmud MM, Teh JKL, Azizan SN. "Hyflex learning and student engagement in higher education: a systematic literature review." Frontiers in Education 11:1773704, 2026, doi:10.3389/feduc.2026.1773704, accessed 2026-11-17. Link ↗
- See PJ, Daniel B, Koh J. "Hybrid learning environments in higher education: A systematic review of emerging learning and teaching modalities." Research and Practice in Technology Enhanced Learning 21:35, 2026, doi:10.58459/rptel.2026.21035, accessed 2026-11-17. Link ↗
- Jakonen T, Rusk F. "The accountability of movement in synchronous hybrid classroom interaction: managing mobility troubles in robot-mediated remote participation." Learning, Culture and Social Interaction 56:100956, 2026 (available online 1 December 2025), doi:10.1016/j.lcsi.2025.100956 (open-access journal PDF read in full from the University of Turku repository), accessed 2026-11-17. Link ↗
- Tudela H, Serrurier M, Mechergui N, Brudieux Y, Jaïdane M. "Collaborative Synchronous Hybrid Learning Environments: Opportunities and Audio/Acoustic Quality Challenges." Proceedings of the 17th International Conference on Computer Supported Education (CSEDU 2025), Volume 1, pp. 512-520, 2025, doi:10.5220/0013352300003932, accessed 2026-11-17. Link ↗
- Acoustical Society of America. "Classroom Acoustics for Architects: A companion booklet for ANSI/ASA S12.60 Parts 1 and 2" (file name dated 18 April 2015), accessed 2026-11-17. Link ↗
- ASA Standards, notice of the revised Part 1, "Now Published: ASA/ANSI S12.60-2026/Part 1 Sets Updated Acoustical Criteria for Permanent Schools" (the standard itself is sold and was not), accessed 2026-11-17; vendor pages (makers' own material): Sennheiser, Singapore Management University customer story. Link ↗ Link 2 ↗
- Catchbox, "The problem with ceiling array mics" Link ↗
- Owl Labs, "Top 8 Classroom Setups Using the Meeting Owl" Link ↗
- Shure MXA910 guide, all accessed 2026-11-17. Link ↗
Revision notes
This brief gets checked against its sources and contains the following revisions. Future revisions get added here.
Worth knowing
- the main source is one doctoral thesis, and its network counts come from five sessions of one voluntary workshop
- the thesis is not consistent on whether the mix of remote and in-room students affects how they interact; the page quotes both of its sentences
Full check
Sources read
- See, doctoral thesis on hybrid learning (University of Otago, 2026): full text
- Raes et al., systematic review of synchronous hybrid learning (Learning Environments Research, 2020): full text (the authors' manuscript)
- Mahmud, Teh and Azizan, systematic review of HyFlex learning and engagement (Frontiers in Education, 2026): full text
- See, Daniel and Koh, systematic review of hybrid learning environments (Research and Practice in Technology Enhanced Learning, 2026): full text (same first author as the thesis)
- Jakonen and Rusk, a telepresence robot in a university language class (Learning, Culture and Social Interaction, 2026): full text
- Barr and Luo, review of HyFlex course design (2025): full text (no claim on the page rests on it)
- Wood, Detyna and Dommett, review of synchronous hybrid teaching (2025): full text (no claim on the page rests on it)
- Sennheiser, Singapore Management University customer story: full text (a maker's page)
- Catchbox, page on ceiling array microphones: full text (a competing maker's page)
- Owl Labs, classroom set-ups for the Meeting Owl: full text (a maker's page)
- Shure, MXA910 product guide: could not be opened (only the page heading loaded)
- ANSI/ASA S12.60 classroom-acoustics standard: could not be opened (sold, not free; the standards society's companion booklet was read instead)
What changed
- Not found in the sources, removed: disability among the reasons students rely on joining remotely; the sources name illness, work, family and distance.
- Stated more strongly than the source: the page said most exchanges ran through the tutor in all five recorded sessions; the first two sessions had student-to-student talk, and the page now says every exchange ran through the tutor from the third session on (also in the section on what would unlock progress); 'remote students said' turning the camera on made them feel in the class; it was one student.
Left open
- what the Shure ceiling array is sold for, beyond being a ceiling microphone [its page did not load]
- what the classroom-acoustics standard itself and its 2026 revision say [sold, not read]
Separate check of specific claims: A company lab test of two calls at once in one room was added, an unsupported sentence on the classroom-acoustics standard was replaced with what its companion booklet says, and the thesis's second, conflicting sentence on the mix of students was added.
Earlier full check: The telepresence robot is now called remote-controlled, not self-driving, one student's account of putting an ear to the speaker is no longer given as what in-room students did, and the per-group microphone pod is now credited to educators in the thesis, who named it as an ideal.
Separate check of specific claims: Consent lines were added to the audit and make doors, and an unsupported claim about how common hybrid classes are was removed.
Earlier full check: The full title was toned down (mixed groups struggled to hear one another, rather than could not), the thesis is now named as one researcher's thesis, and two reviews and the whole-room microphone systems on sale were added.