{"id":"44f2db06-1cda-420d-b3e9-99a74ecf34f4","arxiv_id":"2502.10166","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Interviews with 34 Irish digital educators show learner-led, step-by-step teaching works best, while fear, outdated devices, rural connectivity, funding, and transport gaps block older adults' digital inclusion.","lead":"A study of 34 digital educators in Ireland reports that older adults learn digital skills best when teaching is personalized, step-by-step, and learner-led, and that fear, outdated devices, poor rural connectivity, funding gaps, and transport problems are the main barriers. The findings give policymakers a practical, interview-based map of where to target digital inclusion funding for older adults.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The core 'what helps most' claims rest entirely on educator self-reports; because the protocol prompted for barriers and no outcome data were collected, the ranked themes may be prompting artifacts.","rationale":"I read the paper as an exploratory qualitative study with transparent limitations; it does real service in documenting educator-reported barriers. The reader's conditional verdict is appropriate. My concern sharpens the weakest assumption: the central claim is causal/evaluative ('maximise learning outcomes', 'can help reduce anxiety') but the evidence is entirely self-report, and the protocol's prompting creates a plausible mechanism for inflated theme counts. This is not about fraud or sample size alone; even a perfect sample of 34 educators would not by itself establish what maximises outcomes without outcome triangulation. The proposed re-coding is low-cost and would test whether the rankings are artifacts. The unsupported 25% statistic is a separate but symptomatic issue. I agree with the reader that CONDITIONAL is the right verdict; my test could be added as a condition for acceptance.","tokens_in":21585,"tokens_out":3154,"duration_ms":34371,"concrete_test":"Re-code all 34 transcripts at the utterance level, tagging each mention of a benefit or barrier as either unprompted (raised by the participant before the interviewer's follow-up prompt) or prompted (raised only after a prompt such as 'any other barriers?'). Then recompute the reported theme frequencies (learner-led 23/34, fear 25/34, scams 26/34, etc.) using only unprompted mentions. If any headline theme falls below a majority of interviews or changes rank, the paper must either downgrade the claim or collect outcome data; if the rankings survive, the prompting concern is mitigated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing premise is that educator testimony is a reliable measure of what maximises older adults' learning outcomes. The paper answers RQ1 ('maximise learning outcomes') using prevalence counts such as learner-led (23/34) and step-by-step instruction, and ranks fear as the most prominent barrier (25/34). But no learning outcome was measured: there are no pre/post assessments, no classroom observations, and no older-learner interviews. Section 6 discloses that interviewees 'often hesitated to critique their organisations... leading to initial claims of no barriers until we prompted for more detailed answers.' That means the frequency counts that drive the headline findings are vulnerable to social desirability and leading follow-up prompts. The unsupported claim in Section 4.2.4 that 'our preliminary study found that 25% of older adults have no sense of family support' is an example of the paper converting impression into a numeric fact without a cited source or reported analysis. If the prompting generated the barrier themes, the central claim about which mechanisms matter most is not established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports a qualitative study of 34 digital educators in the Republic of Ireland who teach digital skills to older adults. Using semi-structured interviews and reflexive thematic analysis (Braun and Clarke), the authors derive 16 themes from 39 codes and report prevalence counts for key themes, including learner-led instruction (23/34), fear as a barrier (25/34), inaccessible language (13/34), and lack of family support (11/34). The paper argues that personalized, step-by-step, learner-centered teaching is most effective; that fear, outdated devices, poor rural broadband, limited transport, and underfunding are the main barriers; and that simulated online banking platforms could reduce anxiety. It concludes with policy and programme recommendations for funding, infrastructure, educator upskilling, and corporate responsibility.","tokens_in":21695,"tokens_out":3323,"duration_ms":36241,"significance":"If its conclusions hold, the paper provides a useful counterpoint to the predominantly learner-focused literature by surfacing educators' perspectives on what works in digital inclusion programmes for older adults in Ireland. Its strengths include the use of a recognized qualitative method, the inclusion of representative quotations for each theme, the transparent reporting of theme frequencies, and an explicit limitations section. However, the central claim about what 'maximises learning outcomes' rests entirely on educator self-reports, with no measurement of learning outcomes, and the paper contains one unsupported quantitative claim (the '25%' figure in Section 4.2.4). These issues are fixable through re-framing and tightening, but they currently limit the evidential weight of the headline conclusions.","major_comments":[{"comment":"The statement 'our preliminary study found that 25% of older adults have no sense of family support when it comes to cyber safety' is unsupported. No dataset, methodology, or citation is provided for this figure anywhere in the manuscript, and it is unclear how a percentage was derived from 34 educator interviews, since the preceding text (11 of 34) is expressed as a count of interviews, not a percentage of older adults. This statistic is load-bearing because it is used to argue that family support cannot substitute for formal digital education. It must be removed or replaced with a properly sourced and described statistic.","section":"Section 4.2.4"},{"comment":"RQ1 asks what factors 'maximise learning outcomes', but no learning outcome was measured. The prevalence counts in Section 4.1 (e.g., learner-led 23/34, step-by-step instruction) are counts of what educators said about their practices and perceptions, not evidence of learning gains; the interview protocol asked about programme strengths and barriers, and Section 6 explicitly discloses that interviewees 'often hesitated to critique their organisations' and initially claimed 'no barriers until we prompted for more detailed answers.' Consequently, the paper's central claim that these factors are what most improve outcomes is not established. The findings should be re-framed as educators' reported practices and perceived benefits, and RQ1 or the related conclusions should be softened accordingly.","section":"Sections 3.5, 4.1, and 6"},{"comment":"The convenience sample is heavily skewed toward ETB staff (17 of 34) and unpaid volunteers (10 of 34), with only one for-profit college and one corporate drop-in service represented. The paper nevertheless generalizes to 'Ireland' and makes broad policy recommendations (e.g., increased public funding, corporate responsibility) in Section 5.5. The discussion should acknowledge how sector representation and the geographic distribution shown in Figure 1 limit the generalizability of theme frequencies, and the policy recommendations should be tempered accordingly.","section":"Section 3.3 and Section 5.5"}],"minor_comments":[{"comment":"The quotation contains a duplicated word: 'we we don't really know why it works.' This should be corrected to 'we don't really know why it works' or marked as a verbatim hesitation if that was intended.","section":"Section 4.1.3, quote P26"},{"comment":"The reference line reads 'InNine-teenth International Conference'—the spacing between 'In' and 'Nineteenth' is missing, and 'Nineteenth' should be spelled out without a hyphen.","section":"ACM Reference Format"},{"comment":"The sentence 'This study was approved by the ethical review board of the organisation' is vague; please name the approving institutional review board or ethics committee for transparency.","section":"Section 3.4"},{"comment":"The phrase 'In 2021 the Eurobarometer research findings found' is redundant; consider writing 'In 2021, Eurobarometer research found' for clarity.","section":"Section 4.3.3"},{"comment":"Reference [2] contains a garbled URL ('WH0?sequence=1'), and reference [28] would benefit from a direct stable link or DOI to aid retrieval.","section":"References [2] and [28]"}],"recommendation":"major_revision","confidential_remarks":"The paper's scope is more aligned with digital inclusion and accessibility research than with the tangible/embodied interaction themes typical of TEI, but it does speak to hands-on, embodied learning. The unsupported 25% statistic and the mismatch between RQ1 and the evidence are the main obstacles; they are correctable with careful revision. I would encourage the editor to request an anonymized version of the interview protocol with the revision, since the current description of the protocol is brief and the prompting of barriers disclosed in Section 6 is important for interpreting the theme frequencies."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the short version: this is a competent, useful qualitative study that deserves a serious referee, but the abstract and RQ1 overclaim. The data is what it is—34 self-selected educators' perceptions—and the paper mostly stays honest about that in the limitations.\n\nWhat's new: the Irish context, specifically the mix of ETBs, NGOs, rural transport and broadband delays, and the concrete proposal for simulated banking platforms. The thematic analysis follows Braun and Clarke, the quotes actually support the themes, and the authors cite the prior educator-interview literature (Geerts, Vercruyssen, Tomczyk) instead of pretending the findings are unprecedented. That's honest scholarship.\n\nThe soft spots are real but proportionate. First, RQ1 asks what 'maximises learning outcomes,' but no outcome was measured. The frequency counts (learner-led 23/34, fear 25/34) are prevalence of mentions in interviews, not effect sizes. That's standard for qualitative work, but the framing should be softened. Second, the unsupported claim in Section 4.2.4 that 'our preliminary study found that 25% of older adults have no sense of family support' is a numeric fact with no source or analysis behind it. That should be cut or properly referenced. Third, the sample is convenience-based and half ETB staff; the paper acknowledges this, and the limitation about prompting for barriers is disclosed in Section 6. The stress-test note is right that this makes the ranked themes vulnerable to social desirability, but that's a standard limitation of interview research, not a fatal flaw.\n\nThe paper would be stronger with the interview protocol and codebook released, and with member checking or a second coder. But for what it is—an applied, policy-facing qualitative study—the method is appropriate and the reporting is transparent.\n\nWho it's for: Irish policymakers, digital inclusion programme designers, and researchers working on older-adult digital skills. It won't change scientific understanding, but it's a legitimate empirical contribution. I'd send it to review, ask for revisions on the overclaiming RQ1 framing and the 25% claim, and recommend the authors archive their protocol.","headline":"Solid Irish educator-interview study with useful applied findings; the 'learning outcomes' framing overclaims what self-reported interviews can support, and one unsourced 25% statistic should be cut.","tokens_in":22265,"tokens_out":1866,"would_cite":false,"duration_ms":19094,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Digital educators in Ireland say older adults learn best through personalized, step-by-step, learner-led instruction, with fear and systemic barriers such as underfunding and rural broadband as the main obstacles.","keywords":["older adults","digital divide","digital educators","Ireland","digital literacy","cybersecurity","qualitative interviews","online banking training"],"falsifier":"A randomized comparison in which older adults receive either the described personalized, step-by-step, learner-led instruction or a standardized curriculum, measuring digital literacy gains and anxiety, would settle whether the approach is the active ingredient; the claim is undercut if learning gains are equal in both arms.","tokens_in":21356,"feed_emoji":"📱","tokens_out":3627,"duration_ms":35048,"temperature":0.7,"pith_summary":"The paper tries to establish what makes digital skills programmes work for older adults in Ireland by asking the people who teach them. Analyzing interviews with 34 digital educators, it argues that personalized, step-by-step, learner-led instruction is the most effective approach, while fear and anxiety, inaccessible technical language, outdated devices, and unsupportive family members are the main individual barriers. It also argues that systemic problems—underfunding, poor rural broadband, limited transport, and weak promotion—prevent existing programmes from reaching older adults. If the paper is right, policy and programme design should shift funding toward more one-to-one contact hours, simpler materials, and practical tools like a simulated online banking platform. The value of the paper is in elevating the educator's perspective, which is underrepresented in the digital-inclusion literature.","feed_headline":"Personalized teaching narrows the older-adult digital divide","feed_subtitle":"Irish educators say step-by-step, learner-led training beats fear, funding gaps, and rural broadband.","key_machinery":"The central mechanism is a qualitative study design: semi-structured interviews with 34 digital educators across Ireland, analyzed through reflexive thematic analysis that produced 39 codes and 16 themes. The load-bearing thematic cluster is learner-led, step-by-step instruction, which emerged most frequently (23 of 34 interviews) and is tied to the specific recommendation of a simulated, or 'dummy', online banking platform that lets older adults practice digital transactions without exposing personal financial data.","core_discovery":"The paper claims that, in the words of Irish digital educators, older adults learn digital skills most effectively when instruction is personalized, step-by-step, and learner-led, with the learner's own interests and questions steering the content. It reports that fear of making mistakes and of scams is the most frequently mentioned barrier, appearing in 25 of 34 interviews, and that systemic failings—underfunding, poor rural broadband, scarce transport, and inadequate promotion—prevent programmes from reaching older adults. It further claims that a risk-free simulated online banking platform would reduce anxiety and build digital confidence. The paper presents these findings as a guide for policymakers, programme designers, and corporations digitizing essential services.","pith_inferences":["If educator testimony is accurate, the same barriers likely appear in other countries with comparable rural broadband and transport gaps; the Irish findings could generalize to similar European contexts, but this is an inference rather than a claim in the paper.","The proposed simulated banking platform is testable: a pilot comparing older adults who practice on a dummy platform with those who receive only verbal instruction could quantify the anxiety-reduction effect the paper anticipates.","The paper's emphasis on the educator's role suggests that upskilling educators in cybersecurity may have a multiplier effect on older adults' safety; a cost-effectiveness study could make the funding case explicit.","Because most interviewees were Education and Training Board staff, the views of volunteer-only organisations may be underrepresented; a stratified replication by organisation type could test whether the theme frequencies hold."],"forward_implications":["Programme funders should prioritize flexible, one-to-one contact hours and learner-led curricula over fixed syllabi.","Public investment in rural broadband and transport infrastructure is a necessary condition for digital inclusion, not a luxury.","Banks and other digitizing services should provide safe practice environments, such as dummy online banking platforms, and retain human support channels.","Free, up-to-date cybersecurity training and simple-language course materials for educators would improve programme quality.","Advertising through radio, libraries, and community venues is needed because online ads miss the target audience."],"supporting_citations":[{"why":"Geerts et al. supply the 'instruction gap' concept and evidence that personalized instruction matters for older adults, which this study extends to the Irish context.","marker":"[74]"},{"why":"Schirmer et al. provide the 'lifeworld' approach that grounds the learner-led theme in the paper's findings.","marker":"[73]"},{"why":"Gates and Wilson-Menzfeld offer a systematic review of geragogy and programme delivery that contextualizes the time-constraint findings.","marker":"[75]"},{"why":"Tomczyk et al. present the educator perspective on digital inclusion challenges, a direct precedent for interviewing educators rather than learners.","marker":"[71]"},{"why":"Vercruyssen et al. show that 'basic' digital literacy includes emotional and social dimensions, supporting the paper's interpretation of fear as a barrier.","marker":"[76]"},{"why":"Hogan documents technophobia among Irish older adults and calls for personalized teaching, providing the Irish baseline for this study.","marker":"[66]"},{"why":"Flynn describes Irish older adults' preference for informal peer learning and flexibility, which the paper builds on when discussing learner-led formats.","marker":"[64]"},{"why":"Latulipe et al. examine how close others act as unofficial proxies in older adults' banking, grounding the recommendation for a dummy banking platform.","marker":"[93]"},{"why":"Eurobarometer data on Ireland's rural broadband gap support the paper's systemic-barrier argument regarding connectivity.","marker":"[63]"}],"fun_headline_variants":["Irish digital educators: step-by-step training closes the gap","Personalized lessons beat fear and bad broadband for older learners","For older adults, fear and poor broadband are the real blockers","Step-by-step, learner-led digital training wins for older adults","Digital educators: personalized, step-by-step teaching closes the divide"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper's conclusions rest on the assumption that what 34 educators say about older adults' barriers and learning needs accurately reflects older adults' actual experiences and outcomes, without direct triangulation from learners themselves.","fun_headline_variants_meta":{"raw":{"variants":["Irish digital educators: step-by-step training closes the gap","Personalized lessons beat fear and bad broadband for older learners","For older adults, fear and poor broadband are the real blockers","Step-by-step, learner-led digital training wins for older adults","Digital educators: personalized, step-by-step teaching closes the divide"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001202,"raw_usage":{"total_tokens":4923,"prompt_tokens":881,"completion_tokens":4042,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":497,"completion_tokens_details":{"reasoning_tokens":3958}},"tokens_in":497,"tokens_out":4042,"duration_ms":30279,"temperature":1.0,"reasoning_tokens":3958,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T19:08:04.055714+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A randomized comparison in which older adults receive either the described personalized, step-by-step, learner-led instruction or a standardized curriculum, measuring digital literacy gains and anxiety, would settle whether the approach is the active ingredient; the claim is undercut if learning gains are equal in both arms.","supporting_citations":[{"cited_title":"Informal and incidental learning in the workplace (Routledge revivals)","cited_arxiv_id":null,"evidence_quote":"Geerts et al. supply the 'instruction gap' concept and evidence that personalized instruction matters for older adults, which this study extends to the Irish context."},{"cited_title":"Digital skills training for older people: The importance of the ‘lifeworld’","cited_arxiv_id":null,"evidence_quote":"Schirmer et al. provide the 'lifeworld' approach that grounds the learner-led theme in the paper's findings."},{"cited_title":"What role does geragogy play in the delivery of digital skills programs for middle and older age adults? a systematic narrative review","cited_arxiv_id":null,"evidence_quote":"Gates and Wilson-Menzfeld offer a systematic review of geragogy and programme delivery that contextualizes the time-constraint findings."},{"cited_title":"Ireland one of europe’s worst for rural broadband gap, says eu report, 2021","cited_arxiv_id":null,"evidence_quote":"Tomczyk et al. present the educator perspective on digital inclusion challenges, a direct precedent for interviewing educators rather than learners."},{"cited_title":"How “basic” is basic digital literacy for older adults? insights from digital skills instructors","cited_arxiv_id":null,"evidence_quote":"Vercruyssen et al. show that 'basic' digital literacy includes emotional and social dimensions, supporting the paper's interpretation of fear as a barrier."},{"cited_title":"The digital divide has grown old: Determinants of a digital divide among seniors","cited_arxiv_id":null,"evidence_quote":"Hogan documents technophobia among Irish older adults and calls for personalized teaching, providing the Irish baseline for this study."},{"cited_title":"Keeping up with the times in ireland: Older adults bridging the age-based digital divide together? Studies in the Education of Adults , pages 1–19, 2024","cited_arxiv_id":null,"evidence_quote":"Flynn describes Irish older adults' preference for informal peer learning and flexibility, which the paper builds on when discussing learner-led formats."},{"cited_title":"Unofficial proxies: How close others help older adults with banking","cited_arxiv_id":null,"evidence_quote":"Latulipe et al. examine how close others act as unofficial proxies in older adults' banking, grounding the recommendation for a dummy banking platform."},{"cited_title":"A long term vision for eu rural areas, 2021","cited_arxiv_id":null,"evidence_quote":"Eurobarometer data on Ireland's rural broadband gap support the paper's systemic-barrier argument regarding connectivity."}],"review_version":1}