{"id":"9b7eb197-50f2-42ce-8911-df0185cb82b4","arxiv_id":"2506.14653","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"A systematic literature mapping shows that rebound effects are rarely considered in smart home energy efficiency research and that HCI is relatively more aware than other computing fields.","lead":"This paper maps how often research on smart homes and energy efficiency mentions rebound effects, the economic phenomenon where efficiency gains are offset by increased consumption. It finds such consideration is rare in computing and HCI literature and proposes a taxonomy of actions for HCI to address it.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Rebound-term synonym gap could make the central finding an artifact of query wording; an expanded synonym search is the decisive check.","rationale":"The paper is a careful, well-scoped systematic mapping with disclosed limitations and reproducible scripts; the dual search and multiple database comparison are genuine strengths, and the manual review of the small SIGCHI set helps interpret the full-text counts. The single most load-bearing step, however, is the implicit identification of the concept of rebound with the literal query string 'rebound effect*' (plus two rare variants). The central conclusion is an absence claim: the field rarely considers rebound effects. Absence claims are only as strong as the search's recall. The authors acknowledge false negatives in Section 5.4 but do not measure recall; they assert that the terminology is well-established. That assertion is plausible in energy economics but not self-evident in HCI/computing, where the same dynamics often appear as 'backfire', 'take-back', 'rebound' without 'effect', 'induced demand', or adjacent concepts such as the energy performance gap (which Section 2.1 treats as a related failure mode). Because Search 2 only widens the location of the same terms (full text rather than metadata), it cannot rescue recall. If the expanded synonym test produces materially higher ESR counts, especially in IEEE/ACM or with substantive HCI papers, the headline finding narrows from 'rarely considered' to 'rarely named', and the call to action loses much of its empirical force. Since this is testable with the same protocol and shared scripts, I recommend CONDITIONAL rather than outright acceptance: accept provided the sensitivity analysis confirms the gap. This is not a rejection; the burden of proof is modest and the existing evidence is substantial.","tokens_in":25154,"tokens_out":13655,"duration_ms":150568,"concrete_test":"Re-run the Section 3 protocol on the same databases and date ranges with an expanded rebound set: 'rebound' (standalone), 'rebound effect*', 'backfire', 'take-back', 'takeback', 'Jevons paradox', 'Khazzoom-Brookes postulate', 'induced demand', and, as a separate sensitivity arm, 'energy performance gap'. Recompute the ESR intersections and the ratios in Eqs. (1)-(2), and hand-code a random sample of any new computing/SIGCHI hits to see whether they substantively discuss rebound mechanisms rather than using the terms parenthetically. If the triple-intersection counts rise by more than a small factor (e.g., >5x) or if computing databases gain a non-trivial set of substantive papers, the central claim should be narrowed to 'rarely named' rather than 'rarely considered'. If the counts stay low and the new hits are peripheral, the original conclusion stands.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim (Section 5) — that rebound effects are poorly represented in smart home energy efficiency research — rests on the assumption that the query 'rebound effect*' OR 'Jevons paradox' OR 'Khazzoom-Brookes postulate' reliably detects papers that engage with rebound phenomena. The first term effectively does all the work; the other two are rare. This leaves out 'rebound' used as a standalone noun, 'backfire', 'take-back'/'takeback', 'behavioral response to efficiency', 'induced demand', and adjacent framing such as the 'energy performance gap' that Section 2.1 itself discusses as a rebound-related failure of predicted savings. Search 2 does not fix this: it still requires the same narrow term set, just anywhere in the full text. Section 5.4 acknowledges false negatives but asserts they are minimal because 'terminology ... is well established'; that is a plausible but untested claim. If a meaningful portion of smart home energy research discusses rebound-like dynamics under these other labels, then the near-zero triple-intersection counts in Table 1 and the flat trends in Figure 3 are an artifact of the search vocabulary, not evidence about the field's substantive engagement. Because the paper's call to action is premised on the gap being real rather than lexical, this is the most load-bearing assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports a systematic literature mapping of how often rebound effects are considered in smart home energy efficiency research. The authors search four bibliographic databases (Scopus, ScienceDirect, IEEE Xplore, ACM Full-Text Collection) plus a curated SIGCHI corpus, using three keyword clusters: 'smart home*', 'energy efficien*', and 'rebound effect*' with two rarer alternatives. Searches are performed twice: once on titles, abstracts, and keywords (Search 1) and once on full text (Search 2). The reported counts show very small numbers at the triple intersection, zero in IEEE, ACM, and SIGCHI in Search 1, and only a handful of full-text mentions in Search 2. Temporal analysis shows flat or zero annual counts of rebound-focused smart home papers despite growth in the broader research area. The paper interprets this as evidence that rebound effects are neglected in smart home energy efficiency research and proposes a taxonomy of actions for HCI to identify, measure, explain, and mitigate direct, indirect, and structural rebound effects.","tokens_in":25364,"tokens_out":4662,"duration_ms":50231,"significance":"If the central conclusion is robust, the paper is a useful and timely contribution to Sustainable HCI and to the broader energy efficiency literature: it quantifies an important gap, connects a well-established economics concept to HCI practice, and offers a practical taxonomy. The strengths are the transparent and replicable search protocol, the dual metadata/full-text strategy, the shared scripts on Zenodo, the manual checks on IEEE result sets, and the candid limitations discussion. The paper is also careful to distinguish between 'mentioning' and 'substantively engaging with' rebound effects in its qualitative review of the SIGCHI papers. The main weakness is that the quantitative conclusion rests on a narrow set of search terms, and this vulnerability is not resolved by the reported limitations discussion.","major_comments":[{"comment":"The three-keyword operationalization of 'rebound effects' is too narrow, and the claim that false negatives are minimal is unsupported. The query in §3 uses only 'rebound effect*' OR 'Jevons paradox' OR 'Khazzoom-Brookes postulate'; the first term does nearly all the work in the reported counts (Tables 1 and 2). The paper's own background defines 'backfire' as another name for the same phenomenon (§2.2.1) and discusses the 'energy performance gap' as a rebound-related failure of predicted savings (§2.1), yet neither term is searched. Also absent are standalone 'rebound', 'take-back'/'takeback', 'induced demand', and 'behavioural response to efficiency'. Because Search 2 re-uses the same narrow term set anywhere in the full text, it cannot detect papers that discuss rebound-like mechanisms under these labels. Section 5.4 asserts that the terminology is 'well established' and that the number of missed papers should be minimal, but no false-negative check is reported. Since the central finding of neglect is a count of papers using particular words, this vocabulary choice is load-bearing; I ask for an expanded synonym search, or a manual check of a random sample of smart-home efficiency papers for conceptual engagement with rebound, before the conclusion can be considered robust.","section":"§3 and §5.4"},{"comment":"The comparative claim that HCI is 'most aware' of rebound effects is based on very small counts. Table 2 row 7 gives SIGCHI 6 of 41 energy-efficiency smart-home papers (14.63%), versus single-digit numerators in IEEE and ACM; a one-paper shift changes the SIGCHI percentage by roughly 2.4 percentage points, and the manual reading in Table 3 reduces the six results to five substantive papers, none of which places rebound at the core of the analysis. The paper acknowledges the small counts in places, but §5.2 later uses the comparison as a positive finding ('HCI shows more awareness of rebound effects than all of the other research communities we included in our search'). Either add uncertainty quantification or soften the claim to 'a few SIGCHI papers mention rebound, mostly in passing'.","section":"§4.3 and §5.2"}],"minor_comments":[{"comment":"The search terms include 'Jevons paradox' without the possessive apostrophe, although the paper elsewhere writes 'Jevons' paradox' or 'Jevons' Paradox'; adding both variants would be a cheap robustness improvement.","section":"§3"},{"comment":"The sentence 'not a single paper in the field of computing ... based on Search 1' is accurate for the IEEE and ACM rows in Table 1, but it should be qualified as applying to the metadata search, since the SIGCHI full-text search does return ESR results.","section":"§4.1"},{"comment":"The table title promises a list of papers but the footnote explains that one of the six results is the full proceedings of CSCW'17, leaving five entries; consider retitling the table or stating the reduction in the main text.","section":"Table 3"}],"recommendation":"major_revision","confidential_remarks":"The paper is well executed and the mapping protocol is exemplary in transparency, but the quantitative core rests on one load-bearing assumption: that the chosen rebound vocabulary captures the field's substantive engagement. The decisive check is an expanded-synonym search and/or a manual false-negative audit. If the authors run that check and the counts remain essentially unchanged, I would support acceptance; the SIGCHI 'most aware' claim should be softened regardless because of the tiny counts."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"I'd read this as a solid, well-scoped mapping study rather than a breakthrough, and I think the central finding survives contact with the evidence. The paper's real contribution is the first cross-database count of how often smart home energy efficiency work explicitly engages with rebound effects, and the numbers are stark: near-zero hits at the three-way intersection in metadata searches across all five corpora, and the SIGCHI subset, small as it is, actually shows the highest proportional engagement (6 of 41 full-text papers). The authors also do the right things methodologically: dual search strategy, manual false-positive checks on IEEE, shared scripts, and a clear limitation section. The temporal analysis is honest — flat or zero rather than a growth story. That is a useful, field-internal result that should change how sustainable HCI frames evaluation. The soft spot is the one the stress-test note flags: the keyword set for rebound effects is narrow, and the paper does not test whether alternative labels like 'backfire', 'take-back', or 'behavioral response to efficiency' would change the counts. But I do not think that concern is fatal. For one, the full-text search would have caught many such papers if the terms were in common use, because those papers would likely still mention 'rebound' somewhere in related framing. For another, the term 'rebound effect' is genuinely well established in the energy and policy literature that presumably informs this work; a paper substantively engaging with the concept would nearly always use that name. The deeper worry is that the authors hand-wave this in Section 5.4 ('terminology is well established') without testing it. That is a minor empirical gap, not a load-bearing flaw. An expanded synonym search would have been the clean way to close it, and its absence costs the paper some certainty but not its core conclusion. The taxonomy in Section 5.1 is a synthesis, not a validated framework — fine as a starting point, and the authors do not overclaim it. The small SIGCHI counts (33 energy efficiency papers in Search 1, 3 at ES) mean the HCI comparison percentages are noisy, though the consistency across databases and search modes keeps the qualitative conclusion stable. Self-citations appear in background but are not load-bearing for the quantitative result, so the low circularity burden is earned. Bottom line: this is a competent, useful paper for the sustainable HCI community and for anyone working on smart home energy claims. It deserves a serious referee and a reasonable editor would send it out. With a supplementary synonym check it would be close to airtight, but even as-is it is a credible piece of evidence for the field's collective blind spot.","headline":"A careful, transparent mapping showing rebound effects are rarely named in smart home energy research; the central finding likely holds, with a manageable terminology caveat.","tokens_in":732,"tokens_out":734,"would_cite":true,"duration_ms":27139,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Across five databases, smart home energy efficiency research largely ignores rebound effects.","keywords":["rebound effect","smart home","energy efficiency","sustainable HCI","literature mapping","Jevons paradox","digital rebound","energy savings"],"falsifier":"Run the same Search 1 and Search 2 in Scopus and IEEE with an expanded synonym set—'take-back', 'induced demand', 'behavioural response to efficiency', 'energy performance gap'—joined to smart home and energy efficiency terms; if the expanded intersection yields dozens of papers per database or a rising temporal trend, the paper's central conclusion would be an artifact of terminology rather than a fact about the literature.","tokens_in":24976,"feed_emoji":"🔁","tokens_out":4055,"duration_ms":40859,"temperature":0.7,"pith_summary":"This paper tries to establish that rebound effects—the economic phenomenon in which efficiency gains are partially or fully offset by increased demand—are almost entirely absent from smart home energy efficiency research. It reports a systematic literature mapping across Scopus, ScienceDirect, IEEE Xplore, the ACM Full-Text Collection, and a corpus of SIGCHI venues, using keyword clusters for smart homes, energy efficiency, and rebound effects. Its headline finding is that a title/abstract/keyword search found no computing paper at all combining all three topics, while the small number of intersection papers in general databases has not grown over time even as smart home research has flourished. If this is right, the viability of many efficiency claims in sustainable HCI is not yet established, and reported savings may be upper bounds rather than realized outcomes.","feed_headline":"Rebound effects are missing from smart home energy research","feed_subtitle":"A five-database literature scan finds efficiency savings in smart homes may be overstated until rebound effects are measured.","key_machinery":"The machinery is a systematic literature mapping built as a three-set Venn intersection of keyword clusters: 'smart home*', 'energy efficien*', and 'rebound effect*' (plus 'Jevons paradox' and 'Khazzoom-Brookes postulate'). Two search strategies run in four databases plus a SIGCHI corpus: Search 1 restricted to titles, abstracts, and keywords, and Search 2 full-text searches. The quantitative core is the set-intersection counts together with two ratios, $R_{in E}(\\%) = (E \\cap R) \\times 100 / E$ and $R_{in E \\text{ for } S}(\\%) = (E \\cap R \\cap S) \\times 100 / (E \\cap S)$, which let the authors compare rebound awareness across disciplines and show that computing lags the general energy literature by one to two orders of magnitude.","core_discovery":"The central discovery is that rebound effects are poorly represented in smart home energy efficiency research, in the authors' terms. In the metadata-focused search, Scopus and ScienceDirect returned 4 and 1 papers at the triple intersection, while IEEE and ACM returned zero; full-text searching raised these counts to 25, 12, and 92 in IEEE, ACM, and ScienceDirect, but the proportions remain low relative to the large energy efficiency literature. Temporal analysis shows that smart home energy efficiency research grew strongly from 2011 to 2020 while attention to rebound effects stayed flat or near zero. Within the SIGCHI corpus, 6 of 41 smart home energy efficiency papers (14.63%) mention rebound effects somewhere in the full text, a higher share than in the other computing databases, but closer reading shows the concept is usually mentioned in passing rather than placed at the core of analysis. From this, the authors conclude that the HCI community is relatively more aware of rebound effects than other computing communities yet has not systematically investigated them, and they respond with a taxonomy of actions for identifying, measuring, explaining, and mitigating direct, indirect, and structural rebound effects.","pith_inferences":["The keyword-based conclusion could be tested against a synonym-expanded replication: if terms like 'take-back', 'induced demand', 'behavioural response to efficiency', or 'energy performance gap' surface many smart home efficiency papers, the reported gap may be partly terminological rather than substantive.","A similar mapping could be run for other efficiency domains, such as electric vehicles, teleworking, or industrial IoT, to see whether digital rebound neglect is a general pattern across computing research rather than a peculiarity of smart homes.","A practical next step would be to build a shared measurement protocol for rebound in HCI field studies, combining longitudinal energy data with qualitative accounts of how saved money and time are re-spent, since the paper's own review shows that no existing study combines all three well.","The paper's evidence implies that policy evaluations relying on smart home efficiency estimates should discount those estimates until rebound magnitudes are measured; the authors gesture at this consequence but do not quantify it."],"forward_implications":["If the central claim is correct, reported smart home efficiency savings in the computing literature are best read as upper bounds, because rebound dynamics that would reduce them are generally not accounted for.","HCI is better positioned than other computing fields to take up rebound research, given its higher relative awareness of the concept and its existing expertise in in-situ, qualitative, and systems-oriented methods.","The taxonomy supplies a concrete agenda: for direct, indirect, and structural rebound, researchers can identify cases, measure effect sizes, explain mechanisms, and mitigate rebound through design and policy engagement.","Efficiency-only strategies for smart homes cannot be assumed to deliver climate benefits; sufficiency measures, emissions constraints, or design choices that anticipate rebound may be needed for savings to be real.","If HCI moves rebound to the core of its energy research, it can produce context-specific rebound estimates useful to policy makers and to more technically oriented fields that currently omit the effect."],"supporting_citations":[{"why":"Supplies the definition and survey of rebound effects that ground the paper's keyword choices and framing.","marker":"[1]"},{"why":"Provides the Jevons paradox and backfire evidence used to motivate including 'Jevons paradox' and 'Khazzoom-Brookes postulate' as search terms.","marker":"[98]"},{"why":"The 2011 CHI workshop call to consider rebound effects in sustainable HCI that the paper asks whether the community has heeded.","marker":"[54]"},{"why":"One of the few smart home rebound studies, used as an example of dedicated experimental work and of low ecological validity.","marker":"[18]"},{"why":"Agent-based modeling of rebound in smart homes that informs the taxonomy's discussion of direct and indirect rebound.","marker":"[116]"},{"why":"Introduces the 'digital rebound' concept used to argue that digital technologies are especially prone to rebound effects.","marker":"[22]"},{"why":"Argues that smart home technologies may reinforce unsustainable consumption, supporting the challenge to efficiency visions.","marker":"[108]"},{"why":"Example of HCI's existing qualitative research on energy feedback, used to claim HCI has the expertise to study rebound.","marker":"[40]"}],"fun_headline_variants":["Smart home savings may be overstated; rebound effects largely ignored","Rebound effects: the blind spot of smart home energy research","Why smart home efficiency gains could vanish: rebound overlooked","HCI must account for rebound to make smart home savings real"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The conclusion rests on the assumption that papers engaging with rebound-like dynamics reliably use the search terms 'rebound effect', 'Jevons paradox', or 'Khazzoom-Brookes postulate'; if researchers discuss the same phenomenon as take-back, induced demand, or energy performance gap, the mapping will under-count them.","fun_headline_variants_meta":{"raw":{"variants":["Smart home savings may be overstated; rebound effects largely ignored","Rebound effects: the blind spot of smart home energy research","Why smart home efficiency gains could vanish: rebound overlooked","HCI must account for rebound to make smart home savings real"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000815,"raw_usage":{"total_tokens":3564,"prompt_tokens":929,"completion_tokens":2635,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":545,"completion_tokens_details":{"reasoning_tokens":2567}},"tokens_in":545,"tokens_out":2635,"duration_ms":18767,"temperature":1.0,"reasoning_tokens":2567,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T19:48:53.363309+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same Search 1 and Search 2 in Scopus and IEEE with an expanded synonym set—'take-back', 'induced demand', 'behavioural response to efficiency', 'energy performance gap'—joined to smart home and energy efficiency terms; if the expanded intersection yields dozens of papers per database or a rising temporal trend, the paper's central conclusion would be an artifact of terminology rather than a fact about the literature.","supporting_citations":[{"cited_title":"Greening, David L","cited_arxiv_id":null,"evidence_quote":"Supplies the definition and survey of rebound effects that ground the paper's keyword choices and framing."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"One of the few smart home rebound studies, used as an example of dedicated experimental work and of low ecological validity."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Argues that smart home technologies may reinforce unsustainable consumption, supporting the challenge to efficiency visions."}],"review_version":2}