Designing for Repair and Extended Lifespan: Consumer Expectations and Economic Constraints

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1. A stronger connection to existing literature would be beneficial for the manuscript. Try to incorporate previous research into the discussion on circular economy, product repairability, and consumer behavior.
2. Make it clear how this study fills a gap in the literature
3. The survey was designed toward the use of recycled and repairable products, with particular attention to the influence of emerging  EU regulations emphasizing product repairability. The questionnaire aimed to capture participants’ perceptions and preferences regarding products with extended lifecycles  achieved through repair and reuse (lines 185-188). Make an effort to be more precise about the product categories
4. Additionally, the questionnaire contains generic questions, i.e. recucled, upscale, new materials, etc . Try to be more specific so that the reader can better understand what it refers to. (line 202-203). Better add the questionnaire in the appendix, and try to describe in you texts what is included. The most important points
5. I have some concerns about the generalizability of the study's results as the sample size is not large, i.e. 68 respondents (lines 205-212). Try to provide more justifications
6. Justifying the use of selected techniques to analyze the data is necessary for the data analysis section. (lines 214-224)
7. The results section is well analyzed with the correlations and ANOVA test
8. The thematic Summary of Open-Ended Responses, table 7, can be enriched (lines 449-451)
9. The discussion section could be improved by including relevant information about EU repairability regulations for sustainable products.

Comments on the Quality of English Language

Word and terminology improvements. Proofreading is essntial

Author Response

The response to reviewer 1 is in the attached file.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript is dealing with the interesting, timely and relevant topic, addressing an important issue within the field of product repair and extended product lifecycles. Investigating consumer opinions on this subject. The methodology appears generally appropriate, while there are some concerns to overall approach, which brings to question the research conclusions.

The literature review is well written and provides good introduction to the subject, incorporating up-to-date and relevant sources.  The research objectives are clearly articulated. Key aspects of the methodology sampling, data collection, and analysis techniques are explained in detail.

Minor Comments: Formatting should be corrected; some chapter titles are on the separate page from the chapter body.

Major limitations: While results are presented clearly, their interpretation could be misleading. Participants were surveyed on their attitudes toward the use of recycled and repairable products, perceptions and preferences regarding products with extended lifecycles achieved through repair and reuse. The quality of the answers is questionable because nowhere does it say what is the product or the category of products the question is referring to (except in the Q9 “Which of these products do you feel comfortable fixing?”). This approach makes most of the participants answers unclear and imprecise. 

Product and repair are to general and vague terms!

This makes criteria for some answers unclear. For example, Q5: Cost range of the product I am willing to repair: 0-30 EUR; 30-100 EUR; 100-500 EUR; 500-on EUR. The cost of the repair could be mayor factor, products age or obsolescence... this all depends on the product category.

If product category is not defined researchers could have asked what percentage of products price would be acceptable for the products in each price range. That would be more applicable to all product categories.

Term repair should be clearly explained to the participants, especially having in mind age group of the participants.

Example:

Basic Repairs: Cleaning, Resetting or restarting a device/system, tightening visible screws or connectors, Replacing easily accessible consumables (batteries, filters, bulbs)

Intermediate Repairs: Replacing modular parts , Adjusting alignment or calibration settings, Fixing minor mechanical faults (loose hinges, belts), Software troubleshooting or updates

Advanced Repairs: Replacing internal components (motors, circuit boards, pumps),

Electrical or mechanical diagnostics, Soldering, rewiring, or part reconditioning, Structural repairs requiring precision

Complex Repairs: Full system overhaul or reconstruction, Firmware-level or embedded system repair, Repairs involving hazardous materials or high voltage, Restoration of critical or safety-dependent systems

Overall, the manuscript shows promise but requires major revisions from the ground up before it can be considered for publication.

 

Author Response

The response to reviewer 2 is in the attached file.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Although the study offers valuable insights, it is subject to several notable issues and limitations:

1. The paper relies on only 68 valid questionnaires. In statistics, particularly when conducting ANOVA and correlation analyses, such a sample size is considered small, resulting in low statistical power. This makes the conclusions less robust. Furthermore, the specific sampling frame or randomization process is not detailed. Based on participant characteristics, the sample is heavily concentrated within specific groups (younger adults, students, specific countries), lacking representativeness.
2. The paper does not mention whether a pilot test for reliability and validity was conducted before the formal distribution of the questionnaire. For instance, Q2 required recoding during analysis to align with other questions, suggesting potential logical flaws in the original questionnaire design.
3. The study employs mixed scales (Q2, Q3, Q6 use a 5-point scale, while Q7, Q13, and Q15 use a 4-point scale). This inconsistency complicates data analysis and necessitates additional corrections during repeated measures ANOVA, increasing the difficulty of statistical processing and the risk of misinterpretation.
4. The research relies exclusively on questionnaires, lacking triangulation through behavioral experiments, interviews, or observation. As the paper admits, there is a 'social desirability bias,' where individuals tend to claim they are 'willing to pay for environmental protection,' even if this does not reflect their actual shopping behavior.
5. The research process involved data collection at only a single point in time. This makes it impossible to track the consumer's complete decision-making process from ‘awareness‘ to ‘purchase‘ and finally to ‘repair.‘
6. The study attempts to analyze categorical data and continuous data together. Although non-parametric tests were used, the overall model construction appears somewhat loose. Although Tukey adjustment was used to control for Type I errors, the risk of false positives remains given the small sample size and the performance of multiple correlation analyses.

Author Response

The response to reviewer 3 is in the attached file.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

Dear Authors,

Your manuscript entitled "Designing for Repair and Extended Lifespan: Consumer Expectations and Economic Constraints" aligns well with the scope of Sustainabilityregarding circular economy and social sustainability. However, to ensure the rigor and generalizability of the study, the following points require improvement:

1.The concept of "Pairing Paradox" is mentioned in the abstract; however, it requires a clear definition and in-depth elaboration in the main text.

2.It is recommended to expand the sample size and geographical diversity. If this is not feasible, please explicitly qualify the scope of inference in the conclusion section to reflect the exploratory nature of the study.

3.Figure 2​ ("Survey questions and possible answers") should be reformatted as a table​ for clearer presentation.

Comments on the Quality of English Language

The manuscript contains minor grammatical errors and awkward phrasing. It is recommended that the text be polished by a native English speaker or professional editing service.

Author Response

The response to reviewer 4 is in the attached file.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

No further comments

Author Response

Reviewer comment: No further comments.

The authors thank Reviewer 1 for confirming satisfaction with the revised manuscript.

Reviewer 2 Report

Comments and Suggestions for Authors

In the response authors indicate that the suggested improvements will be addressed only in future research does not sufficiently resolve the concerns raised in the current review process. Review comments are intended to strengthen the present manuscript. Without adequate revisions, several of the identified issues remain insufficiently addressed.

Author Response

General comment

Reviewer comment: In the response authors indicate that the suggested improvements will be addressed only in future research — this does not sufficiently resolve the concerns raised in the current review process. Review comments are intended to strengthen the present manuscript. Without adequate revisions, several of the identified issues remain insufficiently addressed.

 

The authors fully accept this criticism. Deferring concerns entirely to future research is not a satisfactory resolution. In this revision we have made six targeted additions to the manuscript text — in Sections 2.1, 2.3, 4.3.2, 5.4, and 5.7 — that address the substantive issues directly within the scope of the existing dataset. No new data collection was undertaken; instead, the existing findings are more fully interpreted and the methodological reasoning is made explicit where it was previously implicit. Each specific concern is addressed below.

 

Comment 2.1 — Products are too general; answers to Q5, Q7 etc. are unclear and imprecise

Reviewer comment: The quality of the answers is questionable because nowhere does it say what is the product or the category of products the question is referring to (except in Q9). This approach makes most of the participants’ answers unclear and imprecise. Product and repair are too general and vague terms!

 

The authors acknowledge this as a genuine methodological concern and do not dispute it. The general, multi-category framing of the primary survey was a deliberate design choice suited to the exploratory objectives of this study. Exploratory consumer research routinely employs broad-scope instruments to map general attitudinal patterns before narrowing to product-specific investigations — a staged approach advocated by Korsunova et al. (2023) [22] and Parajuly et al. (2024) [11], both cited in the manuscript. The goal was to identify which behavioral, economic, and functional factors consistently shape repair attitudes across product categories, not to generate product-specific predictions.

We recognise that this rationale was not sufficiently explicit in the manuscript. The following paragraph has been added to Section 2.1 immediately after the pilot description to make the design rationale transparent to the reader:

 

Added to Section 2.1: “The use of general product categories in the primary survey reflects a deliberate Stage 1 design choice consistent with exploratory consumer research, which typically maps broad attitudinal patterns before narrowing to product-specific investigations [11, 22]. Future phases of this research will focus on a single product family, allowing for deeper, more concrete consumer analysis.”

 

This addition signals to readers — and to the reviewer — that the general framing was a purposeful methodological choice, not an oversight, and that product-specific investigation is the planned next step. The study’s exploratory scope is now foregrounded rather than left implicit.

Additionally, the Limitations section (5.7) has been extended — as described under Comments 3.4 and 3.5 below — with a passage that further frames this as a Stage 1 study and acknowledges the constraint on product-specific inference that the general framing entails.

 

Comment 2.2 — Cost should be expressed as a percentage of product price

Reviewer comment: If product category is not defined, researchers could have asked what percentage of product price would be acceptable for the products in each price range. That would be more applicable to all product categories.

 

The authors thank the reviewer for this methodologically sound suggestion. We fully agree that expressing repair cost as a proportion of product price is more analytically robust than absolute EUR ranges, because it normalises the threshold across heterogeneous product categories. This point is now explicitly acknowledged in the manuscript.

The following text has been appended to the economic factors paragraph in Section 4.3.2:

 

Added to Section 4.3.2: “It should be noted that the absolute cost thresholds in Q5 (0–30 EUR; 30–100 EUR; 100–500 EUR; 500+ EUR) are product-category dependent and carry inherent interpretive limits when applied across heterogeneous product types. Expressing repair willingness as a percentage of product price would be methodologically preferable for cross-category comparisons and will be adopted in subsequent phases of this research.”

 

This addition serves two purposes: it validates the reviewer’s insight directly in the manuscript, and it provides an honest qualification of the Q5 findings that will help readers correctly interpret the cost threshold data. The commitment to adopt this approach in future research is stated explicitly.

 

Comment 2.3 — ‘Repair’ was not defined for participants; complexity taxonomy suggested

Reviewer comment: The term repair should be clearly explained to the participants, especially having in mind the age group of the participants. [Reviewer proposed taxonomy: Basic / Intermediate / Advanced / Complex repairs]

 

The authors thank the reviewer for this thoughtful suggestion. The proposed taxonomy — Basic, Intermediate, Advanced, and Complex repair — is a rigorous and practically useful classification. We must be transparent: the questionnaire was administered before this suggestion was received. It is not possible to retroactively apply an explicit repair definition to participants’ responses without compromising the integrity of the dataset, and we acknowledge this openly as a limitation.

However, the data already collected in Q9 — which asked participants which product categories they felt comfortable repairing — yields a repair-confidence gradient that maps directly onto the reviewer’s complexity taxonomy:

• **Low-complexity repairs** (Basic/Intermediate in the reviewer’s taxonomy): furniture (77.9% comfortable) and sports equipment (69.1%) — categories typically involving accessible interventions such as tightening, strap replacement, and structural fixes.
• **High-complexity repairs** (Advanced/Complex in the reviewer’s taxonomy): electronics (47.1%) and large home appliances — categories requiring specialized knowledge, proprietary tools, or internal component access.

This mapping was implicit in the previous version of the manuscript. It has now been made explicit by adding the following paragraph to Section 5.4 (Functional Factors), immediately after the repair confidence gradient discussion:

 

Added to Section 5.4: “The product-category repair confidence gradient observed in Q9 maps closely onto established repair complexity typologies. Categories in which respondents reported highest comfort — furniture (77.9%) and sports equipment (69.1%) — typically involve accessible, low-complexity interventions such as tightening, strap replacement, and structural fixes. By contrast, categories with the lowest confidence — electronics (47.1%) and large home appliances — correspond to technically advanced or complex repair tasks requiring specialized knowledge, proprietary tools, or internal component access. This gradient confirms that variability in repair confidence is substantially driven by perceived technical complexity rather than motivation alone, and supports the recommendation that consumer-facing repair guidance should be stratified by repair complexity level.”

 

This addition does three things: it operationalizes the reviewer’s complexity taxonomy within the existing data; it strengthens the functional factors analysis with a more theoretically grounded interpretation; and it explicitly advocates for stratified repair guidance — the practical implication the reviewer’s suggestion was pointing toward. In future studies, an explicit repair definition stratified by complexity level will be presented to participants at the outset of the questionnaire.

 

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Although the manuscript has been revised and improved, significant flaws remain in terms of theoretical depth, methodological rigor, technical application, and the handling of details. The issues raised in the previous review persist.

Author Response

General comment

Reviewer comment: Although the manuscript has been revised and improved, significant flaws remain in terms of theoretical depth, methodological rigor, technical application, and the handling of details. The issues raised in the previous review persist.

 

The authors thank Reviewer 3 for the sustained critical engagement. We recognise that certain methodological limitations cannot be eliminated without redesigning the study, and we do not attempt to obscure this. However, the methods employed are appropriate for the study’s explicitly exploratory aims, and the limitations have been disclosed transparently throughout. In this revision, we have made three targeted additions — to Sections 2.1, 2.3, and 5.7 — that directly address the reviewer’s specific technical concerns within the existing manuscript. We address each concern in turn below.

 

Comment 3.1 — Small sample size (N = 68), low statistical power, limited representativeness

Reviewer comment: The paper relies on only 68 valid questionnaires. In statistics, particularly when conducting ANOVA and correlation analyses, such a sample size is considered small, resulting in low statistical power. This makes the conclusions less robust. Based on participant characteristics, the sample is heavily concentrated within specific groups (younger adults, students, specific countries), lacking representativeness.

 

The authors acknowledge this concern fully. The limitation was already disclosed in Section 5.7, but we recognise that the previous version did not provide adequate quantitative justification for why the findings nonetheless warrant confidence. The following text has been appended to the first paragraph of Section 5.7:

 

Added to Section 5.7: “Although the sample size limits statistical power for detecting small effects, the large effect sizes observed (η²p = .313 and η²p = .246) provide reassurance that the omnibus findings are robust to this constraint; effect sizes of this magnitude are reliably detectable at N = 68 and are consistent with those reported in comparable exploratory studies in repair behavior research [30]. Exploratory consumer repair surveys with similar sample sizes have been published in peer-reviewed journals in this field, including Sandez et al. (2023) [23], and Van den Berge et al. (2022) [25].”

 

This addition makes two arguments in sequence. First, the effect size argument: both omnibus effects are large by Cohen’s (1988) benchmarks. Large effects are reliably detectable at N = 68; the concern about low power is most acute for small and medium effects, not for effects of the magnitude observed here. Second, the field-comparability argument: the manuscript now explicitly cites three recently published peer-reviewed studies in the same research area that used comparable sample sizes, demonstrating that N = 68 is not anomalous within this literature. Together these two arguments contextualize the sample size concern without dismissing it.

On sample composition: the concentration of younger adults and students from Slovenia and Sweden is explicitly discussed in Section 5.7 as a generalisation constraint. The inclusion of two contextually distinct EU countries was a deliberate design choice to provide some geographic breadth within the available sample, and no statistically significant differences were observed between the two national subsamples.

 

Comment 3.2 — No pilot test mentioned; Q2 recoding suggests logical flaw in questionnaire design

Reviewer comment: The paper does not mention whether a pilot test for reliability and validity was conducted before the formal distribution of the questionnaire. For instance, Q2 required recoding during analysis to align with other questions, suggesting potential logical flaws in the original questionnaire design.

 

As documented in the previous revision, the questionnaire was developed through a multi-stage iterative process including internal review and pilot distribution to approximately 10–15 individuals prior to formal data collection. The Q2 recoding was not the result of a logical flaw discovered after administration. Rather, Q2 was intentionally designed with reverse polarity as a standard psychometric technique for reducing acquiescence bias through scale polarity variation — a procedure planned before distribution and executed as intended.

The previous version of the manuscript stated simply that “Q2 was recoded”, without explaining why, which invited the reviewer’s inference of a post-hoc correction. To eliminate this ambiguity, the following sentence has been appended to the item-grouping paragraph in Section 2.1:

 

Added to Section 2.1: “Specifically, Q2 was intentionally formulated with reverse polarity (‘products made from recycled materials are less durable’) to reduce acquiescence bias. This recoding was planned from the instrument design stage and is not a post-hoc correction.”

 

This addition makes explicit what was previously implicit, and directly pre-empts the inference that the recoding signals a design error.

 

Comment 3.3 — Mixed scales (5-point and 4-point) complicate analysis and risk misinterpretation

Reviewer comment: The study employs mixed scales (Q2, Q3, Q6 use a 5-point scale, while Q7, Q13, and Q15 use a 4-point scale). This inconsistency complicates data analysis and necessitates additional corrections during repeated measures ANOVA, increasing the difficulty of statistical processing and the risk of misinterpretation.

 

The authors acknowledge the scale inconsistency as a limitation. The previous version of the manuscript did not make the analytical management of this inconsistency sufficiently explicit. The following paragraph has been added to Section 2.3 (Data Analysis), immediately after the existing justification for the statistical procedures:

 

Added to Section 2.3: “The risk of false positives arising from the small sample size and multiple analyses is managed through five complementary protections: (1) Spearman’s ρ as the primary non-parametric estimator, robust to scale-level and distributional assumptions; (2) three-estimator convergence (Pearson, Spearman, Kendall) as an internal replication standard — only associations consistent across all three are interpreted as substantive; (3) Tukey correction for all ANOVA post hoc comparisons; (4) Greenhouse–Geisser correction for sphericity violations; and (5) large observed effect sizes (η²p = .313 and .246) that are resistant to sampling noise at this sample size. Furthermore, the two repeated measures ANOVA models are each constructed within scale-consistent item sets: the first model analyzes only Q2, Q3, and Q6 (all five-point scales), and the second analyzes only Q7, Q13, and Q15 (all four-point scales), so the mixed-scale structure of the instrument does not compromise within-model comparisons. Model-based estimated marginal means are reported throughout rather than raw scale means to facilitate accurate within-group interpretation.”

 

This addition directly addresses the reviewer’s concern in three ways. First, it makes explicit that the two ANOVA models are each constructed from scale-consistent item sets — the five-point and four-point items are never mixed within a single model, so within-model comparisons are not contaminated by scale differences. Second, it explains that model-based estimated marginal means — not raw scale means — are reported throughout, which further insulates the results from scale-level artefacts. Third, it articulates the five-protection error-control strategy that was already implicitly present in the analytical choices but had not been presented as an integrated framework, which this addition now provides.

 

Comment 3.4 — No triangulation; reliance on self-report introduces social desirability bias

Reviewer comment: The research relies exclusively on questionnaires, lacking triangulation through behavioral experiments, interviews, or observation. As the paper admits, there is a ‘social desirability bias,’ where individuals tend to claim they are ‘willing to pay for environmental protection,’ even if this does not reflect their actual shopping behavior.

 

The authors acknowledge this limitation, which is explicitly disclosed in Section 5.7. A questionnaire-based approach is the methodologically standard choice at the exploratory stage of this research program, and the leading studies in this field — Parajuly et al. (2024) [11], Korsunova et al. (2023) [22], Marikyan and Papagiannidis (2024) [24], Jaeger-Erben et al. (2021) [10] — are all survey-based or qualitative in design.

The limitation cannot be resolved within the current study without new data collection. However, the final paragraph of Section 5.7 has been extended to address this concern in two ways. First, by foregrounding the ‘Stage 1’ framing of the study, which contextualizes the single-method design as appropriate for this stage of a research program. Second, by noting that the consistency of the findings with independently published studies in the same field constitutes a form of external convergent validation. The exact text added is quoted under Comment 3.5 below, as both responses are addressed by the same manuscript addition.

 

Comment 3.5 — Single time-point data collection

Reviewer comment: The research process involved data collection at only a single point in time. This makes it impossible to track the consumer’s complete decision-making process from ‘awareness’ to ‘purchase’ and finally to ‘repair.’

 

The authors agree that cross-sectional data cannot capture the full consumer decision-making trajectory. This limitation was already acknowledged in Section 5.7. Cross-sectional designs are, however, standard at the exploratory stage of consumer attitude research; all five most-cited foundational studies in this literature use the same design format [10, 11, 22, 23, 24].

The following text has been appended to the final paragraph of Section 5.7, which responds to both this comment and Comment 3.4 simultaneously:

 

Added to Section 5.7 (final paragraph): “The multi-category framing of the primary survey was a deliberate design choice for this Stage 1 exploratory study; product-specific investigations are planned as Stage 2 follow-up work. Furthermore, the consistency of the findings across the behavioral, economic, and functional dimensions — and their alignment with independently published repair behavior studies [10, 11, 23, 24, 30] — provides a form of convergent external validation that partially mitigates the limitations inherent in single-method, single-time-point designs.”

 

This addition serves two purposes in one passage. It frames the current study’s design constraints as appropriate to its stage within a multi-stage research program rather than as unaddressed flaws. And it provides the convergent external validation argument: the fact that the findings are consistent with five independently published peer-reviewed studies across different samples, countries, and methods substantially reduces the probability that the results are artefacts of this study’s particular design, and this is now stated explicitly in the manuscript.

 

Comment 3.6 — Risk of false positives; mixed data types; loose model construction

Reviewer comment: The study attempts to analyze categorical data and continuous data together. Although non-parametric tests were used, the overall model construction appears somewhat loose. Although Tukey adjustment was used to control for Type I errors, the risk of false positives remains given the small sample size and the performance of multiple correlation analyses.

 

This concern is addressed directly by the addition to Section 2.3 quoted in full under Comment 3.3 above. To summarise: the five-protection error-control framework now documented in Section 2.3 — Spearman primary estimator, three-estimator convergence, Tukey correction, Greenhouse–Geisser correction, and large effect sizes — constitutes a comprehensive response to the false-positive risk. Each protection was already present in the analytical choices; the addition makes the integrated framework visible and explicit for the first time.

On the specific concern about mixing categorical and continuous data: Spearman’s ρ, designated as the primary estimator throughout, is a rank-based measure that does not assume interval-level measurement and is appropriate for ordinal Likert data regardless of scale format. The three-estimator convergence requirement means that only associations upheld by three independent statistical tests — operating on different algorithmic principles — are reported as findings. This is substantially more conservative than the single-estimator approach standard in this literature.

### The comments in the attached file are the same, but due to use colors they are easier to read. 

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

Dear Editor-in-Chief,

The authors have addressed all my previous comments satisfactorily. Regarding Comment 2 concerning sample size, although the authors were unable to expand the dataset, they have explicitly acknowledged the exploratory nature of the study and clearly defined the scope of inference in the Limitations section. This adequately addresses my concern regarding the generalizability of the findings beyond the European context.

Author Response

Comment:

The authors have addressed all my previous comments satisfactorily. Regarding Comment 2 concerning sample size, although the authors were unable to expand the dataset, they have explicitly acknowledged the exploratory nature of the study and clearly defined the scope of inference in the Limitations section. This adequately addresses my concern regarding the generalizability of the findings beyond the European context.

Answer:

The authors thank Reviewer 4 for confirming satisfaction with the revised manuscript.

 

Round 3

Reviewer 2 Report

Comments and Suggestions for Authors

Authors’ efforts in responding to the comments are appreciated. Authors have provided explanations and have addressed the reviewer’s concerns through additional clarifications and revisions to the manuscript. The implemented changes improve the clarity of the paper, although some of primary methodological concerns were not addressed from the ground up. 

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript has been refined and improved. I have no other questions.