Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

OceanChat: The Effect of Virtual Conversational AI Agents on Sustainable Attitude and Behavior Change

T0 review · 4 major / 5 minor · reviewed 2026-08-09 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A three-round conversation with an AI marine character significantly increased pro-environmental intentions and sustainable product choices compared with static scientific information or static character narrative.

desk verdict A genuine three-way randomized comparison of interactive vs. static marine characters, but the headline effect estimates come from models that condition on post-intervention mediators, so the abstract overclaims the total effect. read the letter →

arxiv 2502.02863 v2 pith:WN5KHLTW submitted 2025-02-05 cs.HC cs.AI

Pat Pataranutaporn , Alexander Doudkin , Pattie Maes This is my paper · ORCID
classification cs.HCcs.AI
keywords pro-environmentalbehaviorconversationalagentslargelanguagemodelsanthropomorphismsustainabilityeducationmarineconservationvirtualcharactersbehavioralintentions
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to establish that a brief, three-round dialogue with an LLM-powered animated marine creature can move people toward pro-environmental intentions and sustainable choices more effectively than conventional static education or a static first-person character story. In a between-subjects experiment with 900 participants, the Conversational Character Narrative condition produced a significant positive effect on pro-environmental intentions ($\beta = 0.173$, $p<0.001$) and on the ratio of sustainable to non-sustainable products chosen ($\beta = 0.342$, $p=0.031$), while the Static Character Narrative condition did not show significant effects on these outcomes. The paper interprets this as evidence that interactivity, not just character-driven storytelling, carries the measured advantage. It also reports that the beluga whale character produced stronger emotional engagement on several measures, that the jellyfish was perceived as more anthropomorphic and better at conveying vulnerability, and that the intervention did not shift deeper outcomes such as climate policy support or psychological distance.

What carries the argument

The central object is the Conversational Character Narrative condition of OceanChat: an animated, 3D-rendered marine character whose first-person story is followed by a three-round free-text dialogue, with responses generated by a large language model, converted to speech, and coordinated with character animation in a single web interface. The mechanism works by holding the character, visual design, and environmental message content roughly constant across conditions while varying only whether the participant can converse with the character, so that observed differences on intentions and choices can be attributed to interactivity rather than to storytelling or visual polish.

What would settle it

Re-run the main regressions for pro-environmental intentions and sustainable choice preferences without the seven post-intervention perceived variables as covariates, or fit a mediation model that lets empathy, likeability, and climate-impact awareness sit between condition and outcome; if the conversational condition's coefficient drops to null, the claim that interactivity itself drives the effects would not be supported.

Watch

Extended reading notes

Core claim

The central claim is that interactive conversational dialogue with a virtual marine character changes self-reported environmental intentions and immediate sustainable choice preferences where static information and static character storytelling do not. The authors argue that the conversational condition works by making abstract environmental threats feel tangible and personalized, and they support the claim with regression results showing a significant condition effect on intentions and on a product-selection task, alongside qualitative chat responses expressing empathy, commitment, and practical barriers. They also find that the effect is not uniform: character interaction did not significantly change psychological distance toward climate change, climate policy adoption, or self-reported past behavior, and species choice mattered, with belugas eliciting more empathy, likeability, and perceived intelligence while jellyfish were rated higher on anthropomorphism and conveyed vulnerability more effectively.

Load-bearing premise

The load-bearing premise is that the post-intervention feelings about the character, such as empathy and likeability, are background variables to be controlled for rather than part of the mechanism through which the conversation works, and if they are actually the mechanism, the reported effect of the conversational condition could be over- or understated.

Editorial extensions

If this is right

  • If the central claim is correct, a single brief chat with an AI marine character is enough to raise self-reported pro-environmental intentions and shift immediate product choices, which makes conversational agents a scalable complement to conventional environmental education.
  • Interactivity, rather than character presence alone, appears to be the active ingredient: the static character narrative showed no significant effect on intentions or sustainable choice preferences, while the conversational condition did.
  • The intervention does not move entrenched measures such as climate policy support, psychological distance, or reported past behavior, so its expected value lies in nudging near-term decisions and intention setting rather than in converting skeptics or changing deep beliefs.
  • Species selection matters: the beluga whale generated stronger emotional engagement and perceived intelligence, while the jellyfish was rated more anthropomorphic and better at conveying climate vulnerability, suggesting that character choice should be matched to the communication goal.
  • The paper's proposed design principle of graduated anthropomorphism, adjusting how human-like a character appears according to interaction context, follows directly from the finding that both perceived anthropomorphism and authentic species traits contributed to outcomes.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper treats post-intervention perceptions such as empathy, likeability, and climate impact awareness as covariates; if those perceptions are actually mediators on the path from conversation to behavior, the reported condition coefficients may understate the total effect, and a formal mediation analysis would settle that question.
  • A natural extension would be a delayed follow-up: re-contacting participants one or two weeks after the chat to see whether intention gains and product-choice shifts persist or decay, since the study only measures outcomes immediately after the intervention.
  • Using an incentive-compatible choice, such as a real purchase or donation decision rather than a hypothetical product-selection ratio, would test whether the sustainable choice effect reflects genuine behavior change or social desirability in self-report.
  • Because the beluga excelled on emotional measures while the jellyfish excelled on anthropomorphism and vulnerability communication, a multi-species dialogue could be tested in which the featured character is matched to the specific environmental message, such as using a jellyfish for ocean-acidification content and a beluga for empathy-building appeals.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper presents OceanChat, an LLM-powered conversational agent embodied as an animated marine character (beluga whale, moon jellyfish, seahorse), and reports a between-subjects experiment (N=900) comparing three conditions: Static Scientific Information, Static Character Narrative, and Conversational Character Narrative. The main claims are that the Conversational Character Narrative condition significantly increases pro-environmental intentions (beta=0.173, p<0.001) and sustainable choice preferences (beta=0.342, p=0.031) relative to static approaches, that the beluga character elicits stronger emotional engagement across several perceptual measures, and that effects on deeper attitudes (climate policy support, psychological distance) are limited. The paper also describes a synthetic pre-evaluation in which LLM agents serve as simulated participants and reviewers, reporting Kolmogorov-Smirnov divergences between synthetic and real data.

Significance. If the headline effects survive re-analysis, the study is a useful data point for research on conversational agents and sustainability: it uses a deployed, real-time LLM system with a randomized between-subjects design, a relatively large sample, and it transparently reports null results on deeper attitude measures. The synthetic pre-evaluation is an honest exploratory addition; the authors themselves report KS divergences between synthetic and real participant responses. However, the current regression specification undermines the causal reading of the headline effects, and the multiple-comparison situation is unresolved. The paper is therefore of moderate significance pending analytic fixes.

major comments (4)
  1. [Section 4.6.2 and Section 6] The regression model specified in Section 4.6.2 includes the seven post-intervention perceived variables (anthropomorphism, animacy, likeability, intelligence, safety, empathy, climate-change impact awareness) as covariates in every outcome model. Section 4.6.1 lists these measures as post-intervention only, and the paper's own results show they predict the headline outcomes: Section 5.2.9 reports empathy (beta=0.404, p<0.001) as a significant predictor of sustainable choice preferences, and Section 6 states that effectiveness "was mediated by perceptual factors." These variables are therefore not pre-treatment confounders but plausible mediators on the causal pathway from the conversational character condition to the outcomes. Conditioning on them converts the reported condition coefficients from the total intervention effect claimed in the abstract into direct effects net of those mediators. The sentence in Section 4.6.2 that sensitivity analyses "remained consistent across these alternative specifications" is not backed by any reported sensitivity analysis in the manuscript. The authors should report models without the perceived-variable block, and ideally a mediation decomposition, to support the abstract's causal claims.
  2. [Section 5.2] The results section reports separate regressions for at least eleven outcome variables (Sections 5.2.3 through 5.2.11), each with multiple condition and animal contrasts, yet no correction for multiple comparisons is applied or reported. Section 4.4 states that the sample size was increased to maintain power "after any adjustments or corrections," but no adjustments or corrections appear anywhere in the results. The headline sustainable choice result (beta=0.342, p=0.031) and the sustainable consumption result (beta=0.077, p=0.079) are particularly vulnerable; a Bonferroni or FDR correction across the reported family of tests could render them non-significant. Please report adjusted p-values, or clearly designate a pre-specified confirmatory analysis separate from exploratory analyses.
  3. [Abstract, Section 5.2.7, Section 5.2.9, Section 6.1] The abstract and the discussion claim that the Conversational Character Narrative condition significantly increased behavioral intentions and sustainable choice preferences "compared to static approaches." However, the regression analyses in Sections 5.2.7 and 5.2.9 compare each condition only to the Static Scientific Information baseline. The Static Character Narrative condition is not significant for intentions (beta=0.039, p=0.315) or for sustainable choice preferences (beta=0.115, p=0.434). The manuscript does not report a direct contrast between Conversational Character Narrative and Static Character Narrative, so the claim that interactivity itself drives the effect, rather than character narrative alone, is not directly tested. Please report pairwise contrasts among all three conditions or soften the abstract wording to "compared to static scientific information."
  4. [Section 4.5 and Section 5] The manuscript reports that 900 participants were recruited, but that "the final analytic sample was randomly evened out and ranged from n=683 to n=782 across the various outcome measures." The exact numbers of participants excluded for attention-check failure, overly fast completion, or incomplete data are not provided, and the reason for the variable effective N across outcomes is not explained. This makes it difficult to verify the power analysis in Section 4.4 and to assess whether exclusions could bias the condition comparisons. Please provide a flow diagram with exclusion counts and report the effective N for each outcome model.
minor comments (5)
  1. [Section 2, Contribution 3] The word "Synhetic" appears to be a typo; it should be "Synthetic pre-evaluation of Study."
  2. [Section 5.2.9] In the animal-condition paragraph, the text refers to "compared to the baseline" without specifying which baseline. Earlier species comparisons in Section 5.2.2 use the jellyfish as the baseline, but this is not restated here, which can confuse the reader.
  3. [Section 6, Species Selection and Emotional Connection] The reported beta for beluga likeability is 0.361 in Section 5.2.2 but 0.360 in the discussion; please reconcile the discrepancy.
  4. [Section 4.2.2] "vue-js" should be written as "Vue.js," and the description of Replit as a hosting platform could be clarified.
  5. [Section 4.5] The phrase "randomly evened out" is vague; please clarify how the analytic sample was evened out and whether any random subsampling was performed.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the central claims rest on an independent human-subject experiment rather than on fitted inputs or self-citations.

full rationale

The paper's central claims are empirical: a between-subjects experiment with N=900 human participants randomly assigned to Static Scientific Information, Static Character Narrative, or Conversational Character Narrative. The condition coefficients (e.g., β=0.173, p<0.001 for pro-environmental intentions; β=0.342, p=0.031 for sustainable choice preferences) are estimated from participant responses, not derived from the input prompts or LLM outputs. No equation defines the outcome in terms of the intervention by construction, and no parameter is fitted to a subset and then 'predicted' on a closely related quantity. The synthetic pre-evaluation (Section 5.1) is internally self-referential in that LLM agents generated conversations and separate LLM agents reviewed them, but the authors explicitly quantify the divergence from real participants (KS statistics of 0.378 pre, 0.241 post, both p<0.001) and do not use the synthetic results as evidence for the main effects. No load-bearing self-citations or uniqueness theorems are invoked; the cited prior work (e.g., Costello et al. 2024, Giudici et al. 2024) is external. The regression specification in Section 4.6.2 includes post-intervention perceived variables as covariates, which raises a mediation/over-control concern about interpreting the condition coefficients as total effects, but this is a causal-inference validity issue, not circularity: the condition is randomized and the coefficients are not constructed to equal the inputs. The manuscript also claims sensitivity analyses 'remained consistent' without reporting them; while this is missing support, it does not make the derivation circular. Therefore the circularity score is 0.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No free parameters were fitted; assumptions are standard behavioral-study premises plus statistical modeling assumptions. No invented theoretical entities are introduced; the marine characters are software artifacts, and 'graduated anthropomorphism' is a post hoc design principle.

assumptions (4)
  • domain assumption Self-reported survey scales and the product-selection task capture meaningful environmental attitudes, intentions, and preferences.
    Section 4.6.1 defines dependent variables, but no validation or reliability statistics are reported for the custom experimental metrics.
  • domain assumption Post-intervention perceived variables can be treated as covariates without biasing condition coefficients.
    Section 4.6.2 includes them in all models; if they are mediators, estimates are biased.
  • domain assumption The exclusions and 'random even-out' of the analytic sample are ignorable.
    Section 4.5 describes N=900 reduced to n=683-782 with no detailed exclusion audit.
  • domain assumption gpt-4o-mini delivered the intended character personalities and factual content consistently across participants.
    Section 4.2 describes implementation but no output audit or reproducibility artifacts are included.

how reviews work

0 comments
Cite this review

Pith. "Pith review of OceanChat: The Effect of Virtual Conversational AI Agents on Sustainable Attitude and Behavior Change." pith.science (2026). https://pith.science/paper/WN5KHLTW

@misc{pith2026250202863,
  author       = {Pith},
  title        = {Pith review of: OceanChat: The Effect of Virtual Conversational AI Agents on Sustainable Attitude and Behavior Change},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WN5KHLTW}},
  note         = {Machine review of arXiv:2502.02863}
}
read the original abstract

Marine ecosystems face unprecedented threats from climate change and plastic pollution, yet traditional environmental education often struggles to translate awareness into sustained behavioral change. This paper presents OceanChat, an interactive system leveraging large language models to create conversational AI agents represented as animated marine creatures -- specifically a beluga whale, a jellyfish, and a seahorse -- designed to promote environmental behavior (PEB) and foster awareness through personalized dialogue. Through a between-subjects experiment (N=900), we compared three conditions: (1) Static Scientific Information, providing conventional environmental education through text and images; (2) Static Character Narrative, featuring first-person storytelling from 3D-rendered marine creatures; and (3) Conversational Character Narrative, enabling real-time dialogue with AI-powered marine characters. Our analysis revealed that the Conversational Character Narrative condition significantly increased behavioral intentions and sustainable choice preferences compared to static approaches. The beluga whale character demonstrated consistently stronger emotional engagement across multiple measures, including perceived anthropomorphism and empathy. However, impacts on deeper measures like climate policy support and psychological distance were limited, highlighting the complexity of shifting entrenched beliefs. Our work extends research on sustainability interfaces facilitating PEB and offers design principles for creating emotionally resonant, context-aware AI characters. By balancing anthropomorphism with species authenticity, OceanChat demonstrates how interactive narratives can bridge the gap between environmental knowledge and real-world behavior change.

Figures

Figures reproduced from arXiv: 2502.02863 by the authors.

Figure 1
Figure 1. Explanatory Overview of the OceanChat Intervention [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Study Design Overview [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Overview of all three conditions for the seahorse character (top) and OceanChat interfaces of all three animals [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (11 more)
Figure 4
Figure 4. Figure 4: Excerpt of Post-Intervention Product Shopping Sce [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Comparison of synthetic and real participant results for the dependent variable "Sustainable Consumption". [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Regression analysis for animal conditions and per [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: Regression analysis for experimental conditions [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]
Figure 8
Figure 8. Figure 8: Regression analysis for experimental conditions [PITH_FULL_IMAGE:figures/full_fig_p011_8.png]
Figure 10
Figure 10. Figure 10: Regression analysis for experimental conditions [PITH_FULL_IMAGE:figures/full_fig_p012_10.png]
Figure 12
Figure 12. Figure 12: Regression analysis for experimental conditions [PITH_FULL_IMAGE:figures/full_fig_p013_12.png]
Figure 13
Figure 13. Figure 13: Regression analysis for experimental conditions [PITH_FULL_IMAGE:figures/full_fig_p014_13.png]
Figure 15
Figure 15. Figure 15: Regression analysis for experimental conditions [PITH_FULL_IMAGE:figures/full_fig_p015_15.png]
Figure 17
Figure 17. Figure 17: Emotions identified in the chat conversations [PITH_FULL_IMAGE:figures/full_fig_p016_17.png]
Figure 18
Figure 18. Figure 18: Relative change of the dependent variables ap [PITH_FULL_IMAGE:figures/full_fig_p017_18.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. LLM-generated personalized nudges for improving pro-environmental behavior: Field evidence from resource conservation

    cs.CY 2026-04 conditional novelty 7.0 of 10

    LLM-personalized weekly nudges cut electricity use by 0.56 kWh per room-day (18.3 percentage points more than text-only feedback) in a five-week randomized trial.

Reference graph

Works this paper leans on

54 extracted references · 38 canonical work pages · cited by 1 Pith paper

  1. [1]

    Liisa Aavik. 2022. Using Nudging to Support Sustainable Choices. https://www. worldcleanupday.org/post/using-nudging-to-support-sustainable-choices

    work page 2022

  2. [2]

    Abhishek Aggarwal, Cheuk Chi Tam, Dezhi Wu, Xiaoming Li, and Shan Qiao

  3. [3]

    Michal Aibin and Milos Simic. 2024. Energy Consumption of ChatGPT Re- sponses. Baeldung. https://www.baeldung.com/cs/chatgpt-large-language- models-power-consumption Accessed: July 8, 2024. Includes energy values for GPT-3 (0.0003 kWh/query), GPT-4 (0.0005 kWh/query), and GPT-5 (0.001 kWh/query). Carbon intensity assumption: 0.5 kgCO2/kWh

    work page 2024

  4. [4]

    Icek Ajzen. 1991. The theory of planned behavior. Organizational Behavior and Human Decision Processes 50 (1991), 179–211. Issue 2. doi:10.1016/0749- 5978(91)90020-T

    doi:10.1016/0749- 1991

  5. [5]

    Christoph Bartneck, Dana Kulić, Elizabeth Croft, and Susana Zoghbi. 2009. Mea- surement instruments for the anthropomorphism, animacy, likeability, perceived intelligence, and perceived safety of robots.International Journal of Social Robotics 1 (11 2009), 71–81. Issue 1. doi:10.1007/S12369-008-0001-3

    doi:10.1007/s12369-008-0001-3 2009

  6. [6]

    Katharina Breiter, Feline Schnaak, and Henner Gimpel. 2024. Designing Conver- sational Agents to Promote Environmentally Friendly User Behavior. ICIS 2024 Proceedings (12 2024). https://aisel.aisnet.org/icis2024/user_behav/user_behav/4

    work page 2024

  7. [7]

    Alessia Cavaliere, Silvia Pigliafreddo, Elisa De Marchi, and Alessandro Banterle

    work page

  8. [8]

    Galloway

    Matthew Cole, Pennie Lindeque, Claudia Halsband, and Tamara S. Galloway. 2011. Microplastics as contaminants in the marine environment: A review. Marine Pollution Bulletin 62 (12 2011), 2588–2597. Issue 12. doi:10.1016/J.MARPOLBUL. 2011.09.025

    doi:10.1016/j.marpolbul 2011

Show all 54 references
  1. [9]

    Thomas H Costello, Gordon Pennycook, and David G Rand. 2024. Durably reducing conspiracy beliefs through dialogues with AI. Science (New York, N.Y.) 385 (2024), eadq1814. Issue 6714. doi:10.1126/science.adq1814

    2024 doi

  2. [10]

    Zhijun Dong and Tingting Sun. 2018. Combined effects of ocean acidification and temperature on planula larvae of the moon jellyfish Aurelia coerulea.Marine Environmental Research 139 (8 2018), 144–150. doi:10.1016/J.MARENVRES.2018. 05.015

    2018 doi

  3. [11]

    Ithier d’Aramon, Boris Ruf, and Marcin Detyniecki. 2024. Assessing Carbon Footprint Estimations of ChatGPT. (2024), 127–133. doi:10.1007/978-3-031- 59005-4_15

    2024 doi

  4. [12]

    Ghina ElHaffar, Fabien Durif, and Laurette Dubé. 2020. Towards closing the attitude-intention-behavior gap in green consumption: A narrative review of the literature and an overview of future research directions. Journal of Cleaner Production 275 (12 2020), 122556. doi:10.1016...

    2020

  5. [13]

    Edgar Erdfelder, Franz FAul, Axel Buchner, and Albert Georg Lang. 2009. Sta- tistical power analyses using G*Power 3.1: Tests for correlation and regres- sion analyses. Behavior Research Methods 41 (2009), 1149–1160. Issue 4. doi:10.3758/BRM.41.4.1149

    2009 doi

  6. [14]

    Queiroz, Linda Hambrick, Bryan A

    Géraldine Fauville, Anna C.M. Queiroz, Linda Hambrick, Bryan A. Brown, and Jeremy N. Bailenson. 2021. Participatory research on using virtual reality to teach ocean acidification: a study in the marine education community. Environmental Education Research 27 (2021), 254–278. I...

    2021

  7. [15]

    B J Fogg. 2003. Persuasive technology: using computers to change what we think and do . Morgan Kaufmann Publishers. - Triad of Persuasive Technolog (Captology)<br/>- (Gaia) as a social Actor<br/>-

    2003

  8. [16]

    Jambeck, and Kara Lavender Law

    Roland Geyer, Jenna R. Jambeck, and Kara Lavender Law. 2017. Production, use, and fate of all plastics ever made. Science Advances 3 (7 2017). Issue 7. doi:10.1126/sciadv.1700782

    2017 doi

  9. [17]

    Robert Gifford. 2011. The dragons of inaction: psychological barriers that limit climate change mitigation and adaptation. The American psychologist 66 (5 2011), 290–302. Issue 4. doi:10.1037/A0023566

    2011 doi

  10. [18]

    Robert Gifford, Christine Kormos, and Amanda McIntyre. 2011. Behavioral dimensions of climate change: drivers, responses, barriers, and interventions. Wiley Interdisciplinary Reviews: Climate Change 2 (11 2011), 801–827. Issue 6. doi:10.1002/WCC.143

    2011 doi

  11. [19]

    Mathyas Giudici. 2024. Persuasive Conversational Agents to Foster Sustainable Behaviours: Design, Evaluation, and Technology. (2024). https://www.researchgate.net/profile/Mathyas-Giudici/publication/387998521_ Persuasive_Conversational_Agents_to_Foster_Sustainable_Behaviours_ ...

    2024

  12. [20]

    Mathyas Giudici, Giulio Antonio Abbo, Pietro Crovari, and Franca Garzotto. 2024. Delivering Green Persuasion Strategies with a Conversational Agent: a Pilot Study. 811–821. https://hdl.handle.net/10125/106476

    2024

  13. [21]

    Aman Gupta, Amey Band, Shivam Sharma, and Karan Bilakhiya. 2021. Text2Emotion: A Python Package for Emotion Detection in Text. https: //shivamsharma26.github.io/text2emotion/ Detects emotions from textual mes- sages in five categories: Happy, Angry, Sad, Surprise, and Fear

    2021

  14. [22]

    Markus Hadler, Beate Klösch, Stephan Schwarzinger, Markus Schweighart, Re- becca Wardana, and David Neil Bird. 2022. Measuring Environmental Attitudes and Behaviors. Springer International Publishing, 15–35. doi:10.1007/978-3-030- 85796-7

    2022 doi

  15. [23]

    Donna J. Haraway. 2016. Staying with the Trouble: Making Kin in the Chthu- lucene. Staying with the Trouble (8 2016). doi:10.1215/9780822373780

    2016 doi

  16. [24]

    Haws, Karen Page Winterich, and Rebecca Walker Naylor

    Kelly L. Haws, Karen Page Winterich, and Rebecca Walker Naylor. 2014. See- ing the world through GREEN-tinted glasses: Green consumption values and responses to environmentally friendly products. Journal of Consumer Psychology 24 (7 2014), 336–354. Issue 3. doi:10.1016/J.JCPS....

    2014 doi

  17. [25]

    Kirsten Hillebrand and Florian Johannsen. 2021. KlimaKarl – A chabot to promote employees’ climate-friendly behavior in an office setting. SSRN Electronic Journal (2021). doi:10.2139/ssrn.3897674

    2021 doi

  18. [26]

    Perttu Hämäläinen, Mikke Tavast, and Anton Kunnari. 2023. Evaluating Large Language Models in Generating Synthetic HCI Research Data: a Case Study. Conference on Human Factors in Computing Systems - Proceedings (4 2023), 19. doi:10.1145/3544548.3580688/SUPPL_FILE/3544548.35806...

    2023

  19. [27]

    Google Inc. 2025. Model Viewer – Easily Display Interactive 3D Models on the Web and in AR. https://modelviewer.dev/ Accessed: 2025-01-26

    2025

  20. [28]

    International Energy Agency. 2023. The world’s top 1% of emit- ters produce over 1000 times more CO2 than the bottom 1%. https: //www.iea.org/commentaries/the-worlds-top-1-of-emitters-produce-over- 1000-times-more-co2-than-the-bottom-1

    2023

  21. [29]

    Brian Alan Johnson, André Derek Mader, Rajarshi Dasgupta, and Pankaj Kumar

  22. [30]

    Reto Knutti. 2019. Closing the Knowledge-Action Gap in Climate Change. One Earth 1 (9 2019), 21–23. Issue 1. doi:10.1016/j.oneear.2019.09.001 doi: 10.1016/j.oneear.2019.09.001

    2019 doi

  23. [31]

    Lin Long, Rui Wang, Ruixuan Xiao, Junbo Zhao, Xiao Ding, Gang Chen, and Haobo Wang. 2024. On LLMs-Driven Synthetic Data Generation, Curation, and Evaluation: A Survey. (6 2024). https://arxiv.org/abs/2406.15126v1

    2024 arXiv

  24. [32]

    Global Ecology and Conservation 21 (3 2020), e00812

    Citizen science and invasive alien species: An analysis of citizen science initiatives using information and communications technology (ICT) to collect invasive alien species observations. Global Ecology and Conservation 21 (3 2020), e00812. doi:10.1016/J.GECCO.2019.E00812

    2020 doi

  25. [33]

    Markowitz and Meaghan L

    Ezra M. Markowitz and Meaghan L. Guckian. 2018. Climate change communica- tion: Challenges, insights, and opportunities. Psychology and Climate Change: Human Perceptions, Impacts, and Responses (1 2018), 35–63. doi:10.1016/B978-0- 12-813130-5.00003-5

    2018 doi

  26. [34]

    Enikő Mészáros, Attila Bodor, Etelka Kovács, Sarolta Papp, Kamilla Kovács, Katalin Perei, and Gábor Feigl. 2023. Impacts of Plastics on Plant Develop- ment: Recent Advances and Future Research Directions. Plants 2023, Vol. 12, Page 3282 12 (9 2023), 3282. Issue 18. doi:10.3390...

    2023 doi

  27. [35]

    Raffaele Marfella, Francesco Prattichizzo, Celestino Sardu, Gianluca Fulgenzi, Laura Graciotti, Tatiana Spadoni, Nunzia D’Onofrio, Lucia Scisciola, Rosalba La Grotta, Chiara Frigé, Valeria Pellegrini, Maurizio Municinò, Mario Siniscalchi, Fabio Spinetti, Gennaro Vigliotti, Car...

    2017

  28. [36]

    Hung Vu Nguyen, Cuong Hung Nguyen, and Thoa Thi Bao Hoang. 2019. Green consumption: Closing the intention-behavior gap. Sustainable Development 27 (1 2019), 118–129. Issue 1. doi:10.1002/SD.1875

    2019 doi

  29. [37]

    OECD. 2022. Global Plastics Outlook . 201 pages. doi:https://doi.org/10.1787/ de747aef-en

    2022

  30. [38]

    Nabi, Abel Gustafson, and Risa Jensen

    Robin L. Nabi, Abel Gustafson, and Risa Jensen. 2018. Framing Climate Change: Exploring the Role of Emotion in Generating Advocacy Behavior. https://doi.org/10.1177/1075547018776019 40 (5 2018), 442–468. Issue 4. doi:10. 1177/1075547018776019

    2018 doi

  31. [39]

    OpenAI. 2025. Models Documentation. https://platform.openai.com/docs/models Accessed: 2025-01-26

    2025

  32. [40]

    Bondyale-Juez, Miguel González-Pleiter, Stefanie Reinold, Maite Asensio, Ico Martínez, and May Gómez

    Jorge Rapp, Alicia Herrera, Daniel R. Bondyale-Juez, Miguel González-Pleiter, Stefanie Reinold, Maite Asensio, Ico Martínez, and May Gómez. 2021. Microplastic ingestion in jellyfish Pelagia noctiluca (Forsskal, 1775) in the North Atlantic Ocean. Marine Pollution Bulletin 166 (...

    2021

  33. [41]

    OECD. 2022. OECD Data Explorer. https://data-explorer.oecd.org/vis?fs% 5B0%5D=Topic,1%7CEnvironment%20and%20climate%20change%23ENV%23% 7CPlastics%23ENV_PLS%23&pg=0&fc=Topic&bp=true&snb=11&df%5Bds% 5D=dsDisseminateFinalDMZ&df%5Bid%5D=DSD_PW@DF_PW&df%5Bag% 5D=OECD.ENV.EEI&df%5Bv...

    2022

  34. [42]

    Root-Bernstein, L

    M. Root-Bernstein, L. Douglas, A. Smith, and D. Veríssimo. 2013. Anthropomor- phized species as tools for conservation: Utility beyond prosocial, intelligent and suffering species. Biodiversity and Conservation 22 (7 2013), 1577–1589. Issue 8. doi:10.1007/S10531-013-0494-4/FIGURES/1

    2013 doi

  35. [43]

    Yolande Strengers. 2011. Designing eco-feedback systems for everyday life. Conference on Human Factors in Computing Systems - Proceedings (2011), 2135–

    2011

  36. [44]

    Replit. 2025. Replit – Build Apps and Sites with AI. https://replit.com Accessed: 2025-01-26

    2025

  37. [45]

    Environmental Protection Agency

    U.S. Environmental Protection Agency. 2023. Greenhouse Gas Equivalencies Cal- culator. https://www.epa.gov/energy/greenhouse-gas-equivalencies-calculator- calculations-and-references

    2023

  38. [46]

    A. M. van Valkengoed, L. Steg, and G. Perlaviciute. 2021. Development and vali- dation of a climate change perceptions scale.Journal of Environmental Psychology 76 (8 2021), 101652. doi:10.1016/J.JENVP.2021.101652

    2021

  39. [47]

    Madalina Vlasceanu, Kimberly C Doell, Joseph B Bak-Coleman, Boryana Todor- ova, Michael M Berkebile-Weinberg, Samantha J Grayson, Yash Patel, Danielle Goldwert, Yifei Pei, Alek Chakroff, Ekaterina Pronizius, Karlijn L van den Broek, Denisa Vlasceanu, Sara Constantino, Michael ...

    2024

  40. [48]

    Richard H Thaler and Cass R Sunstein. 2008. Nudge: improving decisions about health, wealth, and happiness . Yale University Press. 293 pages

    2008

  41. [49]

    Lacy, Erin Ashe, Ailsa Hall, Stéphane Plourde, Ian H

    Rob Williams, Robert C. Lacy, Erin Ashe, Ailsa Hall, Stéphane Plourde, Ian H. McQuinn, and Véronique Lesage. 2021. Climate change complicates efforts to ensure survival and recovery of St. Lawrence Estuary beluga. Marine Pollution Bulletin 173 (12 2021), 113096. doi:10.1016/J....

    2021

  42. [50]

    Evan You and Vue.js Contributors. 2025. Vue.js - The Progressive JavaScript Framework. https://vuejs.org/ Accessed: 2025-01-26

    2025

  43. [52]

    Lund, Agostino Marengo, Alessandro Pagano, Nishith Reddy Mannuru, Zoë A

    Ting Wang, Brady D. Lund, Agostino Marengo, Alessandro Pagano, Nishith Reddy Mannuru, Zoë A. Teel, and Jenny Pange. 2023. Exploring the Potential Impact of Artificial Intelligence (AI) on International Students in Higher Education: Gener- ative AI, Chatbots, Analytics, and Int...

    2023 doi

  44. [2020]

    12, Page 9627 12 (11 2020), 9627

    Do Consumers Really Want to Reduce Plastic Usage? Exploring the Determi- nants of Plastic Avoidance in Food-Related Consumption Decisions.Sustainability 2020, Vol. 12, Page 9627 12 (11 2020), 9627. Issue 22. doi:10.3390/SU12229627

    2020 doi

  45. [2023]

    Journal of Medical Internet Research 25 (2 2023), e40789

    Artificial Intelligence–Based Chatbots for Promoting Health Behavioral Changes: Systematic Review. Journal of Medical Internet Research 25 (2 2023), e40789. Issue 1. doi:10.2196/40789

    2023 doi

  46. [2144]

    doi:10.1145/1978942.1979252

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.