The DOI matches its Crossref record; 50 of the 50 references it cites were found in a registry or an index; none of the 45 sources its claims cite is checked for support yet.
Writing with AI Lowers Psychological Ownership, but Longer Prompts Can Help
Open accessAbstract
The feeling of something belonging to someone is called "psychological ownership." A common assumption is that writing with generative AI lowers psychological ownership, but the extent to which this occurs and the role of prompt length are unclear. We report on two experiments to examine the relationship between psychological ownership and prompt length. Participants wrote short stories either completely by themselves or wrote prompts of varying lengths. Results show that when participants wrote longer prompts, they had higher levels of psychological ownership. Their comments suggest they thought more about their prompts, often adding more details about the plot. However, benefits plateaued when prompt length was 75-100% of the target story length. To encourage users to write longer prompts, we propose augmenting the prompt submission button so it must be held down a long time if the prompt is short. Results show that this technique is effective at increasing prompt length.
Overview
CoChat Summary
TL;DR
Writing with a generative AI assistant lowers psychological ownership compared with writing independently, but requiring or encouraging longer prompts can raise it. Across three experiments, psychological ownership increased with prompt length up to about a 150-word minimum, then plateaued, and a press-and-hold submission delay nudged users to write longer prompts.
Key findings
- In Experiment 1, psychological ownership was lowest for the 3-word AI prompt condition (m = 1.80, sd = 1.13), then rose for 50–100 words (m = 3.90, sd = 1.61), 150–200 words (m = 4.57, sd = 1.70), and was highest for no-AI solo writing (m = 6.29, sd = 0.80).
- Longer prompts were associated with higher psychological ownership: Word Count correlated with Psychological Ownership at r_s = .72, p < .001; Text Similarity at r_s = .75, p < .001; Mental Demand at r_s = .65, p < .001; and Effort at r_s = .66, p < .001.
- Experiment 2 found a plateau beyond a 150-word minimum: ai-150 (m = 4.69, sd = 1.64), ai-175 (m = 4.70, sd = 1.73), and ai-200 (m = 4.54, sd = 1.81) did not differ significantly, and all AI conditions remained lower than no-AI (m = 6.19, sd = 0.99).
- Experiment 3 showed that time delays increased prompt length: 20-sec (m = 76.91, sd = 68.13) and 60-sec (m = 74.57, sd = 63.80) produced significantly higher Final Word Count than none (m = 34.66, sd = 46.56) and 0-sec (m = 28.09, sd = 32.69), while the strict 150-word minimum condition had the highest Final Word Count (m = 164.65, sd = 19.54).
Methods
- Experiment 1 used a within-subjects design with 31 valid Prolific participants after excluding 10 (24%). Participants wrote 150–200-word short stories either alone or with GPT-3.5 Turbo using prompts of 3 words, 50–100 words, or 150–200 words. Measures included a four-item Psychological Ownership composite (Cronbach’s α = 0.96), NASA-TLX items, logged Duration, Word Count, and Text Similarity via Google’s Universal Sentence Encoder; analysis used Friedman and Wilcoxon signed-rank tests with Holm’s corrections and Spearman correlations.
- Experiment 2 used 34 valid participants after excluding 5 (13%), with the same protocol and measures (α = 0.92) but prompt conditions of 150–200, 175–200, and 200–250 words.
- Experiment 3 used 156 valid participants after excluding 42 (21%) in a mixed design: condition was between-subjects (none n = 32, 0-sec n = 32, 20-sec n = 31, 60-sec n = 32, 150-word minimum n = 29) and trial was within-subjects (1–4). Participants wrote prompts to generate a 150–200-word story; delayed submission required holding the button, with delay proportional to prompt length. Logged measures included Final Word Count, Submission Attempts, Attempted Word Count, and Attempted Duration; analysis used Aligned Rank Transform, Kruskal-Wallis, Mann-Whitney U, Holm’s corrections, and Spearman correlations.
Limitations
- The paper states that it explored prompt length within a single prompt, whereas real users likely refine generated text over multiple prompts; it calls for exploring word count across multiple messages in a conversation.
- The paper states that the Experiment 2 plateau may be due to plateauing semantic text similarity and may disappear as models improve; the authors tested GPT-4 Turbo and 4-o with the same prompts and observed no differences in Text Similarity.
- The paper states that Experiment 3 may lack ecological validity: no additional rewards were given for longer prompts, Prolific crowdworkers may be more conscientious and willing to write longer prompts or less likely to abandon the system, and 20- and 60-second delays may not transfer to real-world systems; shorter delays, feedback, and explanations may also affect willingness to wait.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
0 Valid · 3 Not valid · 26 Not checked · 101 Own results and interpretations
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The second experiment used even higher word minimums, yet psychological ownership plateaued.”
- Also stated in: § 5.2.1 Psychological Ownership. · hedged
“Overall, psychological ownership when writing with AI seems to plateau beyond a 150 word minimum (Figure 8).”
§ Writing with AI Lowers Psychological Ownership, but Longer Prompts Can Help > 1 INTRODUCTION
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 5.2.1 Psychological Ownership. · hedged
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Using a similar creative writing task, we conducted an experiment with 20-and 60-second delays, and found prompt length increased compared to having no delays.”
- Also stated in: § 6 ENCOURAGING LONGER PROMPTS IN A REAL INTERFACE · hedged
“We ran a pilot experiment (n=90, 11-18 per condition) to better understand which modifications were the most promising for increasing prompt length, and our preliminary results 4 suggested that integrating a time delay into the prompt submission button could be effective (Figure C.2).”
§ Writing with AI Lowers Psychological Ownership, but Longer Prompts Can Help > 1 INTRODUCTION
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 6 ENCOURAGING LONGER PROMPTS IN A REAL INTERFACE · hedged
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Results showed that as participants wrote more to meet higher word minimums, they had higher levels of psychological ownership.”
- Also stated in: § 4.1 Psychological Ownership · plain
“Overall, writing more words improved feelings of psychological ownership (Figure 3).”
§ Writing with AI Lowers Psychological Ownership, but Longer Prompts Can Help > 1 INTRODUCTION
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 4.1 Psychological Ownership · plain
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Despite having the same word limits, we also observed a significant difference between ai-150 and no-ai.”
§ Writing with AI Lowers Psychological Ownership, but Longer Prompts Can Help > 4.2.2 Word Count.
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Our piloted interaction techniques all attempted to nudge users towards a known word minimum of 150 words as our first two experiments suggested that psychological ownership plateaued at 150 words for the given task.”
§ Writing with AI Lowers Psychological Ownership, but Longer Prompts Can Help > 9.2 Implementation Approaches and Challenges
hedged- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
61 more claims
Create a free account to see all 130 claims and every source verdict.
- It may state two separate claims as one. 34
- The quote only partly states this claim. 3
- Probably not a substantial claim. 41
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
“• 200-250 words with GPT, representing 100-150% of the maximum final story word count.”
§ Writing with AI Lowers Psychological Ownership, but Longer Prompts Can Help > 5.1 Choice of Prompt Lengths
Reported 150 · recomputed 125
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
“𝜒 2 3,𝑁 =31 = 99.00,𝑝 < .001,𝑊 = 1.00”
§ Writing with AI Lowers Psychological Ownership, but Longer Prompts Can Help > A EXPERIMENT 1: ANALYSIS DETAILS
Reported 99 · recomputed 93
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Writing with AI Lowers Psychological Ownership, but Longer Prompts Can Help > 8.2 Learning Effect
“Specifically, significant main effects of trial on Submission Attempts and Attempted Word Count and post hoc tests only revealed significant differences between trial 1 and all other trials.”
§ Writing with AI Lowers Psychological Ownership, but Longer Prompts Can Help > C PRELIMINARY INTERFACE MODIFICATIONS AND INTERACTION TECHNIQUES
“Table D.2: Omnibus and post hoc statistical tests for Experiment 3: (a) Final Word Count, (b) Submission Attempts, (c) Attempted Word Count for trial = 1.”
Citations
Cites
ChatGPT for Good? On Opportunities and Challenges of Large Language Models for Education
Kasneci, E., Seßler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Günnemann, S., Hüllermeier, E., et al. (2023) · doi:10.35542/osf.io/5er8f · Cited by 7,552
ChatGPT for good? On opportunities and challenges of large language models for education
Kasneci, E., Sessler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Günnemann, S., Hüllermeier, E., et al. (2023) · Learning and Individual Differences · doi:10.1016/j.lindif.2023.102274 · Cited by 7,326 · cited 1×
“We believe these results are promising and can positively impact human-AI collaborations.”
The aligned rank transform for nonparametric factorial analyses using only anova procedures
Wobbrock, J. O., Findlater, L., Gergle, D., & Higgins, J. J. (2011) · Proceedings of the SIGCHI Conference on Human Factors in Computing Systems · doi:10.1145/1978942.1978963 · Cited by 2,854
Toward a Theory of Psychological Ownership in Organizations
Pierce, J. L., Kostova, T., & Dirks, K. T. (2001) · The Academy of Management Review · doi:10.5465/amr.2001.4378028 · Cited by 2,546
The State of Psychological Ownership: Integrating and Extending a Century of Research
Pierce, J. L., Kostova, T., & Dirks, K. T. (2003) · Review of General Psychology · doi:10.1037/1089-2680.7.1.84 · Cited by 2,038 · cited 6×
“Psychological ownership may explain an illusory superiority cognitive bias ; specifically, Blau and Caspi [5, 6] showed that writers believe that edits made by a collaborator decreases quality of text, but paradoxically, the writer believes their own edits improve their collaborator’s text.”
InfluentialProceedings for the Annual Meeting of the Cognitive Science Society
Dziemianko, M., Clarke, A. D. F., & Keller, F. (2013) · Cited by 1,726
Douglas, B. D., Ewell, P. J., & Brauer, M. (2023) · PLOS ONE · doi:10.1371/journal.pone.0279720 · Cited by 1,532
Universal Sentence Encoder for English
Cer, D., Yang, Y., Kong, S.-Y., Hua, N., Limtiaco, N., St. John, R., Constant, N., Guajardo-Cespedes, M., Yuan, S., Tar, C., et al. (2018) · Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing: System Demonstrations · doi:10.18653/v1/d18-2029 · Cited by 1,277
Buçinca, Z., Malaya, M. B., & Gajos, K. Z. (2021) · arXiv · doi:10.48550/arxiv.2102.09692 · Cited by 1,124
An Aligned Rank Transform Procedure for Multifactor Contrast Tests
Elkin, L. A., Kay, M., Higgins, J. J., & Wobbrock, J. O. (2021) · The 34th Annual ACM Symposium on User Interface Software and Technology · doi:10.1145/3472749.3474784 · Cited by 664
Research works printed but not linked: 4
- Ina Blau and Avner Caspi. 2009. Sharing and Collaborating with Google Docs: The Influence of Psychological Ownership, Responsibility, and Student’s Attitudes on Outcome Quality. In E-Learn: World Conference on E-Learning in Corporate, Government, Healthcare, and Higher Education. Association for the Advancement of Computing in Education (AACE), 3329–3335.
- Ina Blau and Avner Caspi. 2009. What Type of Collaboration Helps? Psychological Ownership, Perceived Learning and Outcome Quality of Collaboration using Google Docs. In Proceedings of the Chais Conference on Instructional Technologies Research, Vol. 12. 48–55.
- Mark D Dunlop, Emma Nicol, Andreas Komninos, Prima Dona, and Naveen Durga. 2016. Measuring Inviscid Text Entry using Image Description Tasks. In Proceedings of the 2016 CHI Workshop on Inviscid Text Entry and Beyond.
- Serge Egelman, David Molnar, Nicolas Christin, Alessandro Acquisti, Cormac Herley, and Shriram Krishnamurthi. 2010. Please Continue to Hold. In Ninth Workshop on the Economics of Information Security.
Cited by
Which Contributions Deserve Credit? Perceptions of Attribution in Human-AI Co-Creation
He, J., Houde, S., & Weisz, J. D. (2025) · Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3706598.3713522 · Cited by 36
An Empirical Study to Understand How Students Use ChatGPT for Writing Essays
Jelson, A., Manesh, D., Jang, A., Dunlap, D., Kim, Y.-H., & Lee, S. W. (2026) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3772318.3791056 · Cited by 36
Maier, S., Schneider, M., & Feuerriegel, S. (2026) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3772318.3791185 · Cited by 26
While We Wait... How Users Perceive Waiting Times and Generation Cues during AI Image Generation
Lee, H. M., Yadav, D., Lee, S., Govindarazan, K., Chen, C., & Sundar, S. S. (2025) · CHI Conference Extended Abstracts on Human Factors in Computing Systems (CHI EA) · doi:10.1145/3706599.3719725 · Cited by 13
Kim, N., & Ji, Y. G. (2026) · International Journal of Human-Computer Studies · doi:10.1016/j.ijhcs.2026.103771 · Cited by 12
Metadata
- Authors
- Joshi, N., & Vogel, D.
- Affiliations
- University of Waterloo (CA)
- Venue
- Proceedings of the 7th ACM Conference on Conversational User Interfaces
- Publisher
- ACM
- Published
- 2025-07-07
- Pages
- 1-17
- OpenAlex
- W4412055679
- Cited by
- 50
- References
- 49
- Citations by year
- 2026: 41 · 2025: 9
- FWCI
- 30.93
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- ACM Conference on Conversational User Interfaces (CUI) (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- Natural Sciences and Engineering Research Council of Canada (RGPIN-2024-03827)
- Topics
- Physical Sciences › Computer Science › Human-Computer Interaction › Innovative Human-Technology Interaction
- Physical Sciences › Computer Science › Human-Computer Interaction › Usability and User Interface Design
- Physical Sciences › Computer Science › Computer Graphics and Computer-Aided Design › Digital Media and Visual Art
- Keywords
- psychological ownership, prompt length, short story writing, user interface design, human-AI collaboration
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-11-14, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-11-14 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Writing with AI Lowers Psychological Ownership, but Longer Prompts Can Help. https://app.cochat.ai/papers/10.1145/3719160.3736608
- The paper
- Joshi, N., & Vogel, D. (2025). Writing with AI Lowers Psychological Ownership, but Longer Prompts Can Help. Proceedings of the 7th ACM Conference on Conversational User Interfaces. https://doi.org/10.1145/3719160.3736608
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.
