Is Your Prompt Detailed Enough? Exploring the Effects of Prompt Coaching on Users' Perceptions, Engagement, and Trust in Text-to-Image Generative AI Tools
Open accessThe DOI matches its Crossref record; 53 of the 53 references it cites were found in a registry or an index.
Abstract
Prompts are the primary medium for interacting with generative AI tools. However, users often lack sufficient prompt literacy and motivation to fully benefit from these tools. To address this, we explore whether introducing prompt coaching into a chatbot-based generative AI interface can influence users’ perceptions and engagement of prompting, and further affect their trust in the system. In a user study (N = 132), we found that prompt coaching encourages users to specify more details in their prompts, even though over half initially believed their prompts were sufficient. Furthermore, the coach increased users’ cognitive elaboration, which was associated with higher perceived trust calibration. However, prompt coaching did not significantly enhance UX, although users in the coaching absent condition expressed a strong need for prompt assistance for better user experience. These findings have practical implications for the design of trustworthy and responsible generative AI interfaces.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Although generative systems are stateless in that they do not retain context across interactions unless explicitly prompted [107,108]”
From AI Agents vs. Agentic AI: A Conceptual taxonomy, applications and challenges (2026) · cites this paper as [108]
Citations
Cites
Evaluating Structural Equation Models with Unobservable Variables and Measurement Error
Fornell, C., & Larcker, D. F. (1981) · Journal of Marketing Research · doi:10.2307/3151312 · Cited by 109,314
Evaluating structural equation models with unobservable variables and measurement error
Fornell, C., & Larcker, D. F. (1981) · Journal of Marketing Research · doi:10.1177/002224378101800104 · Cited by 109,190 · cited 1×
“According to the criteria from [14], convergent validity is confirmed when the average variance extracted (AVE) is above 0.50 and composite reliability (CR) surpasses 0.70, indicating internal consistency.”
InfluentialPerceived usefulness, perceived ease of use, and user acceptance of information technology
Davis, F. D. (1989) · MIS Quarterly · doi:10.2307/249008 · Cited by 68,483 · cited 2×
“Adopting the perceived ease of use scale from prior research [13], participants were asked to rate four statements, such as “I find it easy to get FaceFrame AI to do what I want it to do” and “FaceFrame AI provides helpful guidance in performing tasks” ( M = 5.35, SD = 1.23, Cronbach’s 𝛼 = .84).”
InfluentialThe Elaboration Likelihood Model of Persuasion
Petty, R. E., & Cacioppo, J. T. (1986) · Advances in Experimental Social Psychology · doi:10.1016/s0065-2601(08)60214-2 · Cited by 8,822 · cited 2×
“To gauge thoughtful user engagement with AI systems, we focus on cognitive elaboration, a proxy for systematic information processing [49].”
Trust in Automation: Designing for Appropriate Reliance
Lee, J. D., & See, K. A. (2004) · Human Factors: The Journal of the Human Factors and Ergonomics Society · doi:10.1518/hfes.46.1.50_30392 · Cited by 6,605 · cited 3×
“We define AI trust as the degree to which users believe the AI system will act on users’ behalf under uncertainty [29].”
InfluentialAgarwal, R., & Karahanna, E. (2000) · MIS Quarterly · doi:10.2307/3250951 · Cited by 4,726 · cited 1×
“By focusing on the stage of absorption, we define user engagement with prompting as the state of deep involvement with prompting during AI interaction, characterized by temporal dissociation, focused immersion, heightened enjoyment, control, and curiosity [1].”
Machines and Mindlessness: Social Responses to Computers
Nass, C., & Moon, Y. (2000) · Journal of Social Issues · doi:10.1111/0022-4537.00153 · Cited by 4,022 · cited 1×
“Considering that individuals tend to mindlessly treat AI as social actors [38], and coaching users to specify their prompts is to help users generate better outcomes, we predict that prompt coaching will trigger the helper heuristic.”
(2006) · doi:10.1080/01449290500330331 · Cited by 3,026
Development of an instrument measuring user satisfaction of the human-computer interface
Chin, J. P., Diehl, V. A., & Norman, L. K. (1988) · Proceedings of the SIGCHI conference on Human factors in computing systems - CHI '88 · doi:10.1145/57167.57203 · Cited by 1,609 · cited 1×
“Using the User Interface Satisfaction scale [9], we evaluated users’ overall experience with FaceFrame AI through six pairs of adjectives, such as “terrible” vs. “wonderful,” and “difficult” vs. “easy.”
Artificial hallucinations in ChatGPT: implications in scientific writing
Alkaissi, H., & McFarlane, S. I. (2023) · Cureus · doi:10.7759/cureus.35179 · Cited by 1,540
Cognitive and affective trust in service relationships
Johnson, D., & Grayson, K. (2005) · Journal of Business Research · doi:10.1016/s0148-2963(03)00140-1 · Cited by 1,466 · cited 4×
“As a multi-dimensional construct [21, 53], it can have separate cognitive, affective, and behavioral facets.”
InfluentialWhy Johnny Can’t Prompt: How Non-AI Experts Try (and Fail) to Design LLM Prompts
Zamfirescu-Pereira, J. D., Wong, R. Y., Hartmann, B., & Yang, Q. (2023) · Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3544548.3581388 · Cited by 1,274
Shifting attention to accuracy can reduce misinformation online
Pennycook, G., Epstein, Z., Mosleh, M., Arechar, A. A., Eckles, D., & Rand, D. G. (2021) · Nature · doi:10.1038/s41586-021-03344-2 · Cited by 1,155 · cited 1×
“Previous research shows that system-initiated assistance can affect users’ perceptions and behaviors [46, 47, 51].”
Digital Media, Youth, and Credibility
Metzger, M. J., & Flanagin, A. J. (2007) · Cited by 1,060
O’Brien, H. L., Cairns, P., & Hall, M. (2018) · International Journal of Human-Computer Studies · doi:10.1016/j.ijhcs.2018.01.004 · Cited by 1,014
Ashfaq, M., Yun, J., Yu, S., & Loureiro, S. M. C. (2020) · Telematics and Informatics · doi:10.1016/j.tele.2020.101473 · Cited by 987
Design Guidelines for Prompt Engineering Text-to-Image Generative Models
Liu, V., & Chilton, L. B. (2022) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3491102.3501825 · Cited by 639
Design Guidelines for Prompt Engineering Text-to-Image Generative Models
Liu, V., & Chilton, L. B. (2021) · arXiv · doi:10.48550/arxiv.2109.06977 · Cited by 635
Formalizing Trust in Artificial Intelligence: Prerequisites, Causes and Goals of Human Trust in AI
Jacovi, A., Marasović, A., Miller, T., & Goldberg, Y. (2021) · Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency · doi:10.1145/3442188.3445923 · Cited by 545
Explicating Web Site Interactivity: Impression Formation Effects in Political Campaign Sites
Sundar, S. S., Kalyanaraman, S., & Brown, J. (2003) · Communication Research · doi:10.1177/0093650202239025 · Cited by 447
Self-citationKruger, J., Wirtz, D., Van Boven, L., & Altermatt, T. W. (2004) · Journal of Experimental Social Psychology · doi:10.1016/s0022-1031(03)00065-9 · Cited by 425 · cited 1×
“Notably, the perceived effort of prompting was associated with higher perceived AI performance quality, indicating the invocation of the effort heuristic: “if I put effort into designing my prompts, the output of the AI must be good” [27].”
(2018) · Singapore Management University Institutional Knowledge (InK) (Singapore Management University) · Cited by 390
Sundar, S. S., Jia, H., Waddell, T. F., & Huang, Y. (2015) · The Handbook of the Psychology of Communication Technology · doi:10.1002/9781118426456.ch3 · Cited by 361 · cited 1×
“Grounded in the theoretical model of human-AI interaction (HAII; [56]), which is based on the Theory of Interactive Media Effects (TIME; [58]), this study explores how prompt coaching affects users’ trust in AI systems through both the cue route and the action route.”
Self-citationThe CLEAR path: A framework for enhancing information literacy through prompt engineering
Lo, L. S. (2023) · The Journal of Academic Librarianship · doi:10.1016/j.acalib.2023.102720 · Cited by 351
Sundar, S. S., Bellur, S., Oh, J., Jia, H., & Kim, H.-S. (2016) · Communication Research · doi:10.1177/0093650214534962 · Cited by 293
Self-citationPrompt Engineering for ChatGPT: A Quick Guide to Techniques, Tips, and Best Practices
Ekin, S. (2026) · ASEE Journal of Engineering Technology · doi:10.18260/jet.43.1.5 · Cited by 261
Dabholkar, P. A., & Sheng, X. (2012) · The Service Industries Journal · doi:10.1080/02642069.2011.624596 · Cited by 256 · cited 1×
“Following previous studies on performance-based trust [11, 45], we adopted five items to measure cognitive trust in AI systems.”
Measuring Trust In Advertising
Soh, H., Reid, L. N., & King, K. W. (2009) · Journal of Advertising · doi:10.2753/joa0091-3367380206 · Cited by 250
Lallemand, C., Gronier, G., & Koenig, V. (2015) · Computers in Human Behavior · doi:10.1016/j.chb.2014.10.048 · Cited by 248
Kim, K. J., Park, E., & Shyam Sundar, S. (2013) · Computers in Human Behavior · doi:10.1016/j.chb.2013.02.009 · Cited by 234 · cited 3×
“Informed by prior studies [26, 31, 42], we used six items to evaluate the benefit heuristic.”
Self-citationStudying Heuristic‐Systematic Processing of Risk Communication
Kahlor, L., Dunwoody, S., Griffin, R. J., Neuwirth, K., & Giese, J. (2003) · Risk Analysis · doi:10.1111/1539-6924.00314 · Cited by 230 · cited 1×
“Participants were asked to indicate their level of agreement with four statements adapted from previous studies [23, 48].”
The design of guided learner-adaptable scaffolding in interactive learning environments
Jackson, S. L., Krajcik, J., & Soloway, E. (1998) · Proceedings of the SIGCHI conference on Human factors in computing systems - CHI '98 · doi:10.1145/274644.274672 · Cited by 203
Trust in public-sector senior management
Albrecht, S., & Travaglione, A. (2003) · The International Journal of Human Resource Management · doi:10.1080/09585190210158529 · Cited by 200 · cited 1×
“Behavioral trust, resulting from both cognitive and affective trust, involves users taking actions based on AI information [2, 53].”
Oh, J., Bellur, S., & Sundar, S. S. (2018) · Communication Research · doi:10.1177/0093650215600493 · Cited by 172
Self-citationDesigning for Responsible Trust in AI Systems: A Communication Perspective
Liao, Q. V., & Sundar, S. S. (2022) · 2022 ACM Conference on Fairness Accountability and Transparency · doi:10.1145/3531146.3533182 · Cited by 165
Self-citationInvolvement with Local Television News Cognitive and Emotional Dimensions
PERSE, E. M. (1990) · Human Communication Research · doi:10.1111/j.1468-2958.1990.tb00222.x · Cited by 162 · cited 1×
“Participants were asked to indicate their level of agreement with four statements adapted from previous studies [23, 48].”
Bechwati, N. N., & Xia, L. (2003) · Journal of Consumer Psychology · doi:10.1207/153276603768344852 · Cited by 138
Is the influence of privacy and security on online trust the same for all type of consumers?
Riquelme, I. P., & Román, S. (2014) · Electronic Markets · doi:10.1007/s12525-013-0145-3 · Cited by 124
Calling BS on ChatGPT: Reflections on AI as a Communication Source
Sundar, S. S., & Liao, M. (2023) · Journalism & Communication Monographs · doi:10.1177/15226379231167135 · Cited by 113
Self-citationAccuracy and social motivations shape judgements of (mis)information
Rathje, S., Roozenbeek, J., Van Bavel, J. J., & van der Linden, S. (2023) · Nature Human Behaviour · doi:10.1038/s41562-023-01540-w · Cited by 110
Prompt Engineering For ChatGPT: A Quick Guide To Techniques, Tips, And Best PracticesPreprint
Ekin, S. (2023) · doi:10.36227/techrxiv.22683919.v2 · Cited by 95
Cohn, M., Pushkarna, M., Olanubi, G. O., Moran, J. M., Padgett, D., Mengesha, Z., & Heldreth, C. (2024) · Extended Abstracts of the CHI Conference on Human Factors in Computing Systems · doi:10.1145/3613905.3650818 · Cited by 71
Imanova, N. M. (2018) · Journal of applied linguistics and language research · Cited by 64
Online Privacy Heuristics that Predict Information Disclosure
Sundar, S. S., Kim, J., Rosson, M. B., & Molina, M. D. (2020) · Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3313831.3376854 · Cited by 61 · cited 2×
“Previous research highlights that the give-and-take interaction between two actors is likely to trigger the reciprocity heuristic, which can be manifested as the belief that “I should give benefits to those who give me benefit” [61].”
Self-citationMuntinga, D. G., Moorman, M., Verlegh, P. W. J., & Smit, E. G. (2017) · Routledge eBooks · doi:10.4324/9781315623252 · Cited by 47
Which is better? Exploring Prompting Strategy For LLM-based Metrics
Kim, J., Lee, S., Han, S. H., Park, S.-Y., Lee, J., Jeong, K., & Kang, P. (2023) · Workshop on Evaluation and Comparison of NLP Systems (Eval4NLP) · doi:10.18653/v1/2023.eval4nlp-1.14 · Cited by 47
Self-citationBritish Journal of Applied Science & Technology
(2017) · British Journal of Applied Science & Technology · doi:10.9734/bjast · Cited by 45
Depleted egos and affirmed selves: The two faces of customization
Kang, H., & Sundar, S. S. (2013) · Computers in Human Behavior · doi:10.1016/j.chb.2013.05.018 · Cited by 36 · cited 3×
“When the control heuristic is activated, users tend to believe that “the more user control, the better the outcome” [24].”
Self-citationWhen the machine learns from users, is it helping or snooping?
Lee, S., Moon, W.-K., Lee, J.-G., & Sundar, S. S. (2022) · Computers in Human Behavior · doi:10.1016/j.chb.2022.107427 · Cited by 36
Self-citationChen, C., & Sundar, S. S. (2023) · Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3544548.3580805 · Cited by 36
Self-citationYeh, S.-F., Wu, M.-H., Chen, T.-Y., Lin, Y.-C., Chang, X., Chiang, Y.-H., & Chang, Y.-J. (2022) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3491102.3501941 · Cited by 35
Prompt Literacy: A Pivotal Educational Skill in the Age of AI
Gattupalli, S., Maloy, R. W., & Edwards, S. (2023) · Scholarworks (University of Massachusetts Amherst) · doi:10.7275/3498-wx48 · Cited by 13
Research works printed but not linked: 12
- 10.1207/S15327663JCP13-1&2_12 · doi:10.1207/s15327663jcp13-1&2_12 · Paper page
- Tiago Bianchi. 2024. Interest in generative AI on Google searches from February 2022 to June 2024 worldwide. statista. Retrieved from https://www.statista.com/statistics/1367868/generative-ai-google-searches-worldwide/
- Hai Dang, Lukas Mecke, Florian Lehmann, Sven Goller, and Daniel Buschek. 2022. How to prompt? Opportunities and challenges of zero- and few-shot learning for human-AI interaction in creative applications of generative models. Retrieved June 6, 2024 from http://arxiv.org/abs/2209.01390 · arXiv
- Sangwook Lee. 2024. Prompting in Generative AI: The effects of providing user control over display of advertisements in generative AI response. Pennsylvania State University, State College, PA, USA.
- Mengqi Liao. 2024. Does conversationality and verifiability affect trust in content generated by ChatGPT? Pennsylvania State University, State College, PA, USA.
- Robert W. Maloy and Sai Gattupalli. 2024. Prompt literacy. EdTechnica (2024), 209–215.
- Ethan Mollick. 2024. Co-intelligence: Living and working with AI. Portfolio.
- Heather O'Brien. 2016. Theoretical perspectives on user engagement. Why engagement matters: Cross-disciplinary perspectives of user engagement in digital media (2016), 1–26.
- Joe O'Connor and Jacob Andreas. 2021. What context features can transformer language models use? arXiv preprint arXiv:2106.08367 (2021). · arXiv
- Gordon Pennycook and David G. Rand. 2021. The psychology of fake news. Trends in cognitive sciences 25, 5 (2021), 388–402.
- Sheizf Rafaeli. 1988. From new media to communication. Sage annual review of communication research: Advancing communication science 16, 1 (1988), 110–134.
- 10.1093/jcmc · doi:10.1093/jcmc
Cited by
AI Agents vs. Agentic AI: A Conceptual Taxonomy, Applications and Challenges
Sapkota, R., Roumeliotis, K. I., & Karkee, M. (2025) · arXiv · doi:10.48550/arxiv.2505.10468 · Cited by 295
“• Prompt Dependency and Statelessness: Although generative systems are stateless in that they do not retain context across interactions unless explicitly prompted [107,108], recent advancements like GPT-4.”
AI Agents vs. Agentic AI: A Conceptual taxonomy, applications and challenges
Sapkota, R., Roumeliotis, K. I., & Karkee, M. (2026) · Information Fusion · doi:10.1016/j.inffus.2025.103599 · Cited by 671
“• Prompt Dependency and Statelessness: Although generative systems are stateless in that they do not retain context across interactions unless explicitly prompted [107,108], recent advancements like GPT-4.”
AI Agents vs. Agentic AI: A Conceptual Taxonomy, Applications and Challenges
Sapkota, R., Roumeliotis, K. I., & Karkee, M. (2025) · SuperIntelligence - Robotics - Safety & Alignment · doi:10.70777/si.v2i3.15161 · Cited by 186
A typology of physician input approaches to using AI chatbots for clinical decision-making
Siden, R., Kerman, H. M., Gallo, R. J., Cool, J. A., Hom, J., Goh, E., Ahuja, N. K., Heidenreich, P. A., Shieh, L. Y., Yang, D. X., et al. (2025) · npj Digital Medicine · doi:10.1038/s41746-025-02184-y · Cited by 15
“Promptengineeringbestpracticesarenotalwaysintuitivetolay or untrained users, who may employ a variety of different input styles when interacting with LLM chatbots 10,11 .”
Chen, C., & Zhao, X. (2025) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2025.2530762 · Cited by 10
“Following previous studies (Chen, Lee, et al., 2024; Perse, 1990), we used four items to measure users ’ cognitive elaboration, such as “ I thought about what should be displayed about LGBTQ þ patients in the flyer ” and “ I thought about the needs of the LGBTQ þ patients over and over again.”
InfluentialSelf-citation
Metadata
- Authors
- Chen, C., Lee, S., Jang, E., & Sundar, S. S.
- Affiliations
- Elon University (US)
- University of Colorado Boulder (US)
- Pennsylvania State University (US)
- Venue
- Proceedings of the Second International Symposium on Trustworthy Autonomous Systems
- Publisher
- ACM
- Published
- 2024-09-16
- Pages
- 1-12
- OpenAlex
- W4402659120
- Cited by
- 21
- References
- 50
- Citations by year
- 2026: 10 · 2025: 11
- FWCI
- 6.75
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- International Symposium on Trustworthy Autonomous Systems (TAS) (publishedVersion, cc-by)
- 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
- prompt literacy, text-to-image generative AI, chatbot interface, user trust, user engagement, cognitive elaboration, user experience
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.
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
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.