Pith. sign in

REVIEW 4 major objections 4 minor 55 references

Methodological Foundations for AI-Driven Survey Question Generation

T0 review · 4 major / 4 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read The paper proposes that AI-generated survey prompts can be refined before deployment by having one language model simulate a student and another generate follow-up questions, and shows this loop improves prompt quality but does not…

desk verdict A useful, honest methods paper on AI survey generation whose central refinement claim is undermined by a simultaneous model upgrade, and the statistical validation needs work before it should be trusted. read the letter →

arxiv 2505.01150 v1 pith:GA2NOKJ3 submitted 2025-05-02 cs.CY

Ted K. Mburu , Kangxuan Rong , Campbell J. McColley , Alexandra Werth This is my paper · ORCID
classification cs.CY
keywords GenerativeAILargeLanguageModelsActivityTheoryAdaptiveSurveysEngineeringEducationSyntheticQuestion-ResponseAnalysisPromptSurveyValidation
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 is trying to establish a pre-deployment testing method for AI-generated survey questions: one language model writes follow-up questions while another plays a student persona, and the resulting AI-to-AI conversations are mined for tone, similarity, and structure. This Synthetic Question-Response Analysis is claimed to be practically useful for iterative prompt refinement and initial validation, though the authors state plainly that it cannot replicate human response variability. The authors support this by comparing 4,000 simulated question-response pairs with 1,272 pairs from human participants in two engineering courses. A sympathetic reader should come away with a specific division of labor: AI-to-AI simulation can catch and fix obvious survey-prompt defects cheaply, while human piloting remains the checkpoint for real response behavior.

What carries the argument

The load-bearing object is the Synthetic Question-Response Analysis (SQRA) framework: a repeated loop in which a fixed initial question and a persona prompt elicit a simulated student response, that response is fed to the question-generation LLM to produce the next question, and the cycle repeats for four questions per conversation. The comparison and refinement machinery sits on three text-analysis metrics: VADER sentiment scores for tone, spaCy cosine similarity between each response and the follow-up question for alignment, and counts of conjunctions and punctuation as proxies for double-barreled questions and evaluative language. The same loop, run through five prompt-refinement cycles on 100 simulated conversations each, produced measurably shorter and more focused questions. Activity Theory functions as the interpretive scaffold that frames the simulated persona versus human student as different subjects within the same tool-mediated activity system.

What would settle it

Take the original and refined prompts from the paper's refinement section and have a panel of survey-methodology experts blindly rate the resulting questions for clarity, double-barreling, and evaluative bias. If the expert ratings show no improvement after SQRA refinement, or if VADER sentiment, cosine similarity, and conjunction counts do not correlate with expert judgments of question quality, the framework's central value claim collapses.

Watch

Extended reading notes

Core claim

The central claim is that large language models can generate adaptive, context-aware open-ended survey questions on a commercial survey platform, and that a synthetic loop, one LLM generating questions and a persona-driven LLM answering them, can serve as a testbed for refining those questions before human use. The paper supports this by comparing AI-to-AI interactions against AI-to-human interactions using sentiment analysis, cosine similarity, and question-structure counts. The comparison shows both similarities and systematic gaps: AI-to-AI questions run longer, use more "and" and periods, and track the preceding response more closely, while human interactions are more variable and include terse responses that pull similarity scores down. The authors therefore conclude that SQRA is useful for prompt refinement and initial validation but is not a substitute for human responses, and that AI-driven surveys remain complementary to, not replacements for, human piloting.

Load-bearing premise

The load-bearing premise is that VADER sentiment scores, spaCy cosine similarity, and conjunction and punctuation counts are valid proxies for survey-question quality; the paper reports no human ratings of the refined prompts, so if these proxies do not track what makes a question effective for students, the framework's central value claim collapses.

Editorial extensions

If this is right

  • Any AI-driven survey deployed on a commercial platform can be pre-screened with the SQRA loop, so researchers can identify wordy, double-barreled, or evaluative prompts before any human sees them.
  • Iterative SQRA refinement produces measurable structural improvements, meaning prompt-engineering cycles can be run cheaply and safely at scale without consuming student or researcher time.
  • Because the simulated responses are systematically longer and less variable than human responses, findings from SQRA should be treated as provisional until confirmed by human piloting.
  • The metric set of sentiment, response-question similarity, and structural counts gives future validation studies a reusable baseline for comparing AI-to-AI and AI-to-human conditions.
  • The built-in fallback-question design makes AI-driven surveys robust to API outages, which is a necessary condition for classroom deployment.

Reading between the lines

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

  • A natural next experiment is to hand the original and refined prompts to human cognitive-interview panels; if their judgments disagree with the SQRA metrics, the framework's proxy measures are missing something.
  • The low-similarity outliers in the human data were driven by short, terse answers like "True," which suggests the persona prompts should deliberately include low-engagement response styles so the simulation stress-tests the least articulate respondents rather than the most eloquent ones.
  • The same loop could serve as a regression test whenever the underlying model is upgraded: rerun a fixed corpus of simulated conversations through SQRA and flag changes in question length, sentiment, and alignment before re-releasing the survey.
  • The framework's real ceiling is likely as a filter before human piloting, not as a validation of data quality; if applied beyond prompt refinement, it could give false confidence about how human respondents will actually behave.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper proposes and demonstrates a methodological framework, SQRA, for using LLM-simulated student personas to generate, test, and iteratively refine survey questions before deployment with human participants. The authors integrate OpenAI's ChatGPT API with Qualtrics, describe prompt engineering informed by Activity Theory and SSRL, and compare 1,000 AI-to-AI interactions with 318 AI-to-human interactions using sentiment analysis, cosine similarity, and structural metrics. They report that AI-generated questions are contextually relevant but suffer from redundancy, length, double-barreled phrasing, and jargon; they also report large differences between AI-to-AI and AI-to-human interactions and claim that iterative SQRA refinement improved question quality. The paper is transparent about its central limitation that SQRA cannot replicate human response variability and explicitly calls for further validation. The main contributions are a practical implementation guide, a proposed pre-deployment validation workflow, and an empirical comparison that is preliminary rather than confirmatory.

Significance. If the central claims hold, SQRA would be a useful pre-deployment tool for refining AI-generated survey prompts, filling a genuine methodological gap in engineering education research by providing a scalable, lower-risk alternative to human piloting. The paper gives useful, concrete implementation details for integrating LLM APIs with survey platforms, and it is commendably transparent about the limitation that AI-to-AI interactions do not reproduce human response distributions. The paper also reports effect sizes alongside p-values for several comparisons, which is good practice. However, the significance is conditional on two things: the validity of the text-analysis metrics as proxies for question quality, and a controlled demonstration that SQRA-driven refinement, rather than a simultaneous model upgrade, caused the observed improvements. If those conditions are met, this could become a useful methods reference; as it stands, the evidence is suggestive but not decisive.

major comments (4)
  1. [Section 4.4 and Figure 11] The evidence for the paper's central claim that SQRA-driven iterative refinement improves question quality is confounded. Section 4.4 states that, as part of the refinement process, 'we updated the system-prompting AI from ChatGPT-3.5 to GPT-4o' and that 'we observed compounding improvements resulting from both the prompt refinements and the model upgrade.' Figure 11 compares the original prompt with the final refined prompt, but because the generation model changed at the same time, the observed reductions in character count, periods, conjunctions, and question marks cannot be attributed to prompt refinement alone. A controlled comparison that holds the model fixed, or a factorial design that separates prompt effects from model effects, is required before the practical-utility claim in the abstract and conclusions is supported.
  2. [Section 3.4.3 and Section 4.3] The structural analyses use paired t-tests on samples that are not paired. Section 3.4.3 says 'Paired t-tests were used to determine if frequency of conjunction and punctuation usage significantly differed between AI-to-AI and AI-to-Human interactions,' but the two samples are independent: 1,000 AI-to-AI conversations versus 318 AI-to-human interactions. A paired test requires a natural one-to-one matching that is not described and cannot be justified from the design. In addition, the many significance tests across sentiment categories, similarity scores, symbol counts, and character counts are not corrected for multiple comparisons. The reported p-values (e.g., the **** notations in Figures 8–11) should be recalculated with an appropriate independent-samples test and adjusted for multiplicity before inferential claims are made.
  3. [Section 3.4 and Section 4.4] The central utility claim for SQRA depends on treating VADER sentiment, spaCy cosine similarity, and conjunction/punctuation counts as valid proxies for question quality and alignment with research goals. This assumption enters in Section 3.4 and drives the refinement conclusions in Section 4.4. No human ratings, expert review, think-aloud data, or other criterion validation of these metrics is reported, and the refined prompts from Section 4.4 were never administered to human participants. If these text-analytic proxies do not track what makes a survey question effective for students, the claim that SQRA provides 'practical utility for prompt refinement and initial validation of AI-generated survey content' (Abstract and Section 5.2) is unsupported. At minimum, the paper should be reframed as an illustration of a workflow whose validity requires future criterion-based validation.
  4. [Section 5.2 and Appendix A1] The paper describes a randomly assigned control group (Group B, predefined questions) in Section 3.1 and Appendix A1, but Section 5.2 states that analysis of the differential impact between Group A and Group B is planned for future work. Consequently, claims about the effectiveness of AI-driven surveys relative to traditional surveys—such as the statement in the Discussion that AI-driven surveys 'fill a unique role' and 'balance scale with personalization'—are not supported by the presented data. The control comparison should either be reported or the claims about comparative effectiveness should be explicitly deferred.
minor comments (4)
  1. [Throughout] There are several typos and inconsistent spellings: 'rappor' (Section 1), 'doubled-barreled' (Sections 4.1 and 4.4), 'Simulated Participate Rule' (Section 3.3.2, should be 'Participant'), 'Quatrics' (Appendix A2), 'prediefined' (Appendix A3), and 'V ADER' (Section 3.4.1). A careful proofreading pass is needed.
  2. [Section 3.5 and Table 2] Sample-size reporting is inconsistent. The text mentions 'approximately 400 physics and engineering students' and a 'third-year, one-credit seminar with 55 biomedical engineering students,' while Table 2 reports 388 and 40 for the physics courses and Table 1 reports n = 38 for the biomedical course. The paper should clarify which numbers correspond to survey completion, research consent, and demographic response, and should align all reported denominators with the 318 AI-to-human interactions used in the analysis.
  3. [Section 4.3 and Figure 10] The test descriptions are inconsistent. The text says a Mann–Whitney U test was used for Figure 10A and a paired t-test for Figure 10B, while the figure caption repeats this; however, Section 3.4.3 describes only the paired t-test. Please unify the statistical description and ensure the caption and the text are consistent.
  4. [Section 3.2 and Figure 6] The SQRA process description says 'Steps 1-6 repeated iteratively,' but Figure 6 does not have numbered step labels in the text. Adding explicit step numbers to the figure would make the workflow easier to follow.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: SQRA claims are empirical and externally anchored; the model-upgrade confound is a validity threat, not a circularity.

full rationale

The paper contains no fitted constants, no equations, and no claimed first-principles derivation whose output is equivalent to its input. The SQRA framework's central claims are empirical: AI-to-AI questions are compared against AI-to-human questions actually collected from students, and the paper explicitly concludes that synthetic responses do not replicate human variability. The refinement evidence in §4.4 is beset by a real confound—the system-prompting AI was simultaneously upgraded from ChatGPT-3.5 to GPT-4o—so the observed Figure 11 differences cannot be cleanly attributed to SQRA-driven prompt refinement. That is an internal-validity problem, but it is not circularity, because the comparison is not forced by definition or by a fitted parameter. The only self-citation, Werth et al. (2023), supports background claims about student reflection styles and brief responses and is not load-bearing for the SQRA framework. The AI-to-AI evaluation loop is self-referential in the broad sense that an LLM is used to judge LLM-generated content, but the paper anchors the framework against real human interactions and does not rename its own assumptions as predictions. No circular step meeting the quoted-evidence standard is present.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new physical or conceptual entities; SQRA is an analytic workflow, not an entity with independent evidence requirements. The framework's load rests on domain assumptions about text-analysis proxies and about LLM personas standing in for students.

free parameters (3)
  • conversation_length = 4 questions per conversation
    Set by the authors in Section 3.2; determines how many AI-generated questions are produced per interaction and shapes all comparisons.
  • simulated_conversation_count = 1,000 conversations
    Chosen in Section 3.2 for scale; it affects statistical power but is not fitted to any outcome.
  • model_version = ChatGPT-3.5 for main analyses; GPT-4o during prompt refinement
    The model choice shapes all generated text; the paper does not isolate model effects from prompt effects in the main comparison.
assumptions (5)
  • domain assumption VADER sentiment scores are a valid measure of tone for AI-generated educational survey questions.
    Used in Section 3.4.1; VADER was developed for social-media text and its applicability to AI-generated educational prompts is assumed.
  • domain assumption spaCy en_core_web_lg cosine similarity captures meaningful semantic alignment between a student response and the AI follow-up question.
    Used in Section 3.4.2; no evidence is provided that these embedding similarities correspond to human judgments of relevance.
  • ad hoc to paper Conjunction and punctuation frequencies are valid proxies for double-barreled questions and evaluative feedback.
    Section 3.4.3 explicitly uses these counts as proxies; the paper does not validate that the proxies track the intended quality constructs.
  • domain assumption LLM-simulated personas can represent the target student population closely enough to stress-test survey questions.
    Section 3.3.2 assumes persona prompts emulate student reflection; Section 5 itself reports large, statistically significant differences from human responses, making this a fragile assumption.
  • domain assumption Activity Theory is an appropriate theoretical lens for interpreting AI-mediated survey interactions.
    Used throughout as the framing lens; it guides prompt design and interpretation but is not empirically constrained by the data.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Methodological Foundations for AI-Driven Survey Question Generation." pith.science (2026). https://pith.science/paper/GA2NOKJ3

@misc{pith2026250501150,
  author       = {Pith},
  title        = {Pith review of: Methodological Foundations for AI-Driven Survey Question Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GA2NOKJ3}},
  note         = {Machine review of arXiv:2505.01150}
}
read the original abstract

This paper presents a methodological framework for using generative AI in educational survey research. We explore how Large Language Models (LLMs) can generate adaptive, context-aware survey questions and introduce the Synthetic Question-Response Analysis (SQRA) framework, which enables iterative testing and refinement of AI-generated prompts prior to deployment with human participants. Guided by Activity Theory, we analyze how AI tools mediate participant engagement and learning, and we examine ethical issues such as bias, privacy, and transparency. Through sentiment, lexical, and structural analyses of both AI-to-AI and AI-to-human survey interactions, we evaluate the alignment and effectiveness of these questions. Our findings highlight the promise and limitations of AI-driven survey instruments, emphasizing the need for robust prompt engineering and validation to support trustworthy, scalable, and contextually relevant data collection in engineering education.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 38 canonical work pages

  1. [1]

    @esa (Ref

    \@ifxundefined[1] #1\@undefined \@firstoftwo \@secondoftwo \@ifnum[1] #1 \@firstoftwo \@secondoftwo \@ifx[1] #1 \@firstoftwo \@secondoftwo [2] @ #1 \@temptokena #2 #1 @ \@temptokena \@ifclassloaded agu2001 natbib The agu2001 class already includes natbib coding, so you should not add it explicitly Type <Return> for now, but then later remove the command n...

  2. [2]

    \@lbibitem[] @bibitem@first@sw\@secondoftwo \@lbibitem[#1]#2 \@extra@b@citeb \@ifundefined br@#2\@extra@b@citeb \@namedef br@#2 \@nameuse br@#2\@extra@b@citeb \@ifundefined b@#2\@extra@b@citeb @num @parse #2 @tmp #1 NAT@b@open@#2 NAT@b@shut@#2 \@ifnum @merge>\@ne @bibitem@first@sw \@firstoftwo \@ifundefined NAT@b*@#2 \@firstoftwo @num @NAT@ctr \@secondoft...

  3. [3]

    2E @APACcomma

    @open @close @open @close and [1] URL: #1 \@ifundefined chapter * 7 13 REFERENCES\@mkboth \@ifxundefined @sectionbib * \@mkboth * \@mkboth\@gobbletwo \@ifclassloaded amsart * \@ifclassloaded amsbook * \@ifxundefined @heading @heading NAT@ctr thebibliography [1] @ \@biblabel @NAT@ctr \@bibsetup #1 @NAT@ctr @ @openbib .11em \@plus.33em \@minus.07em 4000 400...

    work page 1999

  4. [4]

    @esa (Ref

    \@ifxundefined[1] #1\@undefined \@firstoftwo \@secondoftwo \@ifnum[1] #1 \@firstoftwo \@secondoftwo \@ifx[1] #1 \@firstoftwo \@secondoftwo [2] @ #1 \@temptokena #2 #1 @ \@temptokena \@ifclassloaded agu2001 natbib The agu2001 class already includes natbib coding, so you should not add it explicitly Type <Return> for now, but then later remove the command n...

  5. [5]

    B ! B )bbbĀ-[x PX֡o'+B>] <M-믿| BHQB_B ! B !EVff eCjjvqtρ kc y ҞcVݟ5j ; > :o! RTP B ! BH' |ϟ?[o;e ט-F . D,`mm[ !|j B ! B ) 222:< X(_NߌWX ; @[[[,0g B !S/!Oݻ8 4

    @open @close @open @close and [1] URL: #1 \@ifundefined chapter * 10 13 REFERENCES \@mkboth \@ifxundefined @sectionbib * \@mkboth * \@mkboth\@gobbletwo \@ifclassloaded amsart * 10 13 \@ifclassloaded amsbook * \@ifxundefined @heading @heading NAT@ctr thebibliography [1] @ \@biblabel @NAT@ctr \@bibsetup #1 @NAT@ctr @ @ @openbib .11em \@plus.33em \@minus.07e...

    arXiv 1978

  6. [6]

    write newline

    " write newline "" before.all 'output.state := FUNCTION blank.sep after.quote 'output.state := FUNCTION fin.entry doi empty output.state after.quoted.block = 'skip 'add.period if if write newline FUNCTION new.block output.state before.all = 'skip output.state after.quote = after.quoted.block 'output.state := after.block 'output.state := if if FUNCTION new...

  7. [7]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in "" FUNCTION format.date "" du...

  8. [8]

    write newline

    " write newline " cite write " FUNCTION editor.postfix editor num.names #1 > "( )" "( )" if FUNCTION editor.trans.postfix editor num.names #1 > "( )" "( )" if FUNCTION trans.postfix translator num.names #1 > "( )" "( )" if FUNCTION authors.editors.reflist.apa5 'field := 'dot := field num.names 'numnames := numnames 'format.num.names := format.num.names na...

Show all 55 references
  1. [9]

    APACrefauthors \ 2002 06

    adams_conceptual_2002 APACrefauthors Adams, S. APACrefauthors \ 2002 06 . A Conceptual Model For The Development And Assessment Of Teamwork A Conceptual Model For The Development And Assessment Of Teamwork . ( \ 7.30.1--7.30.9). ISSN: 2153-5965

    2002

  2. [10]

    \ Baiz, C R

    Alasadi APACrefauthors Alasadi, E A. \ Baiz, C R. APACrefauthors \ 2023 . Generative AI in Education and Research: Opportunities, Concerns, and Solutions Generative ai in education and research: Opportunities, concerns, and solutions . Journal of Chemical Education 100 8 2965-...

    2023 doi

  3. [11]

    \ Owusu Ansah, L

    Baidoo APACrefauthors Baidoo-Anu, D. \ Owusu Ansah, L. APACrefauthors \ 2023 Jan . Education in the Era of Generative Artificial Intelligence (AI): Understanding the Potential Benefits of ChatGPT in Promoting Teaching and Learning Education in the era of generative artificial ...

    2023

  4. [12]

    APACrefauthors \ 2000

    becker2000effect APACrefauthors Becker, L A. APACrefauthors \ 2000 . Effect size (ES) Effect size (es)

    2000

  5. [13]

    , Marx, R W

    blumenfeld1996learning APACrefauthors Blumenfeld, P C. , Marx, R W. , Soloway, E. \ Krajcik, J. APACrefauthors \ 1996 . Learning with peers: From small group cooperation to collaborative communities Learning with peers: From small group cooperation to collaborative communities...

    1996

  6. [14]

    , Kumaresh, N

    Bonta_Kumaresh_Janardhan_2019 APACrefauthors Bonta, V. , Kumaresh, N. \ Janardhan, N. APACrefauthors \ 2019 Jan. . A Comprehensive Study on Lexicon Based Approaches for Sentiment Analysis A comprehensive study on lexicon based approaches for sentiment analysis . Asian Journal ...

    2019 doi

  7. [15]

    , Das, D

    cambria2017practical APACrefauthors Cambria, E. , Das, D. , Bandyopadhyay, S. , Feraco, A. \ . APACrefauthors \ 2017 . A practical guide to sentiment analysis A practical guide to sentiment analysis \ ( 5). Springer

    2017

  8. [16]

    APACrefauthors \ 2024

    DeCosta APACrefauthors Costa, P I D. APACrefauthors \ 2024 . Sociocultural influence on engineering students’ collaborative design project: an Activity Theory perspective Sociocultural influence on engineering students’ collaborative design project: an activity theory perspect...

    2024 doi

  9. [17]

    , Nethken, C L R

    csavina2016assessing APACrefauthors Csavina, K R. , Nethken, C L R. \ Carberry, A. APACrefauthors \ 2016 . Assessing student understanding of reflection in engineering education Assessing student understanding of reflection in engineering education . 123rd ASEE Annual Conferen...

    2016

  10. [18]

    , Al-Hossami, E

    Dorodchi APACrefauthors Dorodchi, M. , Al-Hossami, E. , Benedict, A. \ Demeter, E. APACrefauthors \ 2019 . Using Synthetic Data Generators to Promote Open Science in Higher Education Learning Analytics Using synthetic data generators to promote open science in higher education...

    2019

  11. [19]

    APACrefauthors \ 1987

    Engestrom APACrefauthors Engeström, Y. APACrefauthors \ 1987 . Learning by Expanding: An Activity-Theoretical Approach to Developmental Research Learning by expanding: An activity-theoretical approach to developmental research \ ( 2 \ ). Cambridge University Press

    1987

  12. [20]

    , Hane, E

    franklin2018improving APACrefauthors Franklin, S V. , Hane, E. , Kustusch, M B. , Ptak, C. \ Sayre, E C. APACrefauthors \ 2018 . Improving retention Through metacognition Improving retention through metacognition . Journal of College Science Teaching 48 2 21--27

    2018

  13. [21]

    APACrefauthors \ 2011

    Georg APACrefauthors Georg, G. APACrefauthors \ 2011 . Activity Theory and its Applications in Software Engineering and Technology: Literature Search Results and Observations Activity theory and its applications in software engineering and technology: Literature search results...

    2011

  14. [22]

    a rvel \

    hadwin2011self APACrefauthors Hadwin, A F. , J \"a rvel \"a , S. \ Miller, M. APACrefauthors \ 2011 . Self-regulated, co-regulated, and socially shared regulation of learning Self-regulated, co-regulated, and socially shared regulation of learning . Handbook of self-regulation...

    2011

  15. [23]

    , Wakene, A D

    Hanna2023AssessingBias APACrefauthors Hanna, J J. , Wakene, A D. , Lehmann, C U. \ Medford, R J. APACrefauthors \ 2023 August . Assessing Racial and Ethnic Bias in Text Generation for Healthcare-Related Tasks by ChatGPT Assessing racial and ethnic bias in text generation for h...

    2023 doi

  16. [24]

    \ Thompson, C J

    Hite APACrefauthors Hite, R. \ Thompson, C J. APACrefauthors \ 2019 . Activity Theory as Theoretical Framework for Analyzing and Designing Global K-12 Collaborations in Engineering: A Case Study of a Thai-U.S. Elementary Engineering Project Activity theory as theoretical frame...

    2019

  17. [25]

    , Eikelenboom, M

    Horsfall APACrefauthors Horsfall, M. , Eikelenboom, M. , Draisma, S. \ Smit, J H. APACrefauthors \ 2021 . The Effect of Rapport on Data Quality in Face-to-Face Interviews: Beneficial or Detrimental? The effect of rapport on data quality in face-to-face interviews: Beneficial o...

    2021 doi

  18. [26]

    \ Horbach, S P J M

    Hosseini APACrefauthors Hosseini, M. \ Horbach, S P J M. APACrefauthors \ 2023 . Fighting reviewer fatigue or amplifying bias? Considerations and recommendations for use of ChatGPT and other large language models in scholarly peer review Fighting reviewer fatigue or amplifying...

    2023 doi

  19. [27]

    \ Lachenbruch, P A

    hsu2014paired APACrefauthors Hsu, H. \ Lachenbruch, P A. APACrefauthors \ 2014 . Paired t test Paired t test . Wiley StatsRef: statistics reference online

    2014

  20. [28]

    APACrefauthors \ 2025

    ilishkina2025rethinking APACrefauthors Ilishkina, D I. APACrefauthors \ 2025 . Rethinking the evaluation of educational intervention effectiveness through Activity Theory: A mobile app example Rethinking the evaluation of educational intervention effectiveness through activity...

    2025

  21. [29]

    , Katz, A S

    Johri APACrefauthors Johri, A. , Katz, A S. , Qadir, J. \ Hingle, A. APACrefauthors \ 2023 . Generative artificial intelligence and engineering education Generative artificial intelligence and engineering education . Journal of Engineering Education 112 3 572-577 . APACrefDOI ...

    2023 doi

  22. [30]

    , Misevic, D

    Kusters APACrefauthors Kusters, R. , Misevic, D. , Berry, H. , Cully, A. , Cunff, Y L. , Dandoy, L. Wehbi, F. APACrefauthors \ 2020 Nov . Interdisciplinary Research in Artificial Intelligence: Challenges and Opportunities Interdisciplinary research in artificial intelligence: ...

    2020

  23. [31]

    APACrefauthors \ 1978

    leontiev APACrefauthors Leontiev, A. APACrefauthors \ 1978 . Activity, Consciousness, and Personality. Activity, consciousness, and personality. Prentice Hall . APACrefURL http://marxists.org/archive/leontev/works/1978/index.htm APACrefURL [Leontiev, 1978]

    1978

  24. [32]

    \ Guba, E

    LincolnGuba APACrefauthors Lincoln, Y. \ Guba, E. APACrefauthors \ 1985 . Naturalistic Inquiry Naturalistic inquiry . Beverly Hills, Calif. SAGE Publications . APACrefURL https://books.google.com/books?id=2oA9aWlNeooC APACrefURL

    1985

  25. [33]

    , Chen, L

    lu2025machinelearningsyntheticdata APACrefauthors Lu, Y. , Chen, L. , Zhang, Y. , Shen, M. , Wang, H. , Wang, X. Wei, W. APACrefauthors \ 2025 . Machine Learning for Synthetic Data Generation: A Review. Machine learning for synthetic data generation: A review. APACrefURL https...

    2025 arXiv

  26. [34]

    , Ayre, J

    Mac2022ComparisonReadability APACrefauthors Mac, O. , Ayre, J. , Bell, K. , McCaffery, K. \ Muscat, D M. APACrefauthors \ 2022 December . Comparison of Readability Scores for Written Health Information Across Formulas Using Automated vs Manual Measures Comparison of readabilit...

    2022

  27. [35]

    , Gitonga, D

    education2014211A3 APACrefauthors Makewa, L N. , Gitonga, D. , Ngussa, B. , Njoroge, S. \ Kuboja, J. APACrefauthors \ 2014 . Frustration Factor in Group Collaborative Learning Experiences Frustration factor in group collaborative learning experiences . American Journal of Educ...

    2014 doi

  28. [36]

    \ Etkina, E

    may2002college APACrefauthors May, D B. \ Etkina, E. APACrefauthors \ 2002 . College physics students’ epistemological self-reflection and its relationship to conceptual learning College physics students’ epistemological self-reflection and its relationship to conceptual learn...

    2002

  29. [37]

    \ Najab, J

    mcknight2010mann APACrefauthors McKnight, P E. \ Najab, J. APACrefauthors \ 2010 . Mann-Whitney U Test Mann-whitney u test . The Corsini encyclopedia of psychology 1--1

    2010

  30. [38]

    APACrefauthors \ 2023

    Menekse APACrefauthors Menekse, M. APACrefauthors \ 2023 . Envisioning the future of learning and teaching engineering in the artificial intelligence era: Opportunities and challenges Envisioning the future of learning and teaching engineering in the artificial intelligence er...

    2023 doi

  31. [39]

    \ Holmes, W

    Miao APACrefauthors Miao, F. \ Holmes, W. APACrefauthors \ 2023 . Guidance for Generative AI in Education and Research Guidance for generative ai in education and research \ [book]. United Nations Educational, Scientific and Cultural Organization . APACrefDOI 10.54675/EWZM9535...

    2023 doi

  32. [40]

    , Delgado-Vera, C

    mite2018sentiment APACrefauthors Mite-Baidal, K. , Delgado-Vera, C. , Sol \' s-Avil \'e s, E. , Espinoza, A H. , Ortiz-Zambrano, J. \ Varela-Tapia, E. APACrefauthors \ 2018 . Sentiment analysis in education domain: A systematic literature review Sentiment analysis in education...

    2018

  33. [41]

    , Hong, J S

    park2020methodology APACrefauthors Park, K. , Hong, J S. \ Kim, W. APACrefauthors \ 2020 . A methodology combining cosine similarity with classifier for text classification A methodology combining cosine similarity with classifier for text classification . Applied Artificial I...

    2020

  34. [42]

    \ Schneiderheinze, A

    Russell APACrefauthors Russell, D L. \ Schneiderheinze, A. APACrefauthors \ 2005 . Understanding Innovation in Education Using Activity Theory Understanding innovation in education using activity theory . Journal of Educational Technology & Society 8 1 38--53 . APACrefURL [ 20...

    2005

  35. [43]

    \ Globerson, T

    SALOMON198989 APACrefauthors Salomon, G. \ Globerson, T. APACrefauthors \ 1989 . When teams do not function the way they ought to When teams do not function the way they ought to . International Journal of Educational Research 13 1 89-99 . APACrefDOI https://doi.org/10.1016/08...

    1989 doi

  36. [44]

    APACrefauthors \ 2021

    Sammut APACrefauthors Sammut, R. APACrefauthors \ 2021 . Strategies to improve response rates to web surveys: A literature review Strategies to improve response rates to web surveys: A literature review . International Journal of Nursing Studies 123 104058 . APACrefURL https:/...

    2021

  37. [45]

    \ Feinn, R

    sullivan2012using APACrefauthors Sullivan, G M. \ Feinn, R. APACrefauthors \ 2012 . Using effect size—or why the P value is not enough Using effect size—or why the p value is not enough . Journal of graduate medical education 4 3 279--282

    2012

  38. [46]

    , Esparza, D

    treibergs2022introductory APACrefauthors Treibergs, K A. , Esparza, D. , Yamazaki, J A. , Goebel, M. \ Smith, M K. APACrefauthors \ 2022 . How do introductory field biology students feel? Journal reflections provide insight into student affect How do introductory field biology...

    2022

  39. [47]

    , Hubertz, M

    VanCampenhout APACrefauthors Van Campenhout, R. , Hubertz, M. \ Johnson, B G. APACrefauthors \ 2022 . Evaluating AI-Generated Questions: A Mixed-Methods Analysis Using Question Data and Student Perceptions Evaluating ai-generated questions: A mixed-methods analysis using quest...

    2022

  40. [48]

    , Gijselaers, W H

    van2006social APACrefauthors Van den Bossche, P. , Gijselaers, W H. , Segers, M. \ Kirschner, P A. APACrefauthors \ 2006 . Social and cognitive factors driving teamwork in collaborative learning environments: Team learning beliefs and behaviors Social and cognitive factors dri...

    2006

  41. [49]

    \ Ang, G

    volet1998culturally APACrefauthors Volet, S E. \ Ang, G. APACrefauthors \ 1998 . Culturally mixed groups on international campuses: An opportunity for inter-cultural learning Culturally mixed groups on international campuses: An opportunity for inter-cultural learning . Higher...

    1998

  42. [50]

    APACrefauthors \ 1978

    Vygotsky APACrefauthors Vygotsky, L S. APACrefauthors \ 1978 . Mind in Society: Development of Higher Psychological Processes Mind in society: Development of higher psychological processes . Harvard University Press . APACrefURL [ 2025-04-14 ]http://www.jstor.org/stable/j.ctvj...

    1978

  43. [51]

    , Sochacka, N

    walther2013 APACrefauthors Walther, J. , Sochacka, N. \ Kellam, N. APACrefauthors \ 2013 10 . Quality in interpretive engineering education research: Reflections on an example study Quality in interpretive engineering education research: Reflections on an example study . Journ...

    2013 doi

  44. [52]

    , Sochacka, N W

    walther APACrefauthors Walther, J. , Sochacka, N W. , Benson, L C. , Bumbaco, A E. , Kellam, N. , Pawley, A L. \ Phillips, C M L. APACrefauthors \ 2017 . Qualitative Research Quality: A Collaborative Inquiry Across Multiple Methodological Perspectives Qualitative research qual...

    2017 doi

  45. [53]

    , Rao, A C S

    wankhade2022survey APACrefauthors Wankhade, M. , Rao, A C S. \ Kulkarni, C. APACrefauthors \ 2022 . A survey on sentiment analysis methods, applications, and challenges A survey on sentiment analysis methods, applications, and challenges . Artificial Intelligence Review 55 7 5...

    2022

  46. [54]

    APACrefauthors \ 2023

    Watkins APACrefauthors Watkins, R. APACrefauthors \ 2023 . Guidance for researchers and peer-reviewers on the ethical use of Large Language Models (LLMs) in scientific research workflows Guidance for researchers and peer-reviewers on the ethical use of large language models (l...

    2023 doi

  47. [55]

    , Pollard, B

    werth2023investigating APACrefauthors Werth, A. , Pollard, B. , Hobbs, R. \ Lewandowski, H. APACrefauthors \ 2023 . Investigating changes in student views of measurement uncertainty in an introductory physics lab course using clustering algorithms Investigating changes in stud...

    2023

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.