The reading list
The research that informs Project Verbatim - articles we have found interesting
Pavlyshenko & Stasiuk · Electronics and Information Technologies · 2025
Evaluates transformer-based sentence-embedding models on semantic-similarity tasks, weighing accuracy against processing speed to guide model selection (such as MPNet) for English text analysis.
Fergadiotis, Wright & Green · J. Speech, Language & Hearing Research · 2015
A psychometric comparison of lexical-diversity indices in adult speech that quantifies how sample length biases each measure and supports length-robust indices for valid diversity estimates across speakers.
Pennebaker, Chung, Frazee, Lavergne & Beaver · PLOS ONE · 2014
Shows that a function-word analytic-thinking index, with more articles and prepositions and fewer pronouns and verbs, measured in college admissions essays predicts students' later academic grades.
McCarthy & Jarvis · Behavior Research Methods · 2010
A validation study of lexical-diversity measures showing that MTLD, vocd-D and HD-D resist the text-length bias that distorts the simple type-token ratio, and recommending they be used together for reliable scoring.
McNamara, Louwerse, McCarthy & Graesser · Discourse Processes · 2010
Introduces Coh-Metrix, a tool computing hundreds of cohesion and readability indices such as referential overlap, connectives and semantic similarity, and validates them against text comprehension and difficulty.
Lu · Int. J. Corpus Linguistics (L2SCA) · 2010
Presents the L2 Syntactic Complexity Analyzer, which automatically computes fourteen measures of syntactic complexity such as clause length, subordination and coordination, for second-language writing research.
Heylighen & Dewaele · Foundations of Science · 2002
Proposes the F-score, a part-of-speech formula gauging how formal versus context-dependent a text is, and demonstrates that it cleanly separates spoken from written language across many registers.
Kemper, Greiner, Marquis, Prenovost & Mitzner · Psychology and Aging · 2001
A follow-up Nun Study analysis charting how grammatical complexity and idea density change across the lifespan, and confirming that early-life linguistic ability forecasts later cognitive decline, supporting density as a marker of cognitive reserve.
Snowdon, Kemper, Mortimer, Greiner, Wekstein & Markesbery · JAMA · 1996
Tracked Catholic nuns across six decades and found that those whose early-life autobiographies packed more ideas per ten words (higher idea density) were far less likely to show poor cognition or Alzheimer pathology in old age.
Hunt · NCTE Research Report No. 3 · 1965
Foundational work introducing the T-unit, the shortest grammatically complete unit, as a measure of syntactic maturity, showing that clause length and degree of embedding both rise steadily with grade level.
Kanze, Huang, Conley & Higgins · Academy of Management Journal · 2018
At pitch competitions investors asked men promotion questions and women prevention questions; founders mirrored that framing, and promotion-focused answers attracted far more capital, a mechanism behind the gender funding gap.
Rouhizadeh, Jaidka, Smith, Schwartz, Buffone & Ungar · EMNLP · 2018
Classifies locus of control from annotated Facebook posts against questionnaire scores; control is easy to detect, but labeling it internal versus external is harder, with lexical features outperforming syntactic ones.
McAdams & McLean · Current Directions in Psychological Science · 2013
A concise overview of narrative identity, the internalized and evolving story people tell about themselves, and how the way they narrate experiences relates to wellbeing, maturity and development over time.
Rauch & Frese · European J. Work & Organizational Psychology · 2007
A meta-analysis reviving trait research, showing business owners' personality, especially traits matched to entrepreneurial tasks like need for achievement and innovativeness, reliably relates to business success.
Baum & Locke · J. Applied Psychology · 2004
A longitudinal study showing entrepreneurs' traits such as passion and tenacity, plus skills, predict later venture growth, operating largely through specific goals and self-efficacy rather than directly.
Mueller & Thomas · J. Business Venturing · 2001
A nine-country study linking an internal locus of control and innovativeness to entrepreneurial potential, and showing that the prevalence of these traits varies systematically with national culture.
McAdams · Review of General Psychology · 2001
Sets out narrative identity theory: people build evolving life stories whose themes, notably agency (mastery, achievement) and communion, reflect personality and predict psychological adjustment.
Higgins · Advances in Experimental Social Psychology · 1998
Extends regulatory focus theory, detailing how promotion (eager, gain-seeking) and prevention (vigilant, loss-avoiding) orientations shape emotion, strategy choice and decision-making.
Higgins · American Psychologist · 1997
The foundational statement of regulatory focus theory: a promotion focus pursues gains and ideals while a prevention focus avoids losses and seeks security, two motivational systems with distinct strategies and risk profiles.
Bateman & Crant · J. Organizational Behavior · 1993
Introduces and validates the Proactive Personality Scale, capturing the stable disposition to spot opportunities and take initiative to change one's environment rather than passively adapt to it.
Peterson, Schulman, Castellon & Seligman · Cambridge UP (book chapter) · 1992
Describes CAVE, a method to extract and rate the causal attributions people give in any text, enabling explanatory-style (optimism versus pessimism) coding without administering questionnaires.
Rotter · Psychological Monographs · 1966
The seminal monograph introducing locus of control, whether people attribute outcomes to their own actions (internal) or to luck, fate and others (external), together with a scale to measure it.
Gow, Larcker & Zakolyukina · J. Accounting Research · 2021
Builds a measure flagging managers' non-answers to analyst questions on earnings calls; about eleven percent of questions go unanswered, and higher non-answer rates predict negative market reactions and weaker future performance.
Boyd, Blackburn & Pennebaker · Science Advances · 2020
Analyzes thousands of texts to reveal a recurring three-part narrative structure of staging, plot progression and cognitive tension, and shows it generalizes across fiction and everyday accounts alike.
Antonakis, d'Adda, Weber & Zehnder · Management Science · 2015
Field and lab experiments show charismatic rhetoric is more than talk: a charismatic speech measurably raised workers' effort and output, demonstrating the real economic value of charisma.
Parhankangas & Ehrlich · J. Business Venturing · 2014
Studies impression management in entrepreneurs' pitches to business angels, finding that calibrated, two-sided framing, including openly acknowledging weaknesses, can build more credibility than relentless positivity.
Kacewicz, Pennebaker, Davis, Jeon & Graesser · J. Language & Social Psychology · 2014
Across several datasets shows higher-status people use fewer first-person-singular pronouns like I and me and more we and you, so pronoun use reliably signals relative standing within a hierarchy.
Larcker & Zakolyukina · J. Accounting Research · 2012
Identifies linguistic markers, such as extreme positive emotion, few hesitations and vague impersonal references, that distinguish deceptive from truthful executive statements on earnings calls better than chance.
Antonakis, Fenley & Liechti · Academy of Management Learning & Education · 2011
Two randomized experiments show that training people to deploy charismatic leadership tactics, such as metaphors, stories, three-part lists and contrasts, raises how charismatic and influential observers rate them.
Loughran & McDonald · Journal of Finance · 2011
Builds finance-specific sentiment word lists, showing general dictionaries wrongly score terms like tax or liability as negative in 10-K filings; the tuned lists better link disclosure tone to returns and risk.
Ireland, Slatcher, Eastwick, Scissors, Finkel & Pennebaker · Psychological Science · 2011
Using speed-dating and instant-messaging data, shows that linguistic style matching between two people predicts mutual attraction, relationship initiation and later stability.
Chen, Yao & Kotha · Academy of Management Journal · 2009
Across a laboratory experiment and a field study of venture-capital decisions, perceived preparedness, not displayed passion, drove funding, reframing the pitch as a persuasion process centered on competence.
Cardon, Wincent, Singh & Drnovsek · Academy of Management Review · 2009
A theory paper defining entrepreneurial passion as intense positive feeling tied to identities such as inventor, founder and developer, and explaining how it energizes goal pursuit, creativity and persistence.
Cortese & Fugett · Behavior Research Methods, Instruments & Computers · 2004
Provides imageability norms, ratings of how readily a word evokes a mental image, for 3,000 monosyllabic English words, a reference resource widely reused in psycholinguistic, reading and memory research.
Niederhoffer & Pennebaker · J. Language & Social Psychology · 2002
Documents linguistic style matching, the way conversation partners unconsciously converge in their function-word use, and links the degree of matching to engagement and social coordination.
Green & Brock · J. Personality & Social Psychology · 2000
Introduces narrative transportation: the more a reader is absorbed into a story, the more their beliefs and attitudes shift toward it, establishing absorption as a core engine of narrative persuasion.
Scott, Keitel, Becirspahic, Yao & Sereno · Behavior Research Methods · 2019
Offers the Glasgow Norms: ratings of 5,500 words on nine psycholinguistic dimensions including imageability, concreteness, valence and arousal, with analysis of how the dimensions interrelate.
Fast, Chen & Bernstein · CHI · 2016
Introduces Empath, a tool that generates and validates hundreds of lexical categories from deep-learning word embeddings, closely approximating hand-built dictionaries like LIWC at far lower cost.
Brysbaert, Warriner & Kuperman · Behavior Research Methods · 2014
Provides crowdsourced concreteness ratings for roughly 40,000 English words and lemmas, a large and widely used norm set for research on meaning, memory and language processing.
Biber · Cambridge UP (book) · 1988
A landmark multidimensional analysis of register: tracking co-occurring linguistic features across many texts, it identifies dimensions like involved-versus-informational and shows speech and writing differ systematically.
Oo, Jiang, Sahaym, Parhankangas & Chan · J. Business Venturing · 2023
Analyzing 28,000 crowdfunding campaigns with machine learning, shows founders who use a wider variety of speech acts and switch between them more often raise more money; individual speech acts show inverted-U effects.
Gavrilenko et al. · arXiv:2312.06236 · 2023
Expands public-data startup prediction to 171 mostly textual features, including linguistic metrics drawn from tweets, across ten models, forecasting funding with F-scores above 0.73 and beating prior approaches.
Clarke, Cornelissen & Healey · Academy of Management Journal · 2019
A field and experimental study of pitches finding that hand gestures which depict and symbolize ideas, more than the choice of literal versus figurative language, raise investors' willingness to invest.
ECIE 2017 proceedings (authors on repository record) · ECIE Proceedings · 2017
Compares interviews of 51 highly funded Silicon Valley founders with average entrepreneurs and finds the successful group's language reflects action, future, customer, collective and growth orientations.
Not located (per your notes) · —
Per your notes, a multi-model NLP and clustering framework that analyzes founder language for associations with business success; I could not locate a stable source online to verify its specifics.
Tetlock & Gardner · Crown (book) · 2015
A popular account of Tetlock's forecasting tournaments: the best forecasters are not the smartest but those who think probabilistically, update on evidence and stay actively open-minded, directly relevant to epistemic calibration.
Maarouf et al. · arXiv:2409.03668 · 2024
Builds a fused large language model that combines founders' venture-platform self-descriptions with structured features to predict startup success, finding the text adds meaningful signal beyond fundamentals.