Predicting leadership perception with large-scale natural language data

Bhatia, Sudeep; Olivola, Christopher Y.; Bhatia, Nazlı; Ameen, Amnah

doi:10.1016/j.leaqua.2021.101535

Cited by 29 publications

(24 citation statements)

References 108 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…By initializing our model’s response vectors with pretrained word embeddings, we were able to train effective models using far less data than would have been necessary if we had trained these parameters from scratch. We have previously proposed related techniques for modeling judgments of words on various dimensions (Bhatia, 2019; Bhatia et al, in press; Gandhi et al, 2022; Richie et al, 2019; Zou & Bhatia, 2021; also see Hollis et al, 2017; Van Rensbergen et al, 2016; Sedoc et al, 2017). Other work, including ours, has also shown the applicability of such methods for the study of concept knowledge (Bhatia & Richie, in press; Derby et al, 2019; Lu et al, 2019; Richie & Bhatia, 2021).…”

Section: Discussionmentioning

confidence: 99%

“…Semantic judgments about words (e.g., the tastiness of a food) can also be approximated by calculating the relative vector similarity of a judgment target (e.g., apple ) to words high (e.g., delicious , tasty ) and low (e.g., disgusting ) on a judgment dimension (Grand et al, 2022; Richie et al, 2019). However, even better (out-of-sample) approximations of semantic judgments can be made by directly regressing human ratings for a semantic dimension onto the vectors for judgment targets (Bhatia, 2019; Bhatia et al, in press; Gandhi et al, 2022; Hollis et al, 2017; Richie et al, 2019; Utsumi, 2020; Zou & Bhatia, 2021; see Snefjella & Blank, 2020, for a comprehensive list of “semantic norm extrapolation” studies, as well as caveats thereof). The advantage of this approach is that it allows for human data to directly supervise the setting of flexible weights on the attributes of the target representation.…”

Section: Theoretical Backgroundmentioning

confidence: 99%

“…The advantage of this approach is that it has the best of both worlds: the scale that comes from optimizing a generic objective on a large data set as well as the specificity and task appropriateness that comes from optimizing on a particular objective on a potentially smaller data set. In prior work, we have used this approach to predict human judgments and decisions (Bhatia, 2019; Bhatia et al, in press; Gandhi et al, 2022; Richie & Bhatia, 2021; Richie et al, 2019; Zou & Bhatia, 2021). We believe this approach is also suitable for the present application to free association, for which we have labeled data for only a few hundred thousand trials.…”

Section: Modeling Frameworkmentioning

confidence: 99%

See 2 more Smart Citations

Free association in a neural network.

Richie¹,

Aka²,

Bhatia³

2023

Psychological Review

Self Cite

View full text Add to dashboard Cite

Free association among words is a fundamental and ubiquitous memory task. Although distributed semantics (DS) models can predict the association between pairs of words, and semantic network (SN) models can describe transition probabilities in free association data, there have been few attempts to apply established cognitive process models of memory search to free association data. Thus, researchers are currently unable to explain the dynamics of free association using memory mechanisms known to be at play in other retrieval tasks, such as free recall from lists. We address this issue using a popular neural network model of free recall, the context maintenance and retrieval (CMR) model, which we fit using stochastic gradient descent on a large data set of free association norms. Special cases of CMR mimic existing DS and SN models of free association, and we find that CMR outperforms these models on out-of-sample free association data. We also show that training CMR on free association data generates improved predictions for free recall from lists, demonstrating the value of free association for the study of many different types of memory phenomena. Overall, our analysis provides a new account of the dynamics of free association, predicts free association with increased accuracy, integrates theories of free association with established models of memory, and shows how large data sets and neural network training methods can be used to model complex cognitive processes that operate over thousands of representations.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Theoretical Backgroundmentioning

confidence: 99%

Section: Modeling Frameworkmentioning

confidence: 99%

See 1 more Smart Citation

Free association in a neural network.

Richie¹,

Aka²,

Bhatia³

2023

Psychological Review

Self Cite

View full text Add to dashboard Cite

show abstract

“…This approach has also been used to model evaluations of naturalistic entities on a continuous scale, such as evaluations of the risk posed by technological hazards, the effectiveness of prominent politicians, the excitement generated by consumer brands, and the healthiness of food items (Bhatia, 2019; Bhatia et al, 2021; Gandhi et al, in press; Richie et al, 2019). Results have shown that linear models, common in judgment research, can be combined with word vector representations of these natural entities to predict participants’ evaluations on both an aggregate and an individual level.…”

Section: Judgment and Decision Makingmentioning

confidence: 99%

Cognitive Modeling With Representations From Large-Scale Digital Data

Bhatia

Aka

2022

Curr Dir Psychol Sci

Self Cite

View full text Add to dashboard Cite

Deep-learning methods can extract high-dimensional feature vectors for objects, concepts, images, and texts from large-scale digital data sets. These vectors are proxies for the mental representations that people use in everyday cognition and behavior. For this reason, they can serve as inputs into computational models of cognition, giving these models the ability to process and respond to naturalistic prompts. Over the past few years, researchers have applied this approach to topics such as similarity judgment, memory search, categorization, decision making, and conceptual knowledge. In this article, we summarize these applications, identify underlying trends, and outline directions for future research on the computational modeling of naturalistic cognition and behavior.

show abstract

“…Social scientists increasingly use text-based measures as dependent or independent variables (1)(2)(3)(4)(5)(6). Texts are complex, high-dimensional objects; thus, researchers must find simpler, lower-dimensional representations for their texts to use them in scientific analyses.…”

Section: Introductionmentioning

confidence: 99%

How to make causal inferences using texts

et al. 2022

View full text Add to dashboard Cite

Text as data techniques offer a great promise: the ability to inductively discover measures that are useful for testing social science theories with large collections of text. Nearly all text-based causal inferences depend on a latent representation of the text, but we show that estimating this latent representation from the data creates underacknowledged risks: we may introduce an identification problem or overfit. To address these risks, we introduce a split-sample workflow for making rigorous causal inferences with discovered measures as treatments or outcomes. We then apply it to estimate causal effects from an experiment on immigration attitudes and a study on bureaucratic responsiveness.

show abstract

Predicting leadership perception with large-scale natural language data

Cited by 29 publications

References 108 publications

Free association in a neural network.

Free association in a neural network.

Cognitive Modeling With Representations From Large-Scale Digital Data

How to make causal inferences using texts

Contact Info

Product

Resources

About