The text-package: An R-package for Analyzing and Visualizing Human Language Using Natural Language Processing and Deep Learning

Kjell, Oscar; Giorgi, Salvatore; Ha, Schwartz

doi:10.31234/osf.io/293kt

Cited by 33 publications

(32 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

Section: Natural Language Processing and Statistical Analysesmentioning

confidence: 99%

“…The word data were analyzed with the r-package Text 0.9.0 2 (Kjell et al, 2021). The words generated in the current study were given their semantic representations (i.e., vectors of numeric values describing each word) from a previously created semantic space (used and described in Kjell et al, 2021). The semantic space was created using latent semantic analyses (Landauer and Dumais, 1997) based on singular values decomposition (Golub and Kahan, 1965) on the cooccurrences of 1.7 × 10 9 words from the English Google 5-g database.…”

Section: Natural Language Processing and Statistical Analysesmentioning

confidence: 99%

“…This regression model was applied to the word responses in the current study to estimate their valence (i.e., the regression model was a language predicted valence scale). This model was created using ridge regression (with a penalty grid ranging from 10 −16 to 10 16 ), where crossvalidation was used to evaluate the model (for more details, see Kjell et al, 2021).…”

Section: Language Predicted Scalesmentioning

confidence: 99%

“…Finally, all the individual words in the word responses are "projected" onto the aggregated direction embedding line. The projection is achieved by first "anchoring" all of the individual words' representations in space by subtracting the second group's aggregated semantic representation from each word's representation and then using the dot product to project each word's anchored representation (for more details, see Kjell et al, 2021). To statistically test the words, a dot product null distribution is created by calculating the dot product among randomly selected semantic representations and an aggregated direction embedding created from randomly swapping words' semantic representations from the two groups.…”

Section: Supervised Dimension Projection Plotsmentioning

confidence: 99%

“…All analyses were carried out in R (R Core Team, 2020) using RStudio (RStudio Team, 2020). Apart from the text package (Kjell et al, 2021), the following packages were used: tidyverse (Henry and Wickham, 2020), Hmisc (Harrell et al, 2020), dplyr (Wickham et al, 2020), ppcor (Kim, 2015), psychometric (Fletcher, 2010), reshape2 (Wickham, 2007), ggplot2 (Wickham, 2016, p. 2), data.table (Dowle and Srinivasan, 2019), lm.beta (Behrendt, 2014), lattice (Sarkar, 2008), effsize (Torchiano, 2020), and WRS2 (Mair and Wilcox, 2019).…”

Section: R-packagesmentioning

confidence: 99%

See 4 more Smart Citations

Computational Language Assessments of Harmony in Life — Not Satisfaction With Life or Rating Scales — Correlate With Cooperative Behaviors

2021

View full text Add to dashboard Cite

Different types of well-being are likely to be associated with different kinds of behaviors. The first objective of this study was, from a subjective well-being perspective, to examine whether harmony in life and satisfaction with life are related differently to cooperative behaviors depending on individuals’ social value orientation. The second objective was, from a methodological perspective, to examine whether language-based assessments called computational language assessments (CLA), which enable respondents to answer with words that are analyzed using natural language processing, demonstrate stronger correlations with cooperation than traditional rating scales. Participants reported their harmony in life, satisfaction with life, and social value orientation before taking part in an online cooperative task. The results show that the CLA of overall harmony in life correlated with cooperation (all participants: r = 0.18, p < 0.05, n = 181) and that this was particularly true for prosocial participants (r = 0.35, p < 0.001, n = 96), whereas rating scales were not correlated (p > 0.05). No significant correlations (measured by the CLA or traditional rating scales) were found between satisfaction with life and cooperation. In conclusion, our study reveals an important behavioral difference between different types of subjective well-being. To our knowledge, this is the first study supporting the validity of self-reported CLA over traditional rating scales in relation to actual behaviors.

show abstract

Section: Natural Language Processing and Statistical Analysesmentioning

confidence: 99%

Section: Natural Language Processing and Statistical Analysesmentioning

confidence: 99%

Section: Language Predicted Scalesmentioning

confidence: 99%

Section: Supervised Dimension Projection Plotsmentioning

confidence: 99%

Section: R-packagesmentioning

confidence: 99%

See 3 more Smart Citations

Computational Language Assessments of Harmony in Life — Not Satisfaction With Life or Rating Scales — Correlate With Cooperative Behaviors

2021

View full text Add to dashboard Cite

show abstract

Employees Adhere More to Unethical Instructions from Human Than AI Supervisors: Complementing Experimental Evidence with Machine Learning

2023

View full text Add to dashboard Cite

The role of artificial intelligence (AI) in organizations has fundamentally changed from performing routine tasks to supervising human employees. While prior studies focused on normative perceptions of such AI supervisors, employees’ behavioral reactions towards them remained largely unexplored. We draw from theories on AI aversion and appreciation to tackle the ambiguity within this field and investigate if and why employees might adhere to unethical instructions either from a human or an AI supervisor. In addition, we identify employee characteristics affecting this relationship. To inform this debate, we conducted four experiments (total N = 1701) and used two state-of-the-art machine learning algorithms (causal forest and transformers). We consistently find that employees adhere less to unethical instructions from an AI than a human supervisor. Further, individual characteristics such as the tendency to comply without dissent or age constitute important boundary conditions. In addition, Study 1 identified that the perceived mind of the supervisors serves as an explanatory mechanism. We generate further insights on this mediator via experimental manipulations in two pre-registered studies by manipulating mind between two AI (Study 2) and two human supervisors (Study 3). In (pre-registered) Study 4, we replicate the resistance to unethical instructions from AI supervisors in an incentivized experimental setting. Our research generates insights into the ‘black box’ of human behavior toward AI supervisors, particularly in the moral domain, and showcases how organizational researchers can use machine learning methods as powerful tools to complement experimental research for the generation of more fine-grained insights.

show abstract

Know Thyself! Predicting Subjective Well-Being from personality estimation discrepancy and self-insight

2022

View full text Add to dashboard Cite

Discrepancies in views of the Self are suggested to be negatively related to well-being (Higgins, 1987). In the present study, we used a novel concept, Personality Estimation Discrepancy (PED), to test this classic idea. PED is defined as the computed difference between how one view oneself (Self-Perceived Personality) and a standard Big Five test (IPIP-NEO-30). In a pre-registered (osf.io) UK online study (N = 297; Mage = 37, SD = 14) we analyzed: (1) whether PED would predict Subjective Well-Being (SWB; Harmony in Life, Satisfaction with Life, Positive affect, Negative Affect) and Self-Insight, and (2) whether Self-Insight would mediate the relationship between PED and SWB. The results showed that underestimation of Extraversion, Conscientiousness, and Emotional Stability indeed is associated with both high SWB and high Self-Insight. However, these effects mostly disappeared when controlling for the Big Five test scores. Furthermore, Self-Insight largely (42.9%) mediated the relationship between the mis-estimation and SWB. We interpret these finding such that the relationship of mis-estimating one’s personality with SWB and Self-Insight are mostly explained by the Big Five factors, yet the discrepancy is a dependent feature of scoring particularly high or low on certain personality traits.

show abstract

The text-package: An R-package for Analyzing and Visualizing Human Language Using Natural Language Processing and Deep Learning

Cited by 33 publications

References 39 publications

Computational Language Assessments of Harmony in Life — Not Satisfaction With Life or Rating Scales — Correlate With Cooperative Behaviors

Computational Language Assessments of Harmony in Life — Not Satisfaction With Life or Rating Scales — Correlate With Cooperative Behaviors

Employees Adhere More to Unethical Instructions from Human Than AI Supervisors: Complementing Experimental Evidence with Machine Learning

Know Thyself! Predicting Subjective Well-Being from personality estimation discrepancy and self-insight

Contact Info

Product

Resources

About