Datasets:

coping-emotions
/

roleplay-corpus

license: creativeml-openrail-m
task_categories:
  - text-classification
language:
  - en
tags:
  - coping
  - emotions
  - microaggressions
pretty_name: roleplaycoping
size_categories:
  - 1K<n<10K

The COPING datasset: COPIng emotioN strateGies, the first corpus of coping information

Overview
Content folder
Details demographics
Details role-play corpus
Citation
License
Contact

1. Overview

This folder contains the COPING text corpus. We collected the corpus through a role-playing paradigm in which workers had to impersonate a character X and respond to a scenario in which interlocutor Y takes an opposite, hostile stance. The different descriptions for character X correspond to four definitions of coping strategies: attack, contact, distance, and reject. In total, the corpus contains 1,200 data points, balanced among coping strategies, topics, and scenarios. More details on the data and methods can be found in the associated paper (see Section 5).

2. Content folder

The main folder contains the following files:

demographics.tsv: file with the screeners used to accept participants for different topics.
roleplay-corpus.tsv: file stores the answers of people who had to impersonate a character X. In case a person claims they would react differently, this file also contains their personal answers.
README.md: this file
LICENSE: Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License

3. Details demographics

This file stores the screeners used to accept participant in different topics. The file contains the following variables:

round_number: (string) Study round in which the data is collected.
prolific_id: (string) Anonymized worker ID.
topic: (string) Controversial topic against which the worker reacts.
residence: (string) Current country of residence.
ethnicity: (string) Ethnicity (i.e., peoples’ ethnicity describes their feeling of belonging and attachment to a distinct group of a larger population that shares their ancestry, colour, language or religion).
alcohol_therapy: (string; "Yes"/"No") Prior participation in individual therapy for alcohol use.
pro-choice: (string) Pro-choice or pro-life stance when it comes to others having the right to terminate their pregnancy.
gender: (string) Gender.
place-before-18: (string) Residence where worker spent most of their time before turning 18.
immigration: (string) Was worker born in the country they are currently living in?
sexual-orientation: (string) Sexual orientation.

The value "--" means that a given screener was not used in a specific round.

4. Details role-play corpus

This file contains the data that was collected during the role-playing experiment. The role-play is made up out of these variables:

round_number: (string) Study round in which the data is collected.
prolific_id: (string) Anonymized worker ID.
coping: (string) Coping strategy for which text was generated.
topic: (string) Topic of the scenario.
scenario: (string) Fictional dialogue history that the worker has to imagine for role-playing.
X_Y_relationship: (string) Power relationship between characters X and Y.
X_reply: (string) The answer of X imagined by the worker.
X_reaction: (string) How the worker would expect X to behave .
X_[EMOTION]: (int) (e.g., X_anger) The degree of that EMOTION that the worker would expect X to feel, rated on a 5-point scale with 1 representing the lowest degree and 5 the highest.
respondent_belief_clash: (string; "Yes"/"No") Whether the controversial utterance of character Y clashes with the beliefs and values of the worker.
participant_experience: (string; "Yes"/"No") Whether the worker has experienced prejudice on their ideas with respect to the given topic.
you-react-same: (string; "Yes"/"No") Whether the workers would react the same way as X.
respondent_reply: (string) What the workers would have replied if the given scenario happened in the real life. Only answered if you-react-same is "No".
respondent_reaction: (string) How the workers would behave if the given scenario happened in the real life. Only answered if you-react-same is "No".
respondent_[EMOTION]: (int) How much the workers would feel an EMOTION, if the given scenario happened in the real life. The answer is rated on a 5-point scale with 1 representing the lowest degree and 5 the highest. Only answered if you-react-same is "No".

A value of "--" means that the variable was not collected for that instance.

5. Citation

Please cite this paper as:

@InProceedings{Troiano2024,
  author    = {Enrica Troiano and Sofie Labat and Marco Antonio Stranisci and Viviana Patti and Rossana Damiano and Roman Klinger},
  title     = {Dealing with Controversy: An Emotion and Coping Strategy Corpus Based on Role Playing},
  booktitle      = {Findings of the Association for Computational Linguistics: EMNLP 2024},
  month          = {nov},
  year           = {2024},
  address        = {Miami, Florida, USA},
  publisher      = {Association for Computational Linguistics}
}

6. License

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. This means that the repository is freely available for academic purposes or individual research, but any other use is explicitly prohibited. Moreover, any derivative work (e.g., re-using or modifying the existing dataset) has to be distributed under the same terms and conditions. If you want to use the data for commercial purposes, please contact the authors (see contact details below).

7. Contact

For any questions regarding the dataset, do not hesitate to contact us at: