Data Science in Psychology

Screen Shot 2021-05-26 at 11.02.40 AM.png

Audience

Course module was deployed in Spring 2021 as web app accompanied by an assignment.

Students from different levels take PSYCH 1, and math or computer science is NOT a prerequisite for this course. Unfortunately, since students in PSYCH1 come from different background, we only offered a summarized high-level analysis and were not able to get into the mathematical details of the tools we used for it.

Other applicable courses that may benefit from this module: Some language and linguistic course as well as social psychology and sociology course. 

Project Summary

Primary objective:

To show students how real world lexical data can be collected, interpreted and translated into something with which students who have no previous background in data science can experiment with.

Goals

  1. Collect Data

  2. Interpret data visualizations

  3. Develop insights


Content Outline

Module can be found here: https://psych1-frontend-irrg.onrender.com/

Web App

Individual View

  • Students upload their own Zoom transcript

    • a conversation between them and another person

  • Answer questions about the nature of the conversation and relationship (family, small talk, etc.) to be used as filters later

  • Word Cloud

  • % Occurrence of Words in Your File Bar Graph

  • Emotions

    • Sentiment Distribution Radial Chart

    • % Occurrence Empath Categories in your Conversation

  • 5 Dimensional Curiosity Questions- Your Score vs. Global Average

  • Avg/Max Words Per Turn by Speaker

  • Positive vs. Negative Sentiment in Your Conversation Over Time

Aggregate View

  • Select filters entered by individual users

  • Word Cloud

  • % Occurrence of Words in Class Bar Graph

  • Class Emotions

    • Sentiment Distribution Radial Chart

    • % Occurrence Empath Categories in Entire Class

  • 5 Dimensional Curiosity Questions- Class Score vs. Global Average

  • Frequency of the Average Number of Words Spoken per Turn with Outliers Removed Histogram

  • Frequency of the Maximum Number of Words Spoken per Turn with Outliers Removed Histogram

Assignment

  • Short answer questions focused on reflection and analysis

Overall Work Flow

  • Students upload a transcript from their own conversation

  • Website presents analysis and visualizations such as Word Cloud, sentiment analysis, and entity recognition. 

  • Filter the data by different categories, such as Greek life and personality. 

  • Compare their own data to the aggregated data from the rest of the class. 

  • Professors will be able to choose what data is selected and what questions can be asked

Deployment Comments from Professor

  • deployment process was seamless

  • “[students are] able to pull a lot of information from these metrics”

  • “it was nice to see people use the term surprising so often in these assignments because…just reading through your transcript you might not necessarily learn as much”

    • learning they talked significantly less than their partner etc.

  • “students seemed to have fun…digging into the analysis”

  • “How often do you get to compare your conversation with 200 conversations you were never in a room with?”

  • “[students] are learning things that they didn’t anticipate [about themselves] and thats the best”

For more information email: difuse-pi-group@dartmouth.edu

Previous
Previous

Climate Extremes in a Warming Planet

Next
Next

Differential Equations