Hypothesis Testing of Movie Ratings Data

Authors:

Jennifer Rodriguez-Trujillo

Research Question:

In this project, I demonstrate the essential skills involved in hypothesis testing. To do so, I have used a dataset that stems from a replication attempt of published research (Wallisch & Whritner,2017). To run the tests,I set the per-test significance level 𝛼 to 0.005 (as per Benjamin et al., 2018). Throughout the project KS-test, T-test, Mann-Whitney, and a few other significance tests were performed depending on the question at hand.

Data:

This dataset features ratings data of 400 movies from 1097 research participants. A brief description of the data is as follows:

1st row: Headers (Movie titles/questions) – note that the indexing in this list is from 1
Row 2-1098: Responses from individual participants
Columns 1-400: These columns contain the ratings for the 400 movies (0 to 4, and missing)
Columns 401-421: These columns contain self-assessments on sensation seeking behaviors (1-5)
Columns 422-464: These columns contain responses to personality questions (1-5)
Columns 465-474: These columns contain self-reported movie experience ratings (1-5)
Column 475: Gender identity (1 = female, 2 = male, 3 = self-described)
Column 476: Only child (1 = yes, 0 = no, -1 = no response)
Column 477: Movies are best enjoyed alone (1 = yes, 0 = no, -1 = no response)