Info201GroupProj

Check out our website on: https://yan530.shinyapps.io/Info201GroupProj/

Project Description

Data Set : Bechdel Tests of movies from FiveThirtyEight

Background of Data Set

This data was collected by FiveThirtyEight. There is not much information available from them about the dataset. We accessed it using FiveThirtyEight's github folder full of public .csv (_c_omma _s_eparated _v_alue) files.

Update We ended up merging our data set with the Kaggle set IMDB 5000 Movie Database because the OMDB Api we originally intended to use was made private.

Note: The data only goes up to 2013.

What is a Bechdel Test?

According to bechdeltest.com, the Bechdel Test, sometimes called the Mo Movie Measure or Bechdel Rule is a simple test which names the following three criteria: (1) it has to have at least two women in it, who (2) who talk to each other, about (3) something besides a man. The test was popularized by Alison Bechdel's comic Dykes to Watch Out For, in a 1985 strip called The Rule.

The Bechdel Test has since become relatively well known and is often referenced in movie reviews. It has become a rather embarassing test to fail, though many movies do fail it.

Audiences for this project

Our main audience are people who are critical of movies that fail the Bechdel Test and are interested in learning what trends there are in movies being better or worse at the test. More specifically, we will be targeting users who use statistic-based news websites like FiveThirtyEight. We hope by showing these trends that people will be more critical of movies that are likely to fail the Bechdel Test based on the trends they align with.

Questions that will be investigated

This project will answer the question of whether or not the budget of the movie increases the chance of passing or failing the test.
It will also answer the question of how the profit the movie makes can or cannot predict the chance of passing or failing the test.
Lastly, it will answer the question of how the year the movie was made effects the chance of passing or failing the test.

Technical Description

Format of Project: Shiny.io Format of Data Read in: .csv file (2 joined) Major Libraries Used: Plotly, dplyr, ggplot2, knitr, Hmisc, shiny, shinythemes, scales

Data Wrangling Needed

This data is already pretty well formatted. We may need to edit ascii for some titles that are displaying unicode. They also show an IMDB ID, which might lead to editing using the IMDB API. We may need to join tables with IMDB IDs that contain genre so we can then analyze the movies by genre.

Original Questions to be Answered with Statistical Analysis:

See above "questions that will be investigated."

1. Does the budget of a movie impact its likelihood of passing or failing the Bechdel Test? What is the median budget? Average budget?
2. Does the profit the movie made have a relation to how likely a movie was to pass or fail? What is the median profit?
3. Is there a trend in Bechdel Test passing or failing based on the year the movie was made?
4. Does the genre of the movie increase the likelihood of passing or failing the Bechdel Test?
5. How is the IMDB rating related to the movie's pass/fail status?
6. Are there any actors who are more likely to be in passing or failing movies?
7. Does rating (PG-13, R, etc.) affect the likelihood of passing or failing the test?
8. Does release date affect the likelihood of passing or failing the test?

Final Questions Actually Answered

1. Does the budget & profit impact the test result?
2. Does the country impact the test result?
3. Is there a trend based on the year?
4. Does the genre impact the test result?
5. Does popularity relate to the test result?
6. Does rating (PG-13, R, etc.) affect the test result?

Anticipated Major Challenges

The scope of our data is not massive, as it has about 1800 lines of data and some years do not have many movies. We may have to narrow down our scope to a few years where there is a lot of data. It may be challenging to cleanly join data from the IMDB Api, especially if we have many additional columns to add.

Since we are pretty new to using Shiny.io but the INFO 201 class this project is for is going to teach it in more depth, we should be able to work in Shiny.io by the end of this week.

Encountered Challenges

1. The OMDB Api became private the day after we proposed this project. We had to find alternate ways to get more dimensions of data to analyze.
2. The country codes in the data set were not consistent with those the Plotly library used, so those had to be modified.
3. The genres were in an odd format, ex "Comedy|Drama|Musical" which was very tricky to break apart and count.
4. It is worth noting that some of our original questions had to be changed because we could not access the OMDB Api as expected and worked with some different data.

Future Questions to Investigate

1. Does the location of the movie's filming or release affect passage rates?