Movies -a form of entertainment that has been around for centuries. Although research has been performed on movie ratings over time (e.g. Moon et al., 2010; Ramos et al., 2015), the effect of streaming services on ratings remains rather unexplored. This is especially notable since the world-wide adoption of streaming services in 2010 caused shifts in the distribution and quality perception of movies by audiences (Hennig-Hurau et al., 2021; Kumar, 2023). The present study aims to provide academic relevance by filling this gap.
Movie-watchers nowadays have a broader reference frame and thus become more critical in evaluating content (Hadida et al., 2020). In addition, streaming has resulted in a higher output of movies, which results in a saturated market to choose from. Resultingly, a negative relationship is hypothesised between release year and the average rating.
However, this effect is hypothesised to differ between animation and non-animation movies. As such, technological innovations over time have enhanced the storytelling opportunities and visual possibilities of this genre (Sun, 2023). Current research has not specifically compared animation versus non-animation, whereby the present study aims to shed further light on on the moderating role of genre on the relationship between release year and the average rating.
Besides the academic relevance, the insights are also socially relevant for movie producers and distributors to better understand audiences. All in all, the focus of the research can be encapsulated with the following research question:
To what extent does the release year of a movie influence its average IMDb rating, and does this relationship differ between animated and non-animated films?
In order to test H1 and H2 and therefore answer the main research question, secondary data is consulted from IMDb.com. The database holds several datasets, of which two provide relevance in the context of the present study:
title.basics.tsv.gz: information about the title, runtime, genres and release year.
title.rating.tsv.gz: information on ratings derived from IMDb users.
The dataset merged from the secondary data contains the following variables:
| Variable | Type | Definition |
|---|---|---|
tconst |
character | Unique IMDb identifier for each title |
primaryTitle |
character | The international title of a movie |
startYear |
integer | Release year of a movie |
runtimeMinutes |
integer | Runtime expressed in minutes |
genres |
character | Genres separated by commas |
averageRating |
numeric | Average IMDb user rating (0.0–10.0) |
numVotes |
integer | Number of votes received by the title |
before_2010_dummy |
dummy | Whether a movie is released before or after 2010 |
animation_dummy |
dummy | Whether a movie is an animation or not |
Data analysis consisted of three consecutive parts:
- Descriptive statistics were performed on four subsets of the data;
- Animation movies
- Non-Animation movies
- Movies released before 2010
- Movies released since 2010
-
Independent t-tests were performed to uncover significant differences in the average ratings between the above mentioned subsets.
-
Linear regression models were conducted that predicted the dependent variable of average IMDb rating. Three models were compiled;
- Model 1 including the two dummy predictors
- Model 2 including the two dummy predictors, control variables
- Model 3 including the two dummy predictors, their interaction, and control variables.
Elaborate details are deployed in two documents:
-
A data exploration PDF-document is generated that provides descriptive statistics and visualisations of the prepped data set that was used for analysis.
-
The end product is deployed in the form of a PDF-document, which further elaborates on the motivation, literature, methodology, findings and conclusions.
The graph below can be used to describe the general gist of the findings. There are significant differences in rating, with animation movies scoring higher than non-animation movies. Additionally, movies released before 2010 are proven to score significantly higher than movies released since 2010. Nevertheless, results show no significant results for the interaction between release year and animation status in predicting the average IMDb rating.
The findings yielded provide social relevance in that the success of films are independent of when they are released. Despite that, the results show the animation genre is appealing to audiences, as expressed with the higher ratings.
The structure of this repository is as follows:
project-name/
├─ paper/
│ ├─ makefile
│ └─ paper.Rmd
├─ src/
│ ├─ analysis/
│ │ ├─ 10_pre_assumption_check.R
│ │ ├─ 11_regression_analysis.R
│ │ ├─ 12_post_assumption_check.R
│ │ ├─ 9_t-test.R
│ │ ├─ README.md
│ │ └─ makefile
│ ├─ data-exploration/
│ │ ├─ 7_descriptive_statistics.R
│ │ ├─ 8_exploratory_visualisations.R
│ │ ├─ Data_exploration.Rmd
│ │ ├─ README.md
│ │ └─ makefile
│ ├─ data-preparation/
│ │ ├─ 1_download_raw_data.R
│ │ ├─ 2_merge_raw_data.R
│ │ ├─ 3_select_variables.R
│ │ ├─ 4_handling_missing_values.R
│ │ ├─ 5_filtering.R
│ │ ├─ 6_feature_engineering.R
│ │ ├─ README.md
│ │ └─ makefile
│ └─ .gitkeep
├─ .gitignore
├─ README.md
├─ makefile
By using the installation guides as found on Tilburg University's Science Hub, ensure your device has the following software installed:
- Make - installation guide.
- R / R Studio - installation guide.
Within R Studio, ensure the following packages are installed.
# Install the following packages
install.packages("googledrive")
install.packages("tidyverse")
install.packages("mice")
install.packages("car")
install.packages("lmtest")
install.packages("broom")
install.packages("here")
install.packages("lm.beta")Replicating the workflow of this study can be achieved by taking the following steps:
-
Fork the repository using GitHub
-
Open your preferred terminal/command line (e.g., Git GUI)
-
Create a copy of this repository to your local machine by copying the following sentence in your preferred terminal/command line:
git clone https://github.com/course-dprep/animation-imdb-ratings-analysis-over-time.git
- Set your working directory to the just forked repository
animation-imdb-ratings-analysis-over-timeand run the following word in your preferred terminal/command line:
make
- Afterwards, you can clean the working directory if desired by removing the gen and data folder by running the following in your preferred terminal/command line:
make clean
This project is set up as part of the course Data Preparation & Workflow Management of the Master of Science Marketing Analytics program at Tilburg University, based in The Netherlands. The project is implemented by:
- Britt van Haaster
- Isah Huijbregts
- Lars van der Kroft
- Amanda van Lankveld
- Amy Quist
- Stefan Valentijn
Hadida, A. L., Lampel, J., Walls, W. D., & Joshi, A. (2020). Hollywood Studio Filmmaking in the Age of Netflix: a Tale of Two Institutional Logics. Journal of Cultural Economics, 45(2), 213–238. https://doi.org/10.1007/s10824-020-09379-z
Hennig-Thurau, T., Ravid, S. A., & Sorenson, O. (2021). The Economics of Filmed Entertainment in the Digital Era. Journal of Cultural Economics, 45(2), 157–170. https://doi.org/10.1007/s10824-021-09407-6
Kumar, L. (2023, April). A Study On The Impact Of The OTT Platform On The Cinema With The Special Reference On The Cinema Audience. ResearchGate; unknown. https://www.researchgate.net/publication/376650380_A_Study_On_The_Impact_Of_The_OTT_Platform_On_The_Cinema_With_The_Special_Reference_On_The_Cinema_Audience
Moon, S., Bergey, P. K., & Iacobucci, D. (2009). Dynamic Effects among Movie Ratings, Movie Revenues, and Viewer Satisfaction. Journal Of Marketing, 74(1), 108–121. https://doi.org/10.1509/jmkg.74.1.108
Ramos, M., Calvão, A. M., & Anteneodo, C. (2015). Statistical patterns in movie rating behavior. PLOS ONE, 10(8), e0136083. https://doi.org/10.1371/journal.pone.0136083
Sun, Z. (2023). What does cgi digital technology bring to the sustainable development of animated films?. Sustainability, 15(14), 10895.https://doi.org/10.3390/su151410895