My hypothesis is that there is a large discrepancy between the standard of deviation of music reviews and that of film reviews. Meaning that any individual record is more likely to polarize critical responses to it than a film. I'm collecting the raw data from metacritic.com, and as a way of randomizing I'm limiting the reviews pulled from date ranges.
- How many reviews do I need to collect to give me a reasonable error rate?
- How big would the difference in standard deviation need to be to make it statistically significant?
- Is my selection method actually random?
- Are there are any better statistics I could look at other than standard of deviation that would make my case better?
Thanks for reading, hope to hear from you,