Machiavellian said:
The key is not the question but the type of questions and how they are asked. If they only asked a limited amount of questions and only seek one answer than you do not have a complete scope for you stats. If nelson would to check what I do on my PS3 it would look like I only watch netflix and only select the kids section because my son and daughter is always watching netflix. That stat would not be good or acurate as to the purpose of the PS3 in my household. Without having a very clear and percise way of know exactly how the questions are asked, who was asked and how the research was conducted, drawing conclusions on such limited data would not make you a good researcher. Most researchers compile data from multiple sources not just one. If you really want to make your data at least appear more accurate you would need to gather way more sources then just one. |
1. You'd have a point if this was only 1 year, but it's been consistent throughout the past generation as well. If the questions and how they're asked were the problem, you'd expect a more randomized result. Unfortunately for your point, they've been consistent.
2. Nielson does not work that way. It asks what you are using the console for.
http://www.vg247.com/2013/03/14/ps3-used-for-streaming-media-more-than-gaming-nielsen-study/
You can easily draw conclusions if you know what you're talking about.