Editor’s note: We received the following note from a reader in response to an item in the Survey Monitor section of our October 2013 issue (“#twitterontv – How tweeting can increase on-air TV ratings”), which was compiled from a Nielsen Company press release. The note has been edited for clarity and to conceal the identity of the author, who wishes to remain anonymous.
I am a huge fan of Quirk’s but I was a bit disappointed to read your item in the October issue about Twitter and TV ratings. Specifically, the idea that tweets can “cause” significant changes in TV viewing.
Your item says that the study was “independent” but that isn’t true. Nielsen and Twitter have a business partnership around this that is profitable for both of them and has led to new Twitter ratings sold by Nielsen.
Next, the idea that any statistical test can show “causation” for anything is probably unrealistic. To my knowledge, statistics have not even been able to “prove” a link between smoking and cancer, or the idea that the planet is currently getting warmer due to increases in burned fossil fuels.
In addition, specific aspects of Nielsen’s methodology and results in this study are open for debate. If they found that 29 percent of TV episodes are influenced by Twitter, does that support or refute their hypothesis? After all, that means that 71 percent of episodes were not influenced by Twitter. How low would that percent have to be for them to admit that causality was not there? (And many in the TV industry would assert the opposite relationship, namely that high TV ratings lead to a higher number of tweets!)
But the main problem might be in their methodology. I believe that they included commercial minutes in their analysis. Well, during commercials, it is common for TV ratings to drop (as some people change channels) and for tweets to rise (easy for people to tweet during commercials). Then, when the program comes back from commercial, TV ratings tend to rise, possibly producing a spurious positive correlation between tweets during commercial and later ratings.
As a member of the TV industry, it’s important to me to see that studies about TV are properly reported in the media. And I have to say that I was particularly disappointed by this study and press release from Nielsen. But that’s just my opinion.
In any event, thanks for all you do. I really am a big fan of Quirk’s and what you do for the research industry!