Tuesday, December 12, 2006

Reporting Raw Data

Okay, I was yelled at for a previous post. I was condescending and, further, unclear as to exactly why the article made me mad. The data is fine, there may be more to it than that, but the article is bad for this very simple reason: raw data has very little meaning. This is true of any raw data. Statistics, in particular are rather devoid of meaning, as even their generation can be influenced by those who pose the questions compared with those who interpret the questions. An example from the article/study would be related to the point about "imagining life without instant messaging." One person may see that as "would my life be noticeably different if IM's didn't exist?" Another may read that as "do I think IMing will ever die out?" and another as "do I think people could get along in a meaningful way if there was no instant messaging?" Those are three very different questions. Exactly how the question was posed and how the respondents interpreted it is very relevant to the meaning of the results. I deliberately assumed mocking interpretations of the possible questions from the data given to demonstrate the meaninglessness of the report as a whole. People will read the CNN article and believe that there is value of some sort contained within. That is a problem.

There may be value in the data, but it is open to the ideas of the reader, not the reality of the report. In any case of reporting in which the reader/listener must make their own conclusions with nothing but raw data and no explanations then not only is the journalism bad, it is in fact non-existent. No relevant information beyond raw data was given. Further, there is no extrapolation within the report regarding what significance the data may contain. The additional reporting discusses the addictiveness of messaging. This is tangentially relevant, at best, but even then is not related to the data at all. This is not journalism. This does not belong on CNN. It may be amusing, but in the absence of analysis, reporting data is not a story.

No comments: