Fact Checking and Opinions

I’ve previously said a few good things about fact checking operations such as Politifact, but I’ve also noticed a few questionable items where opinion interfered with facts.

But today I read the article Lies, Damned Lies, and ‘Fact Checking’, and having looked into this a bit further, I must say that the few issues I have found personally with the service are not really representative of the facts. There is, in fact, quite a serious problem here.

The errors that Hemingway points out are of the same type as I have noticed, but his stories make it clear that the problem is much more pervasive.

It is, however, a very easy problem to spot and to correct for. The issue here is being able to distinguish between fact, theory, and opinion. As I’m using these terms here, that means between the data on which one’s opinions are based, theories that connect these facts together, and opinions, which are built on the previous two.

Some might question my distinction of theory and opinion, but I would maintain it is a valid one. Theories provide a consistent way in which we tie various facts together. A theory can be checked against another one by how well each theory explains the facts at hand.

Fact checking is important, and I think it is what an organization like Politifact should do. I do believe that in many cases they do check facts. But they also treat their own opinions as better facts than those of the people they “check.” There is a certain amount of journalistic arrogance involved in that. Who checks the journalists?

Read the complete article. I found myself in agreement with this article in general. The one issue I have is one I have not myself done enough research on. The question is just how valid a favorable and unfavorable story count is in determining bias. I think it has to be corrected based on who is providing the most opportunities.

For example, I would expect a greater number of fact checks done on Republicans right now due to the presidential debate. They are simply making more claims. This doesn’t mean that the balance is not biased. I haven’t made a count that would let me say something like that. It just makes me take the claim of bias with a grain of salt.