Flu Vaccine: Half a Statistic Is Worse Than None 1


NBC Nightly News had a story Growing concern over children dying of the flu.


There is one statement in the report that is a perfect example of how the media mislead you with half a statistic.

The report never told you what this number means. What did they expect you to learn from this? I can think of three possible conclusions you could take depending on what is the value of the statistic they did not report. What they failed to report was what percentage of the children who survived were never vaccinated.

In the figures below I have chosen three possible values for the missing statistics of the percentage of children who were not vaccinated that survived. Above each graph, I have put a label of what you might be able to conclude given any one of the green bars compared to the red bar.

In the above figure, of the children who survived they had a lower percentage of not being vaccinated so you might conclude there was an advantage to being vaccinated.

In the above figure, of the children who survived they had the same percentage of not being vaccinated so you might conclude there was neither an advantage nor a disadvantage to being vaccinated.

In the above figure, of the children who survived they had a higher percentage of not being vaccinated so you might conclude there was a disadvantage to being vaccinated.

Without seeing a green bar, there is nothing you can conclude from seeing the red bar alone. You might have concluded that certainly the vaccine had whatever advantage or disadvantage you had assumed before seeing the number. In other words, this half statistic may have made you more sure of the wrong thing.


Leave a comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.

One thought on “Flu Vaccine: Half a Statistic Is Worse Than None

  • SteveG Post author

    I do not know the value of the green bar statistic, so I don’t claim to know anymore about the efficacy of the vaccine than I knew before this report. If you have a reliable source for the missing data, then let’s have it. If you have no reliable source, then there is no use arguing over the numbers I made up for its possible values. By making up these imaginary numbers, I was not trying to say that anyone of them was close to the truth. I made the numbers up as examples of what they could be. I don’t know how more plainly I can state it.