See All Boards Subscribe To Thread
Statistics Public Forums
Probability: Smallpox in Boston: observational study or experiment?
alexisvdb Oct 15 ReplyFlag
Hi All,

I have a question about Guided Practice 2.45 (p 93). This is about the smallpox in Boston in 1721, where we have data on survival or death of people who were either inoculated or not.

Guided Practice 2.45 asks: is this study observational or is this an experiment? There is a clear explanatory variable (inoculation) and response variable (survival or death), and one is clearly affecting the other. I suspect that the inoculation was performed by a doctor (or similar), even though people were not randomly assigned to treatment or control groups. We are not merely observing living habits and survival rates here, I feel. As p. 24 says: "Studies where the researchers assign treatments to cases are called experiments." So, I would call this an experiment (with a number of major flaws, which we should expect from a study done 200 years ago).

However, the answer in the footnote says: "observational". So, I am a bit confused.

I wonder if I could get any comments on this?

Thanks,

Alex
David Oct 16 ReplyFlag
I wish I could recall the source for this data, which we ran across 8-9 years ago, and I don't recall the exact detail. There are two possibilities:

1) It was an observational study, in that the doctor may have treated patients who agreed to the treatment, or patients asked to be treated.

2) It was an experiment, e.g. a doctor (or set of doctors) selected some patients for inoculation while not inoculating the other set.

Based on the past note that this was an observational study, my suspicion is that this is indeed an observational study. However, I'm not extremely confident. If it is an experiment, I would be very surprised if it's a randomized experiment.

Best,
David

Your User ID will appear with your posts.