4 ms·

There's the question of timing too. It's not clear if the number includes all CVST incidences or only those ocurring in a time window after vaccination. (Update

There's the question of timing too. It's not clear if the number includes all CVST incidences or only those ocurring in a time window after vaccination. (Update: according to NYT these cases are 1-3 weeks after vaccination, so the background rate is too low to matter here). If there is a strong correlation in timing you can not simply use frequency data to rule out causation.

jjk166 6y ago
Of course the cases are shortly after vaccination, the data is symptoms reported shortly after vaccination which might be side effects. You're not going to have anyone who got a clot before the vaccine or long afterwards in the dataset, but those cases still did/will exist. There is no evidence yet of a correlation in timing.
- fspeech 6y ago
  The background rate you quoted are expected to be spread out independent of vaccination. If you pick a two-week window after vaccination you should only expect 1/6 of the cases that are expected in 3 month, which is the number you calculated.
  - jjk166 6y ago
    But we're not observing people for 3 months and picking an arbitrary two week window. People do not get added to the population until they have taken the vaccine, and they are all observed for different lengths of time. We did not hypothesize a 2 week period beforehand and compare how many landed within to how many were outside that range, the 2 weeks is just a circle drawn around the datapoints after the fact.
    - fspeech 6y ago
      Here is the chain of logic: 1. First assume independence, that CVST is unrelated to vaccination. 2. Take any two weeks, you can calculate the expected cases using population background rate because we assume independence. 3. The number is around 1 using Wikipedia data. 4. The observed case number 6 greatly exceeds the expected number based on independence assumption. 5. We conclude that with high probability that our independence assumption is wrong, i.e. there's correlation with high confidence.
      - jjk166 6y ago
        Your logic breaks down at step 2. As a counter example, what are the odds that in the other 7 weeks you would have zero cases? Of course the answer is that this isn't how the data works. You are not sampling a random, normally distributed event over a fixed interval, you are sampling the spacing between two different events where one of them is systemically linked to how you define the population and the interval.

jjk166 6y ago

Of course the cases are shortly after vaccination, the data is symptoms reported shortly after vaccination which might be side effects. You're not going to have anyone who got a clot before the vaccine or long afterwards in the dataset, but those cases still did/will exist. There is no evidence yet of a correlation in timing.

fspeech 6y ago
The background rate you quoted are expected to be spread out independent of vaccination. If you pick a two-week window after vaccination you should only expect 1/6 of the cases that are expected in 3 month, which is the number you calculated.
- jjk166 6y ago
  But we're not observing people for 3 months and picking an arbitrary two week window. People do not get added to the population until they have taken the vaccine, and they are all observed for different lengths of time. We did not hypothesize a 2 week period beforehand and compare how many landed within to how many were outside that range, the 2 weeks is just a circle drawn around the datapoints after the fact.
  - fspeech 6y ago
    Here is the chain of logic: 1. First assume independence, that CVST is unrelated to vaccination. 2. Take any two weeks, you can calculate the expected cases using population background rate because we assume independence. 3. The number is around 1 using Wikipedia data. 4. The observed case number 6 greatly exceeds the expected number based on independence assumption. 5. We conclude that with high probability that our independence assumption is wrong, i.e. there's correlation with high confidence.
    - jjk166 6y ago
      Your logic breaks down at step 2. As a counter example, what are the odds that in the other 7 weeks you would have zero cases? Of course the answer is that this isn't how the data works. You are not sampling a random, normally distributed event over a fixed interval, you are sampling the spacing between two different events where one of them is systemically linked to how you define the population and the interval.