Skip to main content
CT Zhu's user avatar
CT Zhu's user avatar
CT Zhu's user avatar
CT Zhu
  • Member for 11 years, 8 months
  • Last seen more than 2 years ago
awarded
revised
Loading…
Loading…
awarded
awarded
Loading…
awarded
revised
Loading…
awarded
Loading…
revised
Loading…
revised
Loading…
Loading…
answered
Loading…
awarded
awarded
comment
Number of hours instead of frequency count in a chi-squared test
I agree with @gung 's comment a lot. It is a case of survival analysis. If you have an subject, that you followed for 100hrs without seeing the 'event', it can be treated as a right-censored. If you have an obs. for which the 'event' time is observed at 99 hours, then it is a exact time-event. If you have multiple events observed for one subject, it becomes a repeated event or Recurrent event survival analysis. You can have a subjects tracked for only 10 hours (without seeing an event), then it is right censored at 10hrs. You can have subjects that were tracked for different amount of time.
comment
How many people need to be screened if the disease has an R0 of 2, and starts with 1 infection in 1000?
Sure, this question is regarding the logistic growth model, used mainly in ecology, microbiology and medical research. The model is better illustrated in a form of differential equation: $\frac{df(t)}{dt}=r f(t)(1-\frac{f(t)}{K})$, $r$ is often called intrinsic growth rate and $K$ called carrying capacity, $f(t)$ is the population size/number of animal/bacterial/new product sold at time $t$. $f'(0)$ is the grown rate at zero time also labeled as $R_0$, or I hope that's what the OP means. It couldn't be $r$ for sure.
awarded
awarded