I'd like to look at the effect of lockdown policies(announcements) over time on features of the COVID-19 epidemic. I'm not very familiar with different time-series, so i'd like advice on what type of analysis might be appropriate here. Some information about the data:

  • Outcome/Dependent Variable is continuous, but does not change at regular intervals. For example, it may stay constant and then abruptly change on specific dates.
  • Each region has implemented multiple government responses in the form of lockdowns. The announcement of these lockdowns does not occur at regular intervals.
  • Each instance of a government policy in regards to lockdowns can increase restrictions or decrease restrictions, so we would hypothesize an increase/decrease depending on the policy towards lockdowns.

Here's a picture of what the data might look like in a hypothetical region (in reality, i'd be doing this separately for many regions).

enter image description here

Some questions i'd like to answer:

  1. What is the average time from a lockdown policy/announcement to a change in the outcome? Perhaps this question lends itself to some kind of survival analysis?
  2. What is the magnitude of that change, on average?
  • $\begingroup$ Survival analysis is appropriate when you want to investigate the times until events of one or several specific types. With a continuous outcome survival analysis won't be useful. $\endgroup$ – EdM Mar 23 at 18:13
  • $\begingroup$ I figured survival might not be very useful here, but the part where I thought it might work is in searching for the median time to the change in outcome after a lockdown announcement is made. $\endgroup$ – RNB Mar 23 at 18:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.