Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Help a newbie with coding event/ performance from baseline

    Hello,

    I am diligently learning Stata and I have come across a problem that I'm hoping someone can help with. I have a data set such as the one below (but much larger) and actually six years long (2017-2022).
    SID year score
    1 2019 123
    1 2020 145
    1 2021 133
    2 2018 320
    2 2019 333
    2 2020 318
    etc...
    This is test score data before and after the pandemic so 2021 is treated as an event. I've done mixed effect regression for my main analysis with time splines or knots. I have those set up (precov, covid, postcov) and have had success with the following command:
    mixed score precov covid postcov || sid:
    I've gone into other demographics (grade for example) from there, but to keep this simple let us just ignore other variables.

    Now what I'm trying to answer is how prior performance affected covid performance. So for example did student 1 (above) fare differently than student 2 because they were lower in proficiency, to begin with?

    I think I need to figure out how to get Stata to output the covariance matrix so that I can examine how the intercept and slopes are correlated over the 3 periods of time. If the correlation is positive, then higher starting points are associated with faster growth. If the correlation is negative, then higher starting points are associated with slower growth.

    Any ideas on how to do this? As always I appreciate this forum!
    Last edited by Brynne Baruch; 21 Jan 2023, 07:42.
Working...
X