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

  • Testing for outliers with multiple imputation in panel data

    As the title says, I have a outlier in my panel data (2 time periods) with MI. I was wondering what statistical test I could use to see if I should or shouldn't include this single observation.

    As far as I understand it, tests like cook'st distance or leverage-versus-squared-residual plot do not work with panel and MI data. Which test is then usable to help me get an impression if I want this data point or not?

  • Marcos Almeida
    Since you have already posted a similar query here, and you haven't get a reply so far, hopefully this article may be helpful to you.

    Leave a comment: