+ Reply to Thread
Results 1 to 2 of 2

Thread: Using median case load as a predictor

  1. #1
    Fortran must die
    Points: 58,790, Level: 100
    Level completed: 0%, Points required for next Level: 0
    noetsi's Avatar
    Thanked 915 Times in 874 Posts

    Using median case load as a predictor

    I am testing the impact of a contractor or a public unit working with a customer [ I want to know if a private or public unit had more success] . A big part of this is generating control variables so I can separate out the impact of these from private and public units.

    I assume that median caseload probably impacts the results. I can, at great difficulty, find manually what the case load of each private unit and each public unit was when the member was in the system. So if a member was at unit 5 [which is public] for march 2015 I could list what that median caseload was for that month and year. And if he was in the system in April I can find the caseload for them in that month. And I could use that, along with private/public and other control variables, to predict success.

    But I am not sure if that even makes sense [calculating the impact of caseload this way]. The case load will vary across the time a customer is in the system. Is it legitimate to generate an average caseload as a predictor [one problem with that is that the resulting value would not be a dummy, but would have a single numeric value, the average of the median caseload]. Another problem is that a customer might be there six months, and the predictor would be calculated for the entire time of the study.

    I think I am doing this wrong, but I am not sure how to do it right... Something else is that most of the control variables apply to the customer, I am seeking to eliminate these and certain program differences to get at the impact of public and private unit. Here the difference would be specifically associated with the unit [I know median caseload per unit].
    Last edited by noetsi; 03-07-2016 at 06:02 PM.
    "Very few theories have been abandoned because they were found to be invalid on the basis of empirical evidence...." Spanos, 1995

  2. #2
    Omega Contributor
    Points: 38,374, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Not Ames, IA
    Thanked 1,186 Times in 1,147 Posts

    Re: Using median case load as a predictor

    I didn't really read your post but quantile regression uses medians and are good heterogeneous residuals.
    Stop cowardice, ban guns!

+ Reply to Thread


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts

Advertise on Talk Stats