What to do about attrition

• Welcome to the Outcome Assessment forum by John Grohol, 9/18/96
  • (...)
    
    • Minimizing Attrition by Ellen Dornelas, 10/3/96
      
      

Perhaps we can discuss what to do about attrition, besides minimizing it. For example, determining the most important classes of groups in our overall populations, finding out how much these categories are under or overrepresented in the sample, and then compensating by weighting the subgroups accordingly (e.g., if women/men are 50/50 in the population and 60/40 in the follow-up sample, weight the sample men X 1.5. Of course this does not tell us anything about differences between those men and women who we didn't get in touch with, vis-a-vis those we did. I wonder if we can get some more expert advice on this problem.

• Addressing sampling bias concerns by Mike Placona Ph. D., 10/10/96