1 Reply Latest reply on Oct 14, 2013 2:32 PM by Alexandra Dunn

    Weighting Survey responses to match General Population

    Amy Smith

      I have read a lot of threads in the Forums about how to weight data in Tableau, but haven't found a solution that I'm sure is the

      right approach I need to get my survey data in line with how the data looks in the representative population.  So I'm looking for help.

       

      The survey data I have collected for Age shows a different distribution than what my client knows is the age distribution

      for their total customer base.  Therefore, I've got people taking the survey at a disproportionate rate to how they really

      exist in the customer database.  Below is a chart of my Age break out in the survey data compared to the population.

       

      AgeSurvey DataGeneral Population
      18-25< 1%1%
      26-351%6%
      36-454%16%
      46-5516%24%
      56-6521%25%
      66+47%29%

       

       

      I am looking for help trying to get my survey data to reflect the distribution in the rep. pop, so for example

      the opinions in my survey data are not over represented by the Age 66+ group, or under represented by the

      Age 36-45 group.  I need to weight all of my survey data (gender, success rating, satisfaction rating, etc.)

      by the weighting scheme that is applied to getting the age data back in alignment to what the true distribution

      should be in the general population.

       

      Any ideas about what approaches I should take?  Is there a calculated field or formula that would make the most sense

      to use?  I need help establishing the weighting criteria, and then knowing how to apply it to all of my charts that

      use the other dimensions from my survey data.

       

      Thank you so much for any tips you can offer!

      Amy -