Communities - Sample Balancing

A balance is basically a combination of sought percentages of the values in member profile fields of the total available response count. For example, you may want 40% male, 15% people who fall under 25-34 years of age group in your total response count.

Click to download video

It's the process of ensuring that the sample data used for analysis is representative of the entire population.

This is particularly important when dealing with imbalanced datasets, where certain groups or categories are overrepresented or underrepresented. Balancing the sample can help in reducing bias and improving the accuracy and reliability of the analysis.

The procedure for adjusting the sample to match the population response rate is called Sample Balancing.

You can add a Sample Balancer by performing the following steps-

Go to: Analytics >> Balancing >> Add Sample Balancer

• Click on Add Sample Balancer button.
• Provide a name to the balancer.
• Type in the total response size that is required.
• Select the balancing parameters from the list. A maximum of 3 can be specified at a time.
• Type in the percentage combinations that you are looking for. Ensure that the numbers under each paramter total to 100.
• Once, all the steps are done, click on Add Sample Balancer button.
Only the single-select type member profile fields can be used as balancing parameters - hence if you do not have any in the member profile then the balancing parameters list will be empty.

This is done automatically. After following the above steps, the following screen will display the computation of feasibility.

How to Read this -

• In the example below, I've considered number of responses = 100
• Gender has 3 variations- Male (40%), Female (40%) and Others (20%)
• Age group has 7 variations - Under 18 (10%), 18-24 (15%), 25-34 (15%), 35-44 (15%), 45-54 (10%), 55-64 (15%), Above 64 (20%)
• Now, if we look at row 1, it has numbers mentioned in the dark i.e. 4,4,2 which is 10%of 100 = 10 and then further 40% of 10 = 4 (that's how those numbers are calculated automatically)
• The numbers displayed in green are the values of the profile fields that were entered by the audiences.