How do you handle missing values in a dataset when those values represent critical KPIs?

Question

Anonymous · Accepted Answer

Missing values in a dataset could happen due to one of the following reasons:

* Issue in measurement definition - aka instrumentation. Common scenarios
   include:
   
   * metric was not valid for that particular category. e.g:- total conversions
     may not be defined for a user that never reached the target page)
   
   * ratio metric causing div by 0
     
     * Solution: Either treat the value as missing or impute with a zero
       depending on context.

* Issue in capturing data. Scenarios include:
   
   * collection machinery failed at source. for e.g.:- pixel to capture click
     was not transmitted before page load)
   
   * manual error during a survey
     
     * Solution: Impute using mean/median substitution or with more
       sophisticated imputation methods such as kNN imputation

How do you handle missing values in a dataset when those values represent critical KPIs?

Practice this question with AI

Go Premium

Community Answers

Unlock Full Access

Practice More Questions