Data Issues

Waves 1 to 7
Data issues - Waves 1 to 7 – February 2019

15 Minor changes for weight, BMI & and height percentiles and z-scores

As part of the investigation into the date of birth corrections (outlined in section 14), some inconsistency was discovered in the method for calculating 'SC age in months' across the waves:

  • Wave 1 and Wave 2 were calculated using the integer method
  • Wave 3 and Wave 4 were calculated using a custom method that correctly derived the number of completed months to determine age in months
  • Wave 5 and Wave 6 were calculated using the rounding method.

The calculation method for Waves 1, 2, 5 and 6 has now been changed to match the custom method used for Waves 3 and 4. This method was chosen for all waves as it correctly derives the number of 'true' completed months (using the date of interview as a reference point for age calculation) as opposed to the integer method that simply divides the number of days since the interview by 30.44.

As result of this method change, 'age in months for SC' was recalculated for Waves 1, 2, 5 and 6. A large number of records were altered by +1 or -1 for age in months in these waves. However, only very minor changes were observed, when before and after distributions were compared.