Activity-derived variables
During the processing of the final data, ‘derived variables’ were created. These variables combine data from multiple questions to create activity level measures of participation.
These variables were created using SPSS syntax to calculate the duration, frequency and intensity with which people participate in activities.
These variables were then used to create headline measures of activity, such as twice monthly participation.
Creation of composite sports
Once derived variables had been created at activity level, they were then aggregated up to create measures of participation spanning whole sports or groups of sports.
This was done by summing the measures for each individual activity (or, for binary yes/no measures, by counting a mention of 'yes' at any individual activity). For example, data for the following activities was combined into a single code of ‘Racquet sports’:
- Tennis;
- Badminton;
- Squash & Racquetball;
- Pickleball;
- Padel;
- and Table Tennis.
Overall measures of activity were also created. These measures reflected the Chief Medical Officers' recommendations on activity levels, and aggregated minutes, frequency, and intensity of activity per week across all activities, grouping individuals into those who are active, fairly active and inactive.
Demographic-derived variables
In addition to the activity and composite measures, demographic and geographic variables were created from the raw questionnaire data and from sample data.
These included, for example, variables grouping age, ethnicity, physical impairment, socio-economic classification and educational status.
Other individual-level health measures were also created, including BMI score and portions of fruit and vegetables eaten the day before the participant completed the questionnaire.