Stats NZ has a new website.

For new releases go to

www.stats.govt.nz

As we transition to our new site, you'll still find some Stats NZ information here on this archive site.

  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+
Classification and coding process

Classification criteria

For this classification, a person’s age is calculated as the time (in whole years) between their birth date and the date when data is collected. 

Classification

Age classification

Age is a flat classification with single-year categories from zero years to 119 years inclusive, plus a category for 120 years and over. 

The full classification is available in Download of classification.

Age is a fundamental variable that is often used in cross-tabulated outputs. In line with international practice, it is Statistics New Zealand policy to impute a person’s age if there are missing values in the data they have given (eg the person gave the year of their birth, but no month or date). This increases the usefulness of data. It means there are no residual categories in this classification. 

Date of birth classifications

Classifications for date of birth are required because age is derived from the day, month, and year a person is born. Date of birth is derived from three flat classifications: date of birth day, date of birth month, and date of birth year. Day has single categories 01 to 31 (representing the days of the month). Month has single categories 01 to 12 (representing the months of the year). Year has single categories (representing 120 years back from the current year). For example, in 2007, the year classification range starts with 1887, which is 120 years from 2007. All three classifications also have residual categories.

The residual categories are defined in Glossary and references.

The full classification is available on the main page.

Coding process

To derive age from date of birth, see the process below.

To derive age

  1. Where day, month and year of birth are all specified, the respondent’s true age is calculated as a whole year with respect to the reference day.
  2. Where only month and year of birth are specified, and the month specified is not the reference month, age can again be calculated as a whole year with respect to the reference day.
  3. Where only month and year of birth are specified, and the month equals the reference month, compute X=(reference day)/(number of days in reference month). Generate a random number between 0 and 1. If the random number is less than X, day of birth is assumed to fall before reference day. If the random number is greater than or equal to X, the day of birth is assumed to fall on or after reference day.
  4. Where only year of birth is specified, a process similar to step 3 is followed, with X being computed as (reference day, specified as number of days since 1 January in reference year)/ (total number of days in reference year).
  5. Where only day and year of birth are specified, a random number between 1 and 12 inclusive is generated for the month. Age can then be calculated as a whole year, with respect to the reference date.
  6. Where only day and month of birth are specified or none of the three fields are given ( the missing field may be populated with a residual code eg 99 rather than a blank) or the values specified are later than the reference date, then age is imputed.

To impute age

When the information has not been provided by the respondent or is not available, a value for age may be imputed if necessary. The methodology for imputation varies from survey to survey. For example, social surveys often use other known information to impute age when it is missing.

  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+
Top
  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+