Users of census statistics must take the responsibility to determine the errors inherent in the data. Further, users must be sure to understand all the terms used in the presentation of the data. For example, does the term ‘income’ refer to family income or individual income? What does the word ‘unemployed’ mean? Users should take care to review all definitions and source-of-error information carefully and to be aware of the data limitations in any subsequent analysis.
As an example of the above, Statistics South Africa reports that the proportion of persons who completed matric or higher has increased from 23.4% in 1996 to 40.5% in 2011. This should be seen against the backdrop of the requirements to pass matric over time. For example, the current requirements for a matriculation pass consist of a minimum of 40% for a pupil’s home language, life orientation and 2 elective subjects, plus a minimum of 30% for mathematics, an additional language and one elective subject.
The significance of data error to the users of census data depends on the nature of the error, the intended use of the data, and the level of detail involved. Some errors occur more or less at random and tend to cancel out when individual responses are aggregated for a sufficiently large group. For example, some people may overestimate their incomes, while others underestimate them. On the other hand, if there is a tendency for people to err in a particular direction (for example, if incomes are generally understated), then the average reported will be different from the true average. The bias they create in the data persists no matter how large the group, and it is difficult to measure.
Numerous errors may arise from a census such as coverage errors (for example, an enumerator missing a dwelling entirely), content or response errors (where a respondent misinterprets a question), misstatement of age and processing errors (such as in coding of the data). Statistics South Africa conducts a Post Enumeration Survey in an attempt to minimise coverage error.
South Africans should take their duty to provide accurate census data seriously as this has an impact on all aspects of their lives such as the provision of health care, schooling and employment programs. Politicians would do well to examine aspects such as reaching their electorate - for example, Census 2011 reveals that 64.8% of households in the country have no access to the internet.
Gregory Whittaker is an independent actuary. www.algorithm-ca.com.
(Photo: President Jacob Zuma receives the 2011 Census results from Statistician General Pali Lehohla on 30 October 2012. Picture: GCIS)