Digit preference and its implications on population projections in zambia: evidence from the census data

Introduction: Tests of the age preference are not only important because of the major role it plays in describing the population but also because of its importance in making population projections and planning for resource distribution by government agencies and other stakeholders. A common error that most developing countries face in age data is tendency by respondents to round-off or misstate their ages.

Objective: The aim of the study was to assess whether age preference exists in Zambia and its implications on population projections and resource allocation.

Methods: This study utilised age data in single years to determine digit preference as reported in the 1969, 1990 and 2010 censuses by using the Whipples and Myers Blended Indices used. In addition, for grouped data the United Nations Age-Sex Accuracy Index was used.

Results: Although the quality of census data has been improving with successive census undertakings, the fact is age heaping in ages ending with 0 and 5 still exist in Zambia, with more males than females, having heaped ages at the two terminal digits. However, besides remarkable improvement from a score 72.9 in 1969 to a score of 27.9 in 2010 as indicated by the United Nation Age-Sex Accuracy Index, age data still remains inaccurate.

Conclusion: Therefore, this study shows that, age data from censuses is misreported and inaccurate in Zambia, which if not adjusted results in wrong projections for age dependent variables. Therefore, whenever age data is being collected in future censuses date of birth should also be collected.