PROVIDE Project Technical Paper 2005:1
originally contained ten or more ‘missing values’ are included in the table. The table shows
that the vast majority of these ‘missing values’ are in fact uncoded (column A). True missing
values make up a very small number of the total number of observations (columns B and F).
A number of observations were also miscoded, i.e. households that were expected to report
zero expenditure (or a missing value if Statistics South Africa had been consistent in their
treatment of optional sections) (column D). The sum of the ‘missing values’, miscoded
observations and the remaining non-zero values (column E) gives the total number of
observations in the database.
February 2005
39
© PROVIDE Project
More intriguing information
1. The name is absent2. FASTER TRAINING IN NONLINEAR ICA USING MISEP
3. Consumer Networks and Firm Reputation: A First Experimental Investigation
4. Partner Selection Criteria in Strategic Alliances When to Ally with Weak Partners
5. Dementia Care Mapping and Patient-Centred Care in Australian residential homes: An economic evaluation of the CARE Study, CHERE Working Paper 2008/4
6. Policy Formulation, Implementation and Feedback in EU Merger Control
7. The name is absent
8. The name is absent
9. The name is absent
10. The name is absent