The Cambodia Socio-Economic Survey (CSES) asks questions to a country wide sample of households and household members about housing conditions, education, economic activities, household production and income, household level and structure of consumption, health, victimization, etc. There are also questions related to people in the labour force, e.g. labour force participation.
CSES was conducted intermittently in the period 1993 to 2004 but since 2007 the survey is annual. The 2004, 2009 and 2014 were large sample surveys (roughly 12,000 households), whereas the years between have small samples (about 3,600 households).
The data from the CSES provides important information about living conditions in Cambodia and have a wide range of use. Results from CSES are used for monitoring the National Strategic Development Plan (NSDP) and progress towards the Millennium Development Goals. Furthermore, the data are used for developing poverty lines and calculating poverty rates. Data have also been used for food security analyses. The CSES micro data at NIS is available for research and analysis by external researchers after approval by Senior Minister of Planning. The interesting research questions that could be put to the data are many; NIS welcomes new research based on CSES data.
Note to the user
All statistical surveys contain errors and the results – the estimates – from surveys are in most cases not equal to the target values (the true values). If there was a perfectly designed and executed survey, conducted over the whole population not just a sample, the estimate would be equal to the true value. But perfect design and execution can never be reached and the whole population cannot be covered (except in a census) so there will always be inaccuracy in the survey estimates. There are many types of errors in a survey, e.g. measurement errors, coverage errors, non-response, data processing errors and in sample surveys there are also sampling errors. When designing and conducting a survey it is important to control the total error so that accurate estimates can be produced. NIS has put a large effort in the work of minimizing the errors but recommends the user to be aware of the uncertainty in the estimates. The standard error of an estimate is a measure of the uncertainty. Standard errors and confidence intervals for selected estimates are presented in the technical section as well as in an appendix in the CSES 2014 report.
As the published results are estimated values, all percentages and numbers are rounded off. Numbers are rounded to nearest hundreds/thousands and percent to nearest one decimal. It is worth noting that computed percentages are always based on original data. A '0' (zero) means that there is a value close to (and rounded off to) zero and therefore too small to be published. Therefore some tables with percentage do not sum up to 100 percent. In the tables the symbol (-) is used and means no responses in the cell. The symbol (..) means too few observations to be published, whereas (N/A) means that data was not collected that year.