Gathered from randomised multistage stratified sampling, Syndicated Surveys is a nationally representative public opinion poll
An omnibus survey which can be used to poll the economic and political pulse of Sri Lanka with an error margin under 3%
The answers are tabulated by demographic to give detailed insight into the polled opinion
Syndicated surveys by Verité Research is a product where the cost of implementing a national level survey is distributed among a few clients (each with one or two questions), thereby creating a cost-efficient tool for each client to obtain the national sentiment on key issues.
The survey results, if the client chooses to make them public, are backed by Verité Research, a Think Tank in Sri Lanka with established capability and recognition.
The sample for this survey is drawn from a national database built and updated periodically since 2019. It includes households selected through previous survey iterations. The inclusion of households in this database, to participate in such surveys, is based on explicit consent received during a physical meeting with enumerators, in which the household visited has been chosen through a randomised process of Stratified Multistage Sampling.
The Stratified Multistage Sampling method: This allowed to ensure both, (i) a randomized sample selection process to avoid selection bias, and (ii) a detailed geographic stratification of the sample across all 25 districts of the country. The Primary Sampling Unit (PSU) was defined as the Grama Niladhari (GN) division. There are 14,022 GN divisions, and they are the smallest administrative unit in Sri Lanka. Data from the Department of Census and Statistics in Sri Lanka was used as source information on the population distribution and geographic stratification. Those invited to respond to the questions are adults of age 18 and above.
The sample size was determined such that statistical results at the national level could be obtained at a 95% confidence level within a maximum error margin of 3%.
This means that in 95 out of 100 survey samples taken to estimate a mean value in the national population, the mean value of the survey sample would vary above or below the true mean value in the national population within an interval of at most 3 percentage points.
Sri Lanka has a population of c. 22.2 million persons and c. 5.7 million households. Therefore, a nationwide randomised sample of around one thousand people is required to achieve the above level of accuracy when calculating the population mean at a national level. The sample size of the survey was 1,018 individuals, each from a different household.