MST POLL METHODOLOGY & STANDARDS

A standard MST Poll involves interviews with a statistically representative sample of at least 2,500 registered voters, which yields an error margin of +/-2% at the national level. All provinces (except for Batanes) and highly urbanized cities are included in the generated sample areas.  The MST Poll adheres to the best practices and high standards set by the World Association for Public Opinion Research (WAPOR). The polls are initially conducted on a monthly basis up to March 2010, and then weekly tracking polls will be implemented starting April 2010 up to a few days before the actual elections.

=========================================

For the November 2009 Standard Poll, it was conducted among a sample of 2,500 registered voters nationwide, interviewed in-person from November 2-9, 2009. Multi-stage probability sampling was applied allocated into 500 random samples each in the National Capital Region, North & Central Luzon, South Luzon & Bicol, Visayas and Mindanao. The error due to sampling for results based on the nationwide sample could be plus or minus two percentage points. The error for subgroups is higher. Regional estimates were weighted by projected number of voters to obtain the estimates for Total Philippines.

=========================================

For the December 2009 Standard Poll, it was conducted among a sample of 2,500 registered voters nationwide, interviewed in-person from December 6-12, 2009. Multi-stage probability sampling was applied allocated into random samples of 300 in the National Capital Region, 555 in North & Central Luzon, 540 in South Luzon & Bicol, 505 in the Visayas and 600 in Mindanao. This breakdown of sample sizes by major areas reflects the latest proportion of registered voters (October 2009 data of the COMELEC).  The error due to sampling for results based on the national sample is at plus or minus two percentage points. The error for subgroups is higher. Regional estimates were weighted by projected number of voters to ensure proper estimates for Total Philippines.

=========================================

For the January 2010 Standard Poll, it was conducted among a statistically-representative sample of 2,500 registered voters nationwide, interviewed in-person from January 20-27, 2010. Multi-stage probability sampling was applied allocated into random samples of 300 in the National Capital Region, 550 in North & Central Luzon, 540 in South Luzon & Bicol, 510 in the Visayas and 600 in Mindanao. This breakdown of sample sizes by major areas reflects the latest proportion of registered voters (October 2009 data of the COMELEC).  The error due to sampling for results based on the national sample is at plus or minus two percentage points. The error for subgroups is higher.

The registered voters were asked their intention to vote and only those who signified they definitely will vote in the May elections were asked to accomplish a ballot and shade the oval space which corresponds to their choices; after which their ballot is dropped in a box.  This interview technique simulates what will happen during the computerized elections in May.

=========================================

For the February, March and April 2010 Standard Polls, these were conducted among a statistically-representative sample of 2,500 registered voters nationwide, interviewed in-person [survey dates: February 20-26; March 21-23; April 4-6].  Multi-stage probability sampling was applied allocated into random samples of 300 in the National Capital Region, 550 in North & Central Luzon, 540 in South Luzon & Bicol, 510 in the Visayas and 600 in Mindanao. This breakdown of sample sizes by major areas reflects the latest proportion of registered voters (February 2010 data of the COMELEC).  The error due to sampling for results based on the national sample is at plus or minus two percentage points. The error for subgroups is higher.

The registered voters were asked their intention to vote and only those who signified they definitely will vote in the May elections were asked to accomplish a ballot and shade the oval space which corresponds to their choices; after which their ballot is dropped in a box.  This interview technique simulates what will happen during the computerized elections in May.

Advertisements
%d bloggers like this: