About GMI Poll

Home > Press > GMI Poll > About GMI Poll

Advanced market research tools

Survey data for GMI Poll are collected by using Net-MR®, GMI's web-based suite of market research tools that includes data collection, analysis, and reporting.

Using this application service provider (ASP) software, we carefully construct a sample of qualified panelists from 20 of the world's leading economic nations, ensuring that specific demographic requirements are met. The qualified panelists are then emailed survey invitations that include a link to the on-line poll for their completion.

Statistically accurate results start coming in within as little as 24 hours and are usually completed within one to three days. This compares to traditional survey methodologies such as telephone interviews, on-the-street polling, and mailed forms, which can take between three to six weeks or more, and may or may not be representative samples.

The Process

Upon clicking the survey link within the email invitation, respondents are led through a series of web pages, which are hosted on secure servers at our data centers, to select answers to the survey questions. Respondents are able to click through all the questions in approximately 15 minutes. Once an interview is completed, each respondent is prevented from completing the interview again, thereby protecting the representative sample of the survey.

Respondents' answers are uploaded to Net-MR's servers in real time for review and on-the-fly analysis. Access to the data is secured by Secure Socket Layer (SSL) encryption, user accounts and passwords.

At the end of the data collection stage, the data is reviewed by our statistical team to analyze and verify that the survey results are comparably representative before release.

Representative sample in 20 countries

GMI Poll surveys are representative of a demographically balanced population of informed individuals. GMI considers these respondents "informed" because they have access to the Internet and maintain email accounts. To the extent people with email accounts are different from people without email accounts, the samples are demographically representative, but include a selection bias for online, or informed, individuals. This potential selection bias is mitigated by high response rates, with participation percentages typically over 30%, as compared to the participation rates for telephone-survey respondents that fall below 10%.

Selection biases can be present in any sample and any survey. For example, most political polls that attempt to predict the outcome of elections interview only "likely voters" or "registered voters." Most market research studies interview only highly selected groups who make purchasing decisions or who intend to buy a specific product. In contrast, respondents for the GMI Poll are selected randomly, using a method that produces a sample balanced by age, gender, geographic region, and where applicable, race and ethnicity.

Country-to-country international comparisons using these data are valid because the same sampling methodology is used in all 20 countries, and therefore, any selection bias affecting these data is similar across all countries. The results for GMI Poll surveys are an internationally comparable set of results based on responses from 1,000 informed respondents per country.