RED C use a telephone approach for national political polling, as it has been found worldwide to gain a more representative sample than a face to face approach when conducting political opinion polls.
To ensure our sample is totally representative we use an approach called random digit dialling to create a database of telephone numbers. This method is commonly used in market research, and involves first selecting a random sample from directories of all numbers across Ireland and then adding +1 or + 3 to this number. By doing this we are able to ensure we reach all telephone households, including those that are ex-directory.
RED C are also one of the first research companies in the world to change their sampling approach in order to take account of the growing number of voters who only have a mobile phone. This was incorporated prior to the general election in 2007. Half of our sample is now conducted through calls to mobile numbers, again using an RDD approach. This ensures that our sample reaches 98% of all adults in the population, including those people who only have a mobile, those who only have a landline and those who have both.
Quotas are then set on demographics such as age, sex, social class and region to ensure that the people we speak to are representative of all adults aged 18+, based on the very latest Census statistics. The final data is then also weighted to these demographic parameters as a final check to ensure the sample is entirely representative.
For all national population opinion polls RED C interview a random sample of 1,000+ adults aged 18+ by telephone. This sample size is the recognised sample required by polling organisations for ensuring accuracy on political voting intention surveys. The accuracy level is estimated to be approximately plus or minus 3 per cent on any given result at 95% confidence levels.
For constituency surveys we will conduct among a smaller sample size of 500 adults 18+, either by telephone if transfer analysis is not required or face to face if it is required.
In all respects the surveys we conduct are within the ESOMAR (European Society for Opinion and Marketing Research) guidelines for political and opinion polling.
For constituency surveys where more than basic first preference analysis is required we recommend a face to face approach in order to use a ballot paper approach. The ballot paper approach allows us to count votes as they would be in an actual election and provide detailed transfer analysis.
While this means we cannot use our preferred telephone approach to conduct the survey we do still continue to use our specific analysis techniques to ensure a representative sample. These include:-
- Careful question wording
- Weighting by past voting
- Analysing by those likely to vote
- Adjusting for refusers/Don’t Knows
International guidelines for polling recommend that any published local constituency opinion poll should be conducted on a sample size of at least 500 interviews, and this is the sample we recommend for constituency surveys.
This sample will provide a margin of error of 4.5% at 95% confidence, and is an accepted sample size for polls which are conducted in areas smaller than the total population.
The sample points for the survey are selected to be representative of the constituency at DED level. Interviews are given demographic quotas to achieve a set number of interviews within each selected DED, and therefore be representative of the constituency as a whole.
Quotas are then set based on the known population characteristics within each of the areas of the constituency, in terms of gender, age and social class; and the results are finally be weighted back to known population characteristics to ensure it is entirely representative of the voting population