The way in which a survey is conducted affects the estimates it produces. The way in which the sample is obtained is arguably the most important factor, particularly for media measurement surveys.
The NRS is based on a random sample of adults aged 15+ in Great Britain.
Only specifically selected individuals can be interviewed to ensure that the sample is as representative as possible – there are no quotas and no substitutions.
People with identical demographic characteristics will have different reading habits according to their lifestyle and availability. NRS uses a random sample to reduce the bias towards people who are more likely to be available for interview than others.
It is difficult to correct for the ‘availability’ bias by weighting. It matters more for certain sorts of publication than others. Hence the importance of ensuring the most representative sample possible.
In order to ensure that the sample is random there are strict rules governing each stage of the decision of whom to interview:
- The areas – based on Output Areas, as defined by the Census
- The specific addresses to approach – drawn from the Postal Address File
- Who should be interviewed at each address – a random selection procedure based on the ages of household members
If you would like to know more about the sampling procedures, please refer to the Technical Information.
Sometimes we are asked why NRS doesn’t use an Internet Panel for the sample, as this would be less expensive and more convenient.
It is not possible to obtain an Internet Panel sample which is representative enough to provide a media currency. Online panels tend to be skewed towards particular sorts of participant, especially heavy Internet users, and cannot represent non-Internet users. While weighting can help correct some of these differences, it is not enough for NRS purposes. Indeed, some Internet panels, such as YouGov, use NRS readership data to weight their own findings.
The following chart shows the proportion of the population likely to be represented by different sorts of sample: The NRS sample with a response rate of over 50% a high quality quota sample; and an online panel.
NRS recognises that the London and South East area is particularly important to many advertisers and publishers.
London and the South East is the most difficult area to interview in – not so much because participants are unwilling to be interviewed, but because they are less often at home.
Apart from the usual measures of making multiple call-backs, and focusing on evenings and weekends, NRS uses an additional special measure to increase the response rate in London:
- Since September 2006 London participants are offered an incentive of £20 in store vouchers for a completed interview. This led to an increase of 10% points in the London response rate, which is now at 40%.
The increase in the response rate in London has also led to an increase in the sample size. Between 2005 and 2012 the total London sample size increased by 45%; the sample size of those of social grade AB increased by 46%, and of those aged 15-24 by 69%.
South East participants have been offered an incentive since 2012 and initial reports have shown an increase of 9% points in the response rate.
A series of weighting procedures are used to ensure that the sample is balanced by sex, age, region, social grade and household size. For details of these procedures, please see the Technical Information.
The sample is grossed to population estimates provided by JICPOPS.