Quality Control Measures

All our work is conducted according to the Market Research Society code of conduct. In terms of specific interviewer/data checks, we ensure the following are conducted;

1. All questionnaires are manually checked for obvious errors - if a batch contains a high number of obvious errors then interviewer/respondents are re-contacted and asked to clarify responses.
2. 10% of all respondents are re-contacted following the interviews (back-checks) to ensure they were interviewed, they were happy with the format of the interview, that they were satisfied that the interview was conducted fairly etc.
3. 10% of all questionnaires are re-entered manually and a report is issued which shows comparisons of original and re-entered responses for the 10% i.e. different codes to each other, codes outside the coded responses on questionnaire etc. Any errors are then corrected. If a large number of errors are highlighted, 20% are checked and so on.
4. A 'no reply' report is produced to highlight missing responses.
5. An 'error' report is produced i.e. any nonsensical responses highlighted.
6. Routing is checked by assessing base figures on frequency counts i.e. no questions will have responses that shouldn't have.