Quality Control measures

In order to ensure accurate and reliable results of fieldwork, Research Matrix Limited emphasises quality control as an integral part of any project undertaken. Specifically, the following procedures apply to all our projects:

  • Central briefing of all key field personnel.

  • Pre-briefing and selection of only experienced interviewers in each study locations.

  • Piloting of questionnaire and modifying for clarity and comprehension

  • Field supervisors work first as interviewers during training before assuming supervisory roles.

  • This is done in order to familiarise them with the methodology and intricacies of the study. The supervisors carry out the following tasks in the course of any project:

    Reviewing of all completed questionnaires for legibility, accuracy and consistency Monitoring the accuracy of individual interviews Spot field checks on interviewers to eliminate fraud and in accurate form filing Accompanying and back checking of each interviewers work; as a rule a minimum of 20% accompaniment and back checks is carried out by supervisors and quality control officers (ACCP = 5%, BC = 15%) Group sampling method is adopted.

    This technique requires that a field team (comprising one supervisor and between 4 to 6 interviewers) complete a sector or an area before moving to another sector/area as a group. The greatest advantage of this method is the relative increase it affords in the precision and reliability of results as it eliminates the possibility of fraud by individual interviewer.

  • The Applied Quality Control Technique (AQCT) is an additional quality control measure employed by Research Matrix Limited. This is basically done by Research Matrix Limited Quality Control Officers and it includes;

    1. –Spot checks - unannounced visits to study locations
      –Editing – Random selection of completed questionnaires for editing
      –Complete back checks - Going back to respondentshouse by Quality Control to confirm interview (PEC)
      –Random editing - Cutting across dates, sectors and interviewers
      –Checks on supervisors activities
  • All interviewers are given unique identity codes (ID number), which they must write on all the questionnaires they operate. No duplication of interviewer ID is allowed. Where an interviewer is dismissed or leaves before the end of a project, the replacement interviewer is be given a new ID.
  • Rigorous (100%) editing of all operated questionnaires is done in house. Any questionnaire found to contain errors or inconsistencies is rejected.
  • Only experienced coders are employed in the coding of questions
  • Double verification and cleaning of punched data (25%) is carried out.