Quest Research Services has a full functioning Data Processing department that directly reports to the Research Director. Our team is composed of qualified and experienced Statisticians, Systems Analysts/Programmers and data capturers with more than 10 years’ experience as individuals in their respective areas of specialisation.
Data is processed in batches or index and various softwares are used, conditional to client requirements and deliverables. We employ a comprehensive data entry system to ensure that only valid data is entered, followed by a thorough data integrity check, coding of open-ended responses and categorisation of all responses to create meaningful and useable data.
Frequency tables and crosstabs are run to compare different variables to have an optimal true reflection of the field under study.
In case of any anomalies, our Quality Control team, Call Centre agents and Field coordinators will be on standby to do all back checking or redoing the interviews to gather more, reliable and verifiable data. Over the years, Quest Research Services proudly has produced quality, timely and value added data. QRS has extensive knowledge in all the stages of the survey process. This includes tabulation plans, metadata specifications, system designs and questionnaire design. As specialists in data analysis our expertise includes:
- Data Cleaning and Validation (Logic/Consistency Checks)
- Conversion of Captured Data into Utilizable Data Files using relevant software (i.e. SPSS, Excel, GIS Formats, ASCII etc.)
- Basic Data Analysis i.e. Runs such as frequencies
- Advanced Statistical Analysis
- Data Weighting
- Data Dissemination and Reporting
Our analysis software program can perform over 10 basic types of analysis of the captured data, including examining raw data, averaging triggered sweeps sorted according to a number of criteria, obtaining time series, correlation between variables, scatter plots as well as all types of statistics and probability distributions.
After data sets are cleaned and the data is validated on consistency, correctness and completeness, it will then be weighted with weights generated in consistency with current subscriber population estimates.