Many sample companies with little research expertise will talk about quality as part of a memorized sales pitch. But do you they really understand how to determine if sample is representative of a population? If you ask them for recommendations on design, would they be qualified to give them without their head exploding? We insist on full transparency from our sources in terms of the composition of their active panelists and the sources from which they are recruited. We look not only at demographic biases but psychographic biases. We will put a soft launch data set in SPSS and look at the data from the perspective of researchers. We will clean your data to your specifications even if we are not programming upon request. Clients can ask us for design recommendations and we will make them based on your research objectives, not the strengths or weaknesses of any one panel or sample source. As researchers who take the time to understand the objectives, business applications, the potential problem areas- we are in a much stronger position to add value in the area of quality than people who do not make this effort. Of course, we have Relevant ID to prevent duplication and we will recommend thresholds for speeding, straightlining, and help construct red-herring questions appropriate to the subject matter. We will frequently redesign a screener or finalize a questionnaire that a client has drafted. When quality is a function of both the survey instrument, the programming, AND the data collection methodology, how can non-researchers speak credibly about quality when they haven’t designed or tested (rigorously) thousands of surveys like we have?
