It appears that during interviews in chat communities, when asked more than 3 questions using the iResearch methodology, consumers loose interest. Based on our analysis of this phenomenon, the value to the consumer in answering up to 3 questions can be broken down into 4 reasons:

a) the quality of the interaction,
b) the innovation of the approach,
c) the satisfaction of having the choice to share an opinion or meeting someone new to interact with, and
d) the convenience of the approach, short and sweet, completely different from traditional requests for feedback.

More than three questions and the value the consumer perceives diminishes exponentially.

As a result, al berrios & co. divides multi-question surveys into smaller surveys of two or three questions, fielded by iResearchers also divided into smaller units. Each unit only asks their assigned questions. However, multiple units can survey in the same communities without impacting results. Stability is achieved from quantity.

Here is an example: A member from iResearch Unit 1 asks 10 individual Country Music Community members about their favorite programming and favorite actors. A member from iResearch Unit 2 asks 10 more Country Music Community members whether or not they have cable and a particularly favorite cable network. The likelihood of these community members overlapping is approximately 70% to 80% (since active community members - leaders - are usually the only ones that respond to our survey requests consistently), meaning that the same community member was likely asked all 4 questions, with the exception that the questions came from two individual interactions (1). This approach yields an approximate 90%+ response rate (depending on the type of survey), compared to a close-to-0% completion rate for an in-depth 20-question survey in a chat community. And based on side-by-side comparisons with traditional in-field qualitative surveying, this approach yields superior results.

Footnotes
(1) All estimates based on informal review of historical samples.