An independent researcher, partway through fieldwork on a customer satisfaction study, is asked by the client to add a new set of questions about a related but distinct topic — a reasonable-sounding request, usually discussed and negotiated purely in terms of scope, timeline, and additional payment. What often gets missed in that negotiation is a separate, more consequential problem: adding new questions or objectives mid-study can directly compromise the methodological integrity of the study already underway, independent of whether the extra work gets fairly compensated.
Why mid-study additions create a comparability problem
A questionnaire fielded consistently from the first respondent to the last produces data where every respondent answered the same set of questions under the same conditions, which is precisely what makes their answers comparable to each other. Adding a new question partway through fieldwork means early respondents never answered it and late respondents did, which isn't simply a missing-data problem to patch around statistically — it's a change in what was actually measured partway through the study, introducing a systematic difference between early and late respondents that has nothing to do with any real difference in the population being studied.
A study built for one objective rarely serves two well
A questionnaire designed and structured around answering one specific research question is built with that objective's logic embedded throughout — question order, response scale choices, screening criteria, sample composition. Stretching that same instrument to also answer a second, different objective added after the fact usually means neither objective is served as well as it would have been by a study designed around it specifically from the start, because the original design's choices, appropriate for the first objective, are frequently suboptimal or actively counterproductive for the second one bolted on afterward.
Why this gets framed as a business problem instead of a methods problem
Scope creep conversations tend to happen between the researcher and the client relationship, focused on fairness, payment, and timeline — practical, legitimate concerns that are easier to discuss directly than the more technical, harder-to-explain methodological cost of a mid-study change. This framing isn't wrong, but it's incomplete: a client who agrees to pay extra for an added question set has addressed the business problem while the methodological problem — a compromised, less internally consistent study — remains, regardless of how fairly the additional work gets compensated.
What actually protects against this
Treating a scope change request as a methodological question first, not just a commercial one, means asking specifically whether the addition can be incorporated without breaking comparability across the sample — sometimes it genuinely can, if it's added cleanly at a natural fieldwork boundary or run as a genuinely separate module. Where it can't be added without compromising the existing data, the more honest response is recommending it as a separate, follow-on study rather than incorporating it into the study already in progress, even if that's a harder conversation to have with a client eager for a single, combined answer.
What this means for independent researchers managing scope
- Evaluate a mid-study scope change for its effect on data comparability and study validity, not just its effect on timeline and payment
- Be willing to recommend a genuinely separate study rather than folding a significant addition into a study already in the field
- Communicate the methodological cost of a scope change explicitly to the client, not just the commercial cost, since clients often aren't aware this cost exists at all
- Build natural checkpoints into fieldwork timelines where legitimate additions can be incorporated cleanly, rather than treating any point in the field period as equally safe for changes
Protecting scope isn't just protecting the researcher's time — it's protecting the actual, specific value of the data the study was designed to produce in the first place.