Research Reactivity
The possibility that being studied, observed, prompted, or placed in a research setting changes participant behavior or responses.
Download Markdown (.md)Soroush’s Product Wiki · https://wiki.soroush.uk/topic/research-reactivity/
01What it means
Research reactivity occurs when the research situation influences what participants do or say. Awareness of observation, facilitator cues, study demand characteristics, unfamiliar tools, and social desirability can all produce behavior that differs from ordinary context. The Hawthorne effect and observer effect are related labels, not guarantees that every observed change has the same cause.[1][2]
02Why it matters
Product teams may interpret compliant task behavior, positive feedback, or unusually careful performance as normal use. Recognizing reactivity helps researchers design less intrusive studies, separate observed evidence from inference, and avoid treating laboratory behavior or self-presentation as an unfiltered view of reality.
03When to use it
Consider reactivity in moderated sessions, field observation, diary studies, think-aloud work, sensitive interviews, and any study where participants can infer a preferred answer or behavior. It matters most when the behavior is socially judged, habitual, safety-related, or easily altered by attention.
04How it works
- Identify how observation, instructions, incentives, recording, setting, and researcher identity could change behavior.
- Use neutral language and facilitation, realistic tasks, familiar tools, and the least intrusive setup that answers the question.
- Allow acclimatization where appropriate and record interventions or unusual session conditions.
- Compare self-report with behavioral, longitudinal, contextual, or unobtrusive evidence when ethical and feasible.
- State likely reactive effects and avoid claiming that the study reproduces natural behavior exactly.
05A practical example
During a budgeting-app test, participants carefully categorize every purchase while narrating aloud because they want to appear financially responsible. Live product data later shows frequent uncategorized transactions. The team treats the session as evidence about interface comprehension, not proof of everyday maintenance behavior.
06Where it breaks
Reactivity cannot be eliminated completely and is difficult to measure directly. Less visible observation raises consent and privacy concerns, while naturalistic methods sacrifice control and may still influence participants. The popular historical explanation of the Hawthorne effect is contested, so it should not be used as a universal causal story.
Common mistakes
- Assuming behavior observed in a session is identical to routine behavior.
- Leading participants through cues, praise, or rescue and then treating the outcome as independent.
- Trying to hide observation in ways that violate informed consent or participant expectations.
07Sources
- The Hawthorne Effect or Observer Bias in User ResearchNielsen Norman Group · Professional guidanceThe Hawthorne Effect or Observer Bias in User Research
- Reactivity and good data in qualitative data collectionEuropean Journal for Philosophy of Science · Peer-reviewed researchReactivity and good data in qualitative data collection