Summary
In this session, we’ll unlock the transformative power of experimental research methods, inspired by best practices in the social sciences. We’ll explore what these techniques are and how they can be used to reveal powerful insights, drive innovation. You’ll also learn how to get started, how to apply them in ethical and rigorous ways, and why these methods are essential for deep, psychology-driven insights.
Key Insights
-
•
Experimental research goes beyond AB testing by revealing why changes work, revealing causation instead of just correlation.
-
•
Testing 20 trust signals simultaneously in one experiment is far more resource-efficient than running 20 separate AB tests.
-
•
Trust is multi-dimensional; measuring it at both platform and individual pro levels offers a more holistic understanding.
-
•
Random assignment and control groups are essential for isolating effects and attributing results to manipulated variables.
-
•
Half of the 20 trust signals tested had statistically significant positive effects on overall trust in Thumbtack.
-
•
Ten trust signals demonstrated broad impact across all eight trust metrics, providing a clear prioritization roadmap for product teams.
-
•
Experimental research can help product teams prioritize smartly by focusing on hypotheses likely to succeed before costly AB tests or full launches.
-
•
Ethics in experimentation require transparency, informed consent, and minimization of harm to participants.
-
•
Hypothetical scenarios in experiments improve control but have limitations in ecological validity, which can be addressed by validating in live products.
-
•
Understanding that correlation does not equal causation is crucial for UX researchers to avoid misleading conclusions and to rely on experiments.
Notable Quotes
"When we hear experiment, most think AB testing, but experimental research goes beyond to uncover the why behind outcomes."
"The independent variable is what you manipulate, dependent variable is what you measure to see if there’s an effect."
"Randomization ensures differences between groups come from the variable tested, not external factors."
"Half of the trust signals had a statistically significant positive effect on overall trust in Thumbtack, which is a big deal."
"Testing 20 trust signals simultaneously saved significant time and resources compared to running individual AB tests."
"Trust is complex, so we measured eight different trust-related outcomes at both the platform and professional level."
"Ethics are foundational in experiments since manipulating variables can impact participants' perceptions and emotions."
"Correlation does not equal causation. To say X causes Y, you’ve gotta do an experiment."
"Experimental research bridges the gap between curiosity and action by turning ideas into evidence and evidence into impact."
"Start small with one clear testable hypothesis and build from there to make experimental research approachable."
Dig deeper—ask the Rosenbot:
More Videos
"Tracking the results of design ops initiatives allows you to tell data-informed stories, putting you on the same level as other leaders who speak data."
Jules MonzaUse These Words and Count These Things
September 25, 2024
"Retrieval Augmented Generation lets you supply specific relevant information to the model dynamically rather than everything at once."
Sarah BarrettAI in Real Life: Using LLMs to Turbocharge Microsoft Learn
February 13, 2025
"Managers don’t have to go it alone; principal ICs can be key partners for coaching and strategic initiatives."
Edward CuppsThe Principal Path: Journeying from Management to Individual Contributor
June 11, 2021
"Large language models don’t understand language like we do; they generate statistically likely text based on training data."
Savina HawkinsHarnessing AI in UXR: Practical Strategies for Positive Impact
March 26, 2024
"Who deals with chaos, ambiguity, and disruption best? It’s us, design operations humans."
Alnie FigueroaThe Future of Design Operations: Transforming Our Craft
September 10, 2025
"Qualitative reasoning about causal mechanisms usually comes from lived experience, either the researcher or participants."
Joshua NobleCasual Inference
October 6, 2023
"We hadn’t really thought about governance up front at all, which is not the ideal way to do something like that."
Dave Malouf Amy ThibodeauPanel: Design Systems and Documentation
November 7, 2017
"Our hands make all the things, and our brains form the teams that run the hands making the things."
Husani OakleyBias Towards Action: Building Teams that Build Work
June 14, 2018
"I want my collaboration tool to be like the Nest thermostat of my workday, knowing when I’m heads down and when I’m open to communication."
Abby Covert Tomer SharonPanel: Collaboration Tools
November 6, 2017
