Research Methods
A practical best-practice guide for designing rigorous studies on Terac, from framing your question to analyzing your results.
Good data starts long before you launch an opportunity. This guide walks through the methodological decisions that determine whether your results are trustworthy: how you frame your question, write your items, choose your sample, and interpret what you find.
It is written to be useful whether you are running a quick survey or a multi-wave study with domain experts. None of it is required to use Terac, but following it will materially improve the quality of your data.
These pages cover research design. For how to configure the platform itself, see Creating an Opportunity, Screening, and Filters.
The Four Phases
A well-run study moves through four phases. Each page below maps to one of them.
- Design the study. Decide whether you are exploring or confirming, define your population, and commit to a plan.
- Build the instrument. Write clear, unbiased questions and lay them out so participants can answer accurately.
- Sample the right people, at the right size, with enough statistical power to detect what you care about.
- Analyze with a framework you chose in advance, and aim for conclusions that hold up.
Start Here
Research Design
Exploratory vs confirmatory research, defining your target population, and avoiding HARKing
Preregistration
Lock in your hypotheses and analysis plan before you collect data
Writing Good Surveys
Phrasing, question order, response scales, and visual design
Sampling and Power
Sampling methods, representativeness, and how big your sample needs to be
Analysis and Inference
Significance testing, Bayesian methods, and building a real theory
Avoiding Bias
Common biases, measurement error, and why piloting matters
Online Research Validity
The strengths and limits of online samples, and when they fit your question