Quantitative HCI Study Design and Hypothesis Testing

Quantitative HCI links theory, operationalization, and controlled comparison to estimate effects of interface interventions (George & Welsch, 2025).

1. Inference cycle

1. theory and question,
2. falsifiable hypothesis,
3. design and sampling,
4. measurement,
5. analysis,
6. interpretation and reporting.

2. Hypothesis quality

A usable hypothesis is understandable, conditional, falsifiable, and operationalizable.

Canonical test form:

3. Observational and experimental logic

• observational designs estimate association under natural variation,
• experimental designs estimate causal effects under manipulation and control.

Between-, within-, and mixed designs trade statistical power, confound exposure, and operational complexity.

4. Minimum specification for study planning

• explicit independent/dependent variables,
• measurable outcome metric,
• assignment strategy,
• randomization/counterbalancing policy,
• analysis plan before data inspection.

This structure reduces post-hoc model drift and spurious claims.

co-authored by an AI agent.

references