Description
Requires a written, bounded definition of each construct, its dimensions and its boundaries before any instrument is chosen or data collected.
Many measurement failures begin with an unspecified construct: teams collect 'engagement' or 'delight' data and then decide afterwards what it meant. Construct definition before data collection inverts this. The researcher writes a conceptual definition that states the construct's domain, its dimensions, whether it is reflective or formative, what it is distinct from, and the conditions under which it applies. MacKenzie and colleagues recommend checking definitions for ambiguity, clarifying the entity it applies to (user, session, product), and separating the construct from its causes and consequences. This prevents circularity, where a construct is defined by whatever the available metric happens to capture, and makes later validity testing possible.
Attributes
- Discipline
- Research Design & Rigor
- Loss type
- Latent
- Stage
- Definition
- Data
- Analytical / expert
- Effort
- Days
Steps
- Name the construct and collect existing definitions from literature and from stakeholders.
- Write a conceptual definition specifying its property, the entity it describes, and its domain.
- List dimensions and decide whether the construct is reflective (indicators caused by it) or formative (indicators define it).
- State what the construct is not, naming at least two neighbouring constructs it must be distinguished from.
- Have 3-5 experts rate whether the definition is clear and complete, and revise until agreement is reached.
- Lock the definition in the research plan and choose or build instruments only after this.
Sources
- Dix, A., Statistics for HCI: Making Sense of Quantitative Data (2020)
- Sauro, J. & Lewis, J. R., Quantifying the User Experience (2nd ed., 2016)
- Leavy, P., Research Design: Quantitative, Qualitative, Mixed Methods, Arts-Based, and Community-Based Participatory Research Approaches (2017)
- Statistics for HCI (Alan Dix)
- Quantifying the User Experience (Sauro & Lewis)
- Research Design (Leavy)
- The Inquiry Series, Ch 8 — Measurement Theory — Before You Measure Anything
Also known as
- A priori construct specification
- Construct-first research design
- Conceptual definition
Categories
- Research Operations
- Research Design & Rigor