Operationalisation is the process of defining an abstract concept in concrete, measurable terms. It answers the question: How exactly will you measure this? Every research study must bridge the gap between the theoretical Construct (e.g., "proficiency," "noticing," "motivation") and the specific instrument or procedure used to measure it.
The way a construct is operationalised determines what the study actually measures, which may or may not match what the researcher claims to measure. Two studies investigating "the effect of Corrective Feedback on acquisition" may operationalise "acquisition" differently:
| Operationalisation | What it actually measures |
|---|---|
| Timed GJT accuracy | Implicit knowledge of the target structure |
| Untimed GJT accuracy | Explicit knowledge: ability to reflect on rules |
| Spontaneous production in free conversation | Automatised productive knowledge |
| Elicited Imitation accuracy | Implicit oral grammar under time pressure |
| Metalinguistic explanation | Declarative, analysed knowledge |
These are different things. A treatment that improves GJT scores but not spontaneous production has affected explicit but not implicit knowledge. Claiming it "promotes acquisition" depends entirely on how "acquisition" was operationalised.
Theory → Construct → Operationalisation → Measurement → DataEach link in the chain introduces assumptions that can be questioned.
Construct Validity asks whether the operationalisation actually captures the intended construct. If "proficiency" is operationalised as a multiple-choice grammar test, the study measures grammatical knowledge in a decontextualised format, not the full construct of communicative language ability. This is why Bachman & Palmer (1996) argued for test tasks that reflect real-world language use (see Authenticity).
| Construct | Common operationalisations |
|---|---|
| Proficiency | IELTS band, TOEFL score, Elicited Imitation, institutional placement level |
| Noticing | Think-Aloud Protocol reports, Stimulated Recall, eye-tracking fixation data |
| Motivation | Questionnaire scales (Dörnyei's L2MSS), interview data |
| Acquisition | Pre-post gain on GJT, emergence in production, Effect Size of treatment |
| Fluency | Speech rate, pausing, hesitation phenomena |
| Complexity | T-unit length, clause subordination ratios |
When a construct is operationalised too narrowly, important aspects are missed. "Writing ability" operationalised as grammatical accuracy alone ignores coherence, task achievement, and lexical range. "Acquisition" measured only by explicit knowledge tests ignores implicit, procedural knowledge.