Quantitative Research
Quantitative research uses numerical data and statistical analysis to test hypotheses, measure variables, and identify patterns. It operates within a positivist or post-positivist paradigm that values objectivity, generalisability, and replicability. In SLA and applied linguistics, quantitative approaches dominate intervention studies, assessment research, and corpus-based investigations.
Core Features
- Measurement — abstract constructs (proficiency, motivation, anxiety) are operationalised as numerical variables (see Operationalisation)
- Hypothesis testing — research begins with predictions derived from theory
- Statistical analysis — inferential statistics determine whether observed patterns exceed chance
- Control — extraneous variables are controlled or accounted for through design or statistics
- Replicability — procedures are documented precisely enough for Replication
Common Designs in SLA
| Design | Purpose | Example |
|---|---|---|
| True experiment | Causal inference with random assignment | Lab study of recasts vs prompts |
| Quasi-experiment | Causal inference with intact groups | Classroom comparison of FonF approaches |
| Correlational | Relationship between variables | Language Aptitude and proficiency outcomes |
| Survey | Attitudes, beliefs, practices at scale | Language Anxiety across populations |
| Meta-analysis | Synthesising effect sizes across studies | Effectiveness of Corrective Feedback |
Statistical Tools
Descriptive: means, standard deviations, frequencies, distributions.
Inferential: t-tests, ANOVA, ANCOVA, regression, structural equation modelling, factor analysis.
Effect-focused: Effect Size (Cohen's d, Hedges' g), confidence intervals — increasingly emphasised over p-values alone.
Strengths
- Large samples enable generalisation across populations
- Statistical controls mitigate confounding variables
- Findings are replicable and comparable across studies
- Meta-analyses depend on quantitative primary studies
- Clear criteria for evaluating evidence (Internal Validity, External Validity)
Limitations
- Reductionism — complex phenomena (identity, motivation, classroom interaction) are reduced to numbers
- Context-stripping — controlled conditions may not reflect real classroom life (see Ecological Validity)
- Construct validity concerns — what is measured may not match the intended Construct (see Operationalisation)
- Overreliance on p-values — "significant" results can be trivially small; non-significant results can mask meaningful effects
Reporting Standards
The field has moved toward more rigorous reporting. Plonsky (2013, 2014) called for mandatory Effect Size reporting, confidence intervals, and transparency about analytical decisions. The APA 7th edition reinforces these standards.
Key References
- Dörnyei (2007) — Research Methods in Applied Linguistics
- Larson-Hall (2010) — A Guide to Doing Statistics in Second Language Research Using SPSS
- Plonsky (2013) — quantitative research methodology in SLA
- Loewen & Plonsky (2015) — An A–Z of Applied Linguistics Research Methods