Longitudinal Study
A longitudinal study follows the same participants over an extended period, collecting data at multiple time points. In SLA research, longitudinal designs track how learners' interlanguage develops over time — capturing acquisition trajectories, developmental sequences, U-shaped development, and fossilization as they unfold.
Why Longitudinal Designs Matter in SLA
Language acquisition is inherently a process that occurs over time. Cross-sectional designs can only infer development by comparing different groups at different stages. Longitudinal studies observe actual change within the same individuals, making them the only design that can directly document:
- The order in which structures emerge
- Periods of rapid progress, plateau, and regression
- Individual variation in acquisition trajectories
- The relationship between instruction and developmental readiness
Duration and Scope
There is no fixed minimum duration. In SLA, longitudinal studies range from weeks (e.g., tracking a single grammatical feature during an instructional intervention) to years (e.g., documenting overall proficiency development in immigrant children). The defining feature is repeated observation of the same participants.
Landmark Longitudinal SLA Studies
| Study | Duration | Focus |
|---|---|---|
| Schumann (1978) — Alberto | 10 months | Naturalistic acquisition; pidginisation hypothesis; social distance |
| Schmidt (1983) — Wes | 3 years | Communicative vs grammatical competence; Acculturation |
| Cancino, Rosansky & Schumann (1978) | 10 months | Negation development in 6 Spanish L1 learners |
| Bardovi-Harlig (2000) | Multi-year | Tense-aspect acquisition across proficiency levels |
| Vainikka & Young-Scholten (1996) | Extended | Organic Grammar in L2 syntactic development |
Challenges
- Attrition — participants drop out over time, potentially biasing the remaining sample
- Time and resources — data collection and analysis are labour-intensive
- Researcher consistency — maintaining consistent procedures across months or years
- Practice effects — repeated testing may improve performance independently of acquisition
- Generalisability — small samples (often case studies) limit External Validity
Longitudinal vs Cross-sectional
| Feature | Longitudinal | Cross-sectional |
|---|---|---|
| Tracks real change | Yes | Inferred |
| Time required | Long | Short |
| Attrition risk | High | Low |
| Individual variation | Visible | Hidden in group averages |
| Sample size | Usually small | Can be large |
The two designs are complementary. Cross-sectional studies generate hypotheses about developmental patterns; longitudinal studies test whether those patterns hold within individuals.
Key References
- Ortega & Iberri-Shea (2005) — review of longitudinal research in SLA
- Larsen-Freeman (2006) — dynamic systems perspective on longitudinal L2 development
- Mackey & Gass (2005) — research methodology guidance for longitudinal designs