Complex dynamic systems theory (CDST) studies have often been set up as single case studies or studies with few participants. This is sometimes considered a limitation of CDST as it endangers the generalizability of the findings (Bulté & Housen, 2020). CDST-researchers have begun to investigate the possibilities of discerning ergodic groups (i.e. groups where the individual is similar to the group and vice versa) as being able to identify ergodic samples makes it possible to make claims both about individuals and groups. Lowie and Verspoor (2019) divided their participants in highly similar groups but found that even in these similar groups there were large differences in learning trajectories over time.
In the present study we will further investigate the possibilities of discerning ergodic groups and thus the possibilities to do group analyses when investigating language development. In this study we will focus on the development of young L2 English learners' speaking skills. 64 learners who attended the first year of secondary school participated in a longitudinal study in which they were asked to do a speaking activity every week over the course of one school year. They were also tested extensively at the start of the study in order to map individual differences between the learners (e.g. differences in prior knowledge, instruction, out-of-school exposure, motivation and cognitive differences).
The learners belonged to three different class groups which each had a different profile. Group 1 (n = 21) attended a school in Flanders. They did not receive any English lessons in primary education and were at the start of formal English education. Group 2 (n = 22) attended a school in Flanders. They did not receive any English lessons in primary education and did not yet have formal English lessons. Group 3 (n = 21) attended a school in the Netherlands. They started with English lessons in primary education.
In order to investigate ergodicity, we will first look into speaking development through group analyses starting from the three pre-defined groups in our study. Through cluster analysis (cf. Peng et al., 2021), we will then further explore whether other groups of learners can be identified based on individual difference variables measured at the start of the study and whether these learner types can be seen as ergodic groups when investigating speaking development.
The results and implications of our study will be discussed during the presentation and suggestions will be done for future studies.
References:
-Bulté, B., & Housen, A. (2020). Chapter 9. A critical appraisal of the CDST approach to investigating linguistic complexity in L2 writing development.In G. G. Fogal & M. H. Verspoor (Eds.), Language Learning & Language Teaching (Vol. 54, pp. 207–238). John Benjamins Publishing Company.
-Lowie, W. M., & Verspoor, M. H. (2019). Individual Differences and the Ergodicity Problem: Individual Differences and Ergodicity. Language Learning, 69, 184–206.
-Peng, H., Jager, S., & Lowie, W. (2021). A person-centred approach to L2 learners' informal mobile language learning. Computer Assisted Language Learning, 1–22.
