Is it reliable and valid if it is not replicable? On the importance of replicability in quantitative research

Abstract

The idea of this paper arose in a reading group of several colleagues at the Faculty of Philology of Vilnius University after a discussion of a review article published by the editors of Language Teaching. Titled ‘Replication studies in language learning and teaching’ (2008), the paper focuses on replication studies and argues that they should be promoted and valued no less than original research. The participants of the reading group agreed that replication studies, understood here primarily as replications of quantitative research, are indeed an important issue that could be of interest to the broader community of applied linguists in Lithuania. The present paper argues that attempts to replicate earlier studies, which are very scarce or non-existent in Lithuania, deserve more attention both from novice and mature researchers. Replications are particularly valuable in developmental studies where replicating a study over a period of time allows the researcher to obtain data for continued analysis. Furthermore, a replication of a published study that deals with data collected in one country offers an opportunity to verify its findings in a different context and this way consolidates our understanding of phenomena under study. Finally, replication is an invaluable learning method to a novice linguist, be it a senior undergraduate or postgraduate student. Thus the authors of this paper would like to promote the idea of replication research in our community as well as encourage everyone interested make use of the increasingly growing amount of open access data available on the internet.