Observations on simplified or weak use of data in linguistics
For many years I have actually quit reviewing a great deal of used grammars research in the journals that I utilized to adhere to or occasionally look at. Aside from time restraints and being spread too thin in my passions, there’s a factor I obtained annoyed with the study high quality in a number of short articles I saw.
The problem
It appears that a great deal of linguists take at many a couple of initial statistics courses, which’s it. And afterwards numerous published articles are basic research studies with just t-tests and ANOVAs, like that’s all they can do. Now, if it’s an initial exploratory study, I can comprehend. However I pertained to see many articles with just those two methods, and maybe basic correlations. In language education, there are a variety of factors that researchers probably require to manage for. With language learning, research study has revealed the impacts of personal and sociocognitive elements like inherent/ extrinsic inspiration, personality aspects (extroversion, cognitive visibility, etc). For any type of study, you can possibly think of a host of possible confounding variables. Okay, again, it’s alright if it’s an initial, exploratory research, and the authors make that clear, and the research study later on obtains reproduced with extra sophisticated layouts and analytical designs. However …