Imagine you wanted to research about L2 Japanese learners approach the kanji learning task, how you would go about trying to measure that say in an experiment?
Trying to work out some ideas, so please lend me your thoughts and ideas.
I wouldn’t use an experiment. I’d start with an exploratory study using unstructured or semi-structure interviews. In between interviews I would start categorizing the data and amend the structure of future interviews as I identified new potential areas of interest, and I would keep doing interviews until new interviews stopped yielding new insights into different approaches to learning kanji. Then, if I wanted to know specifics about quantitative data like how common these approaches are in a particular population of learners, which approaches are usually used together, etc, then I would design a questionnaire around the different approaches identified in the interviews and try to distribute them among a lot of learners.
Whether it’s a general approach to kanji learning or learning how to read (these are 2 different things, mind you ), I would recommend building a more structured response sheet/survey. Ideally you want categorized data to simplify quantification. Also, we would probably need to know a little bit more about the overall goal, methodology, etc.
My response is that experiments are used to test hypotheses (for example, “the effects of repeated application of electric shocks on L2 Japanese learners”). If you just want to gather information (for example, “how L2 Japanese learners approach the kanji learning task”), you do a survey.