Dear Sumia,

I reckon one of the ways you can approach this issue is the fact that classroom observation can be  obtrusive and this situation undermines the naturalness and authenticity of students' behavior including their discourse, cognitive and affective orientations as well their practical activities such as involvement in the tasks. Into the bargain, observational paradoxes might  contaminate the reliability of our findings. One   such paradox is '' relativity'' which signifies the fact that different observers may have  diverse interpretations of students' behavior. Hence , authenticity , naturalness and scientificity of the observational  findings could be debatable. Moreover, there are problems in regard of with the practicality of observations , teachers and students may be reluctant to be under meticulous scrutiny of an intimidating intrusive agent.

This might depend on your research questions. But in general objectivity of survey outcomes might be an argument to confirm a survey versus subjectivity of the observation of a single observer.

I've used both qualitative and quantitative methodologies in my research depending on the research question.

I tend to use a qualitative approach when I'm casting my net wide and I'm working with a research question that doesn't have a lot of previous research as a foundation - thus I don't really know what questions to ask on a survey (for instance). I'd just be guessing.

If there is more foundation for the research question, I can move to a descriptive methodology like a survey. I know what the results of previous research has been (both qualitative and quantitative) and I can build on that. 

I prefer a qualitative approach as it allows the researcher to deal with individual learners and gain data in a bottom up approach. A naturalistic way will allow the learners to reveal their thoughts without having to fit into the researcher's lens.

Again, I would say that it is not a manner global preferring an approach. Your decision has to be guided by the questions you like to answer and the research context. You should also think about combining both methods to generate complementary knowledge by using the advantages of both approaches.

It sounds like you are trying to capture 'natural language' use and therefore observation would be the obvious research method.  I would agree with other respondents that it really depends on the specific research questions you are asking.

Thank you all for responding to my inquiry. I am trying to see how teachers and students perceive specific linguistic features , how teachers implement related tasks in textbooks and what are students responses via questionnaires. Few teachers will be interviewed as well. 

I think if your research is into perception of tasks then you can argue for a survey.  I would only be careful in using survey to uncover natural language use, as we don't always realise how we actually use language.  Good luck with your work.

If you want to know about what teachers think they are doing, then a survey is okay, but this in no way indicates what they are actually doing. If you want to know what they are actually doing, observation really is necessary. So as others have mentioned, it really depends on your research question.  I have found Schoenfeld's "How We Think" really helpful in thinking through the issue of teacher practice in theory and practice.

Hi Sumia,

Why is it that you want to argue that a survey is better than classroom observation? By setting up that argument it appears that you are perhaps considering a survey more objective than an observation. Is this the case? Kevin.

Sumia, if you use questionnaires, there will be a problem of the individuals' perception and their actual behaviour- which may or may not match. For me, this will raise issues of validity.