No, there would be no complications when a researcher uses two units of analysis. Evidently, even when using two units of analysis, the researcher would make one unit the center of focus in his research. According to Abbasi (2017) there can be no complications when using two units of research as long as the researcher makes careful planning and designing. In my study, I meant the unit of design will be learners’ satisfaction though in the context of classroom learning and not literally learners and their satisfaction. In the discussion about population definition and sample size, ‘the reason why the samples are going to be drawn from an accessible population is because the researcher will have narrowed down the aspects of the research’ is used to refer to my own research though with reference to what researchers in this field have done before. So I would say that this statement points to my own research but also looks into what previous researchers have done in the past.

When it comes to sampling, it is necessary that a researcher makes a perfect choice of the right population he wishes to use and definitely random sampling would not deliver to the researcher the best audience. Sampling and choice of research participants should be very specific with particular information to be used in searching for the right participants. It is this kind of sampling according to Cohen, Manion & Morrison (2011) that will bring the researcher to a successful attainment of the results needed in the research. Purposive sampling would be best for this kind of research.