Practice MRF

You may be saying the MRF stands for “Man, Research Frustrating.” For those struggling with Capella’s BMGT8032: Survey of research methods, or for other dissertation students working on their proposal, here are a few thoughts. 

Research questions 1.6.  Your research questions need to have clear measures, you have to be able to get in touch with the participants, and they have to be stated in statistical language.   If you don’t have these three things, you don’t have answerable research questions. 

Sample size 2.2.  Sample size is a tricky thing, and maybe the order of writing has something to do with this.  Capella has this section as 2.2, which talks about the sample—fair enough.  However, since the preponderance of dissertations use a power analysis, and the power analysis is different based on the statistics used, the sample size justification (section 2.2) should go after, not before, the data analysis plan 2.5.  The best thing to do is to make sure you have the correct analysis, then use G-power (which is free) or go to our membership website page basic-membership for a write-up ($29). 

Measurement 2.3: First of all, this will become part of your dissertation, so make sure that you have constructs that are measurable.  If you are the first person to measure a particular construct, expect a few extra months to pilot test the instrument, then you still have to assess the reliability and validity of the new instrument.  Don’t reinvent the wheel—find a reliable and valid instrument that exists. Worst case, adapt a reliable and valid instrument, and use a change cross-walk in the appendix to show how your adaptation is different.

Data analysis plan 2.5.  The data plan is comprised of three components: which analyses are appropriate to assess your research questions, what are the assumptions of the selected analyses, and a justification of why the analyses were select.  The appropriate analysis is selected based on the way the research question is phrased (i.e., “difference” questions presume ANOVA type analyses) and the level of measurement of the variables (i.e., ANOVAs presume an interval or ratio level dependent variable and a nominal level independent variable).  The assumptions of an analysis can be found on our website (www.statisticssolutions.com) or elsewhere on the web.  And finally, the justification of the analysis combines the above two points by simply stating that given the research question phrasing and the level of measurement, this particular statistical test is appropriate. 

Certainly, any questions, feel free to call us (877) 437-8622 or email us [email protected]