Review the use of statistical research methods in your academic area
By statistical research methods I mean the sort of statistical tools covered in the quantitative part of this unit in use to answer research questions. By your academic area I mean the ‘home’ discipline (or sub-discipline), as represented by papers in reputable academic journals. If you start in your particular research niche, you may not find much use of statistics or ‘high quality’ (ABS 3 or 4) journals. In this case I suggest widening the ‘area’, e.g. to ‘retail marketing’, or even wider. Ultimately, in simple academic terms, what wide academic community would be a natural ‘home’ for your research? For example in LBS academics are organised by Subject Area Groups such as ‘Marketing’ or ‘Operations Management’.
I’m really looking for 3 components (A, B & C here)
Start with a short academic-style abstract.
- A survey of the prevalence of the use of statistical research approached
Survey a reasonable number of papers in your field, esp. concentrating on ‘good’ journals that we’d expect to have decent refereeing standards, either a year or two of 1 or 2 top journals in their academic area (best approach) or keyword searches and/or the snowball searches you’re doing for your own intended research niche – but in this case comment on ‘quality’ indicators and any relationship of this with your other conclusions
Yielding perhaps ~60 (minimum) papers to look through the abstract of – e.g. flicking through a year’s worth of a good journal… You should anyway be looking regularly at the top journals in your field! Describe briefly the method you used to identify this population.
This gives a sample (perhaps 30-40 papers??) to draw first conclusions about how often stats methods (as a whole) are used and a pool to use for B and C.
- An overview of what techniques are used and for what
Using the stats methods papers identified in scan through to collect some info for analysis (e.g. frequency tables)
- use of types of techniques (descriptives, hypothesis testing, MLR, CI etc)
- deductive vs. inductive research design
- sample sizes considered adequate
- attempts to use visual representations of data exploration and results
For this make some overall conclusions about how stats is used in the field.
Whilst doing this identify 4 interesting papers for C…
- A critique of stats methods aspects of 4 papers
Critique the statistical aspects of the methodology – compare & contrast with what we’ve talked about on the unit and Field’s guidance on the specific techniques used. I recommend sticking to techniques covered in the quants part of this unit (i.e. up to MLR). By all means go further if you wish (e.g. SEM, Logistic regression), but you will need to consult further in Field: Discovering statistics using IBM SPSS or further texts in order to evaluate whether the methodology is sound.
Include some snips of tables, graphs etc from these papers to illustrate points.
Particularly interesting would be ‘poor’ use of stats (inc. claims made from it) in ‘good’ journals!
A page or two on each. Do not write lots of description of what they did in the papers – a short intro is enough. Concentrate on the critique of how they used the methods technically (and epistemologically if relevant).
Finish with a short Disc & Conclusion, inc. reflection on why your field may have the characteristics found (e.g. maturity, paradigm, general levels of knowledge/expertise) and any evidence of a paradigm shift in the use of stats (cf The Cult of Statistical Significance).
References:
I only expect you to cite & reference the papers used in C, plus methodology texts referred to. Practise using Harvard citation and referencing style. Remember to include the page number for any text or fig or table reproduced. Also remember to exercise proper practice to avoid plagiarism (also see programme guidance). Remember all material is run through detection software.
This is an ‘opportunity’ to get deeper into the literature in your field, see how the methods you’ve encountered are used ‘in anger’, practise writing review and synthesis of literature, picking holes in others’ work, proper practice in citation, referencing & avoiding plagiarism, attention to detail (inc. spelling, formatting, page numbers, section& sub-section headings) etc.
To set expectations: a main body of 12-15 pages (4000 – 5000 words) should suffice. As usual, quality matters much more than quantity. Word count not an issue unless materially inadequate work or gratuitously excessive length…