This article is aimed at students undertaking a research dissertation or thesis and it explains concepts students often find difficult, plus personal reflections on writing for peer reviewed academic journals.
My earlier article, ‘How to Prepare a Research Proposal and Literature Review: Hints for Accounting, Finance, and Investment Students’ is aimed at undergraduate, postgraduate and research students to help them identify their dissertation topic and key literature. This second article guides students through the structure of the literature review and how to articulate research hypotheses.
Even diligent students are often unsure how to structure their literature review and how much detail to include about the papers they have read. While there are several styles, the method I outline here is a useful starting point for students who have not previously written an accounting and finance dissertation, or students whose first language is not English. The purpose of the literature review, is to identify and motivate the research question(s). Questions are then framed in the literature review as formal hypotheses; thus providing a structure that subsequent sections of the dissertation can refer back to.
According to the Oxford English Dictionary, the singular term hypothesis and plural hypotheses means proposed explanation(s) based upon limited evidence that can be used as a starting point for further explanation. For example, supposing you are interested in the general question of whether sunshine influences stock returns. Your theory might be that sunshine makes people happier – happier people are more confident, and thus more likely to take risky decisions, such as buying stocks – so stock prices rise on sunny days. To develop this theory in the literature review, you would review papers from behavioural finance and psychology to see if there is any established theory and evidence linking sunshine, to happiness, and happiness to risk taking behaviour. You would also search for evidence of competing theories as to why stock prices might, or might not, rise on sunny days. If after a lot of reading, you find that the most convincing research seems to support your theory, you then frame your research question in the form of a testable hypothesis as follows:
Hypothesis I: There is a positive relationship between number of hours of sunshine and stock returns
Having identified your theory and hypothesis, a normal approach is to clearly identify and explain your theory, summarising the main supporting and opposing evidence from the literature. If your argument in support of your hypotheses draws upon several topics, or strands of research, it may be advisable to have separate subsections each dedicated to the relevant literature from a given topic. Finally before stating your hypotheses, it is useful to have a paragraph drawing the various topics together, thus organising the evidence in support of your theory and hypothesis.
Some papers and dissertations examine several related hypotheses. However, it is arguably best not to have too many hypotheses in a single undergraduate or masters dissertation in case it gives the impression that you are trying to justify a mixed set of results after the event (ex-post), an approach that is not good practice. Instead, it is usually better to have fewer hypotheses, but justify them more rigorously and in more depth, sometimes just one or two may be enough.
Types of hypotheses
The above example is of a ‘one-tailed hypothesis’ in that the direction of the relationship is specified – more sunshine – higher stock prices. However, if your literature search has identified conflicting evidence, i.e. some in support of a positive relationship and some in support of a negative relationship, you can hedge your bets by setting up your research question as a two – tailed hypotheses. You could then summarise the literature indicating no relationship concluding with your null hypothesis along the lines of:
Hypothesis I 0: There is no relationship between number of hours of sunshine and stock returns
Follow this up by explaining literature in support of a positive relationship concluded with
Hypothesis I a: There is a positive relationship between number of hours of sunshine and stock returns
Next, summarise and explain the evidence and literature supporting a negative relationship and conclude with
Hypothesis I b: There is a negative relationship between number of hours of sunshine and stock returns.
How many papers should I cite in my literature review and in how much detail?
The literature review should identify the most important papers relating to the theory and empirical evidence underpinning your research question and methods. It is likely that you will talk about the most important papers in detail, drawing attention to similarities and differences between their questions, their method of addressing them and yours. Other papers may just replicate the seminarl papers, or they may be tangential to your research question, perhaps dealing with control variables, so you only need to mention them briefly.
It is likely that part of your assessment grade will be based upon your choice of literature and how you justify it. Therefore, which papers you identify and how much time you spend discussing them is more important than the total number. So, rather than worrying about how long your reference list is, or how many papers you have cited, it is better to think about why you have included those papers and how much space you should dedicate to discussing them. As a general rule, a reference list containing between twenty and thirty papers is usually long enough, but there may be exceptions in either direction. If in doubt, you should discuss this with your supervisor.
Isaac T. Tabner 