You can see in Fig 1 that it does not have to be elaborate. Just the most important points and quotes etc is sufficient. You can do this for one article, two articles, and so on. This will serve both as a guide for writing your review AS WELL AS for your generation of questions. You can already see a few questions generated there.
Step 4. Review the arguments and conclusions you have identified (arguments, the explicit premises, implicit premises, and final conclusions)
In this step, you just list as many or as few arguments and conclusions and facts that you can identify. For example, from the Diener review, we get the following:
- Meta analyses suggest that people who have high levels of happiness also have better chance of living longer compared with those who have low levels of happiness and subjective well being.
- Prospective studies with healthy people suggest that those with high levels of happiness have better indices of heart related health parameters (less likely to be suffering from high blood pressure and so on)
- Animal experiments suggest that stress related variables lead to worse health states for the experimental animals
- Human experimental evidence & randomised controlled trials suggest that induced stress levels will lead to increased steroid levels and reduced chance of healing
- High levels of happiness however do not "protect" against or there is no evidence they do for people with illnesses and those with cancers
We could go on, but you will see that overall, these are the facts that leap from the article. The more extensively you study the more you cover such facts. How much will you study? Depends on your question, and context. For this exercise, I am not insistent that you study a lot for your research. One or two articles should give you sufficient number of clues.
Step 5. Identify the pattern that these arguments & premises are leading us to? Can we identify (1) what they have asserted and argued, (2) what they have left unsaid based on what they have stated? (3) Do they form a pattern? What is the pattern? How do we articulate that pattern?
If we review all of the above facts (and let's pretend that these are the only facts that we will base our research on), what pattern emerges? How about the following:
Healthy people who are also happy are LIKELY to have good health and live long and have good heart health BUT it is UNLIKELY that people with cancers enjoy the same benefits;
Step 6. What may explain the pattern? Can we use this explanation to a more abstract statement as at least two theories that can take into account the facts and premises? What hypotheses we draw from them?
From here, your research can take several directions:
- You can only focus on one or more implicit assumptions and questions you raised in your summary of the research (this may lead to descriptive studies and cross-sectional surveys). That way you have already identified some gaps in the literature that you would like to explore
- You take a more formal, explanatory research., and start with a theory building.
We have already asked several questions on this paper, so we will not expand 1 anymore; let's see what theory and hypotheses we can come up with.
Let's have a theory that we can develop to explain why healthy people who are happy will have good heart related health but if people have cancers or illnesses yet they are happy, even then they may not enjoy the same effects (why are the impacts of being happy work differently for people with and without sickness such as cancer?).
- Perhaps illness leads to higher output of cortisol and that is a mediating variable? (cortisol theory), so our hypotheses might be to test for cortisol level for people with and without illnesses after controlling for their levels of happiness. We could then set up our null and alternative hypotheses. Null hypothesis: "People with similar levels of happiness but with illness will have similar levels of cortisol"; Alternative hypothesis: "People with similar levels of happiness but with and without illnesses will have different levels of cortisol (ill people have higher levels of cortisol)
- You can have other theories as well.
What theories you will develop will depend on what facts you uncover and choose to explain. How much facts you uncover will depend on the depth of your prior research and observation. You can start with research or you can start with observations.
Step 7. Can we identify what data we need to validate these hypotheses? How can we obtain those data?
In this case, we may need to set up a case control study and select people with and without disease conditions and measure their cortisol levels after matching these people with their levels of happiness (after giving prior questionnaires). We may need to develop questionnaires and obtain blood or body fluid samples.
For this assignment, we stop at Step 7 and reassess what we have done. It is time to put these into a narrative that includes: (1) the key facts we know, (2) the facts, ideas, and assumptions we do not know or unstated assumptions or missing bits that need to be uncovered, (3) the facts, ideas, assumptions that fit a pattern, (4) explanations and theories based on them, and (6) hypotheses and plans of data collection.