Week 19 - Developing Good Research Ideas Flashcards
What are the key principles when developing good research ideas?
- All new research is based on past findings
- Good research questions are focused, precise and testable
- Qualitative + quantitative research can have different aims and questions
What do we mean by all new research is based on past findings?
- Researchers don’t pluck ideas from thin air
- The best research is often done by specialists with years/decades of experience in that area
-> they’re coming up with research questions based on their deep understanding of a research area - It is possible to do research in new areas - the key is to read lots first
-> what do we already know about this thing?
-> what are the key theories, findings, methods in the area
-> how robust is the current evidence?
-> what future research do researchers suggest?
example: group tasked to conduct a study exploring predictors of happiness in undergraduate university students
-> search for papers about predictors of happiness (generally + in students specifically)
-> then together you identify key trends + theories, find a specific place where the evidence is less certain, and conduct your research on that
How can we use theory to drive research?
- Theories are how we organise our findings and try to explain them
- Popper: Science evolves through developing theory, then trying to falsify it
- Some of the best research does that - takes a theory + tries to show where it is false
-> may not mean a theory is completely wrong, but just find the edges of where it applies
example: brain ageing comparison
- Research using fMRI to explore how ageing changes our brains
- Common finding is that we see more brain activity in older adults
- A prominent theory is that this reflects compensatory use of extra brain regions
-> as we get older, our brain are less able to do their job (certain areas not performing so well) so other areas activate more to try and compensate for them
This theory leads to some clear predictions:
1) Brain regions that usually do the job do not work well
2) Areas with extra activity should hold more information
Why do we need to be aware of the “gap in the literature”?
- Gaps in literature can be a useful source for new research ideas
- But you need to ask two key questions:
-Does the gap definitely exist
-> may be that you haven’t found it
-Why does the gap exist
-> is it because people haven’t got round to testing a very important question or is it because its not a useful question to test?
Also, there doesn’t have to be a gap - psychology needs more replication
Why should we evaluate past research?
- Criticality + scepticism are very important when reading, planning and conducting research
- All research has to simplify things to study them
- In psych, research is based on samples -> can’t study entire population so need to take sample
-> always going to be bias: bias in people who are willing to take part, biases in terms of who you’ve got access to - Resources (time, access, money, etc) are also limited -> restrict the ways in which we conduct research
- Some research is also just wrong
- Knowing the state of existing research helps us develop new questions
- Evaluation does not have to mean diving into a paper’s methods
- Its equally important to consider context:
-> is this one of several studies showing the same thing?
-> do other studies find something different
-> how do the findings relate to relevant theories?
-> how recent is the research?
-> what’s the history of findings in this area?
Where can you find useful synthesis?
This synthesis of findings helps understand the area
- Usually other people have often already done lots of synthesis
-> Introductions synthesise the most relevant literature in an area
-> discussion explain how a given study fits within that relevant literature
-> textbooks give an initial broad understanding
-> review articles are ideal places to find detailed synthesis
These can all explicitly (by saying we think future research needs to test A, B and C) or implicitly (by suggesting what has and hasn’t been done) give info on where research needs to head next + give you clues on where you might want to conduct research
Replicability and power
The replication crisis in psych has taught us that:
- Reported effect sizes are over-inflated
- Many studies lack sufficient power
-> don’t have enough ppts or big enough effects to actually detect what they’re looking for
This has two major implications for research planning:
1) Be careful when choosing past research to base your ideas on
2) Your own research should avoid these pitfalls by considering power
-> do you have enough ppts and its your effect size going to be big enough to find the thing you are looking for?
What are practical outcomes for this section?
- Research should clearly follow from past findings
- Research questions should aim to test theory
- A “gap in the literature” is not sufficient justification for a study
- You can learn about an area, including possible future directions, but reading other people’s synthesis
- Beware of issues of power and replicability (in past research + your own)
Why do good research questions need to be focused, precise and testable?
Focused and precise research questions:
- Most research does not lead to huge theoretical developments
- Instead it tests the edges of our understanding
Testable research questions:
- You should generally have clear hypotheses before you start collecting data
- This is called confirmatory research
-> you have an idea of what you’re expecting + you’re conducting research to confirm whether that is the case or not the case - Some research is exploratory: no clear hypotheses, just collect data and see what turns up
- Replication issues can occur when exploratory analyses are presented as confirmatory
-> preregistration can help prevent this - preregister what they’re planning to do before they collect the data and conduct the analysis
Why is it good to build replication into studies?
- Good studies build in some amount of replication
- Helps to check your research is working
Qualitative goals vs quantitative goals
- Quantitative research hypothesises, tests, analyses, confirms, rejects
- It looks for relationships, effects, predictors, outcomes
- Qualitative research explores experiences, narratives, ideas, language, identity
- Often does not test hypotheses or aim to replicate findings
e.g. ten people, interviewed in depth about the same topic will give different answers - that is okay
-> looking for diversity + richness in qualitative research - The questions and approaches to developing them are therefore different
Many of the broad points still apply:
- Questions should be developed in relation to existing literature
- Have a strong idea of methods and analyses before starting