L1 - Research Questions Flashcards
What are the 4 steps in developing a research question?
Step 1 - Find a broad topic
Step 2 - Review the literature
Step 3 - Identify your research problem
Step 4 - Refine your initial, broad question
Define Step 1 of developing a research question:
- Do not be too narrow initially
- Find a broad topic that you are interested in and motivated by
Define Step 2 of developing a research question:
Literature Review
- What do we and do we not know?
- Is the existing information conflicting?
- Critically appraise the literature
Define Step 3 of developing a research question:
Identify a research problem
- Systematic reviews can be helpful
Define Step 4 of developing a research question:
Refine the question:
- Define specific terms
- Specify populations, outcomes, comparators
What are the criteria for a good research question?
- The question has not already been adequately addressed
- It passes the ‘so what’ test
- The question is answerable
- The question is feasible
What is an independent variable?
AKA predictor variable or exposure variable
An independent variable is the condition, intervention or characteristic that will predict a given outcome
What is a dependent variable?
AKA outcome variable or measure, or predicted variable
A dependent variable is the response or effect that is presumed to change depending on the independent variable
What is an experimental study?
The researcher does something to the participants (manipulates the independent variable) and measures any changes to the dependent variable
What is an observational study?
The researcher observes a group and has no control over what happens. They measure the dependent variable in pre-existing groups of participants.
What is the formula for a RQ in an experimental study?
Is there a difference in (DV) in (population) who (IV 1) and (IV 2)?
What principle is used for the design of an experimental study RQ?
The PICO Principle
Explain the PICO Principle:
P - Patient/Population
I - Intervention/IV 1
C - Comparison/IV 2
O - Outcome/DV
What is the formula for an RQ in an observational study?
Is there an association between IV and DV in population?
What principle is used in the design of observational studies?
The PECO Principle
Explain the PECO Principle:
P = Patient/Population
E = Exposure/IV/Predictor/Risk Factor
C = Comparison IV
O = Outcome
What is the formula for a RQ in qualitative research?
What is (phenomenon under investigation) of population
e.g. What are the school experiences of people who stammer?
What is sampling and why do we do it?
Taking a sub-group that is representative of a full population to take part in the study - this is more resource-effective and feasible than studying a whole population.
What is a sampling error?
The error between the mean in a sample compared to the mean in the population.
Why may samples be different?
- Samples were taken from 2 truly different populations
- Samples were drawn from populations that were not truly different, but a difference was found due to sampling error
Why may samples not have a difference?
- Drawn from 2 populations that did not differ
- Samples were drawn from 2 populations that were different but the difference was not evident due to sampling error
What is inferential statistics?
Looking at the mean from 2+ samples and attempts to correct for sampling errors when looking for difference.
Something is inferred about a population based on the sample.
What is descriptive statistics?
Purely describe the data that is in the sample, with no manipulation/attempt to correct for error.
i.e. establish the mean, median and SD of a data set
What is a hypothesis?
A potential answer to a RQ
What is a null hypothesis?
The hypothesis that there is no significant difference between specified populations, and that any observed difference is due to a sampling or experimental error.
What is the formula for a null hypothesis?
There is no difference between group A and group B in…
What is the principle of falsification?
The principle of falsification states we try to disprove a theory, not prove it.
How do you formulate a hypothesis?
As a declaritive rephrasing of the RQ:
E.g.
Is there a difference in F0 in the voices of people who have read out loud for 2 hours, and people who have read silently?
H0: There is no difference in F0 in the voices of people who have read out loud for 2 hours and people who have read silently
Ha: There is a difference in F0 in the voices of people who have read out loud for 2 hours and people who have read silently
Why are the results of qualitative studies not generalisable?
In qualitative research, samples are typically small and hand-selected, meaning they’re not representative of populations
This means that comparisons between samples cannot be generalised to populations, meaning there’s no need for inferential statistics and hypotheses
