Study designs Flashcards
Name the (8) basic steps in research.
- Identify the research problem
- Review literature
- Form the research question, objectives and hypotheses.
- Choose the study design
- Decide the sample design
- Collect the data from the research sample
- Process and analyse the collected data
- Write the research report, develop a research proposal, disseminate and utilize the results.
What are the 5 components of the methodology cycle for empirical research? Also describe what the 5 components mean.
- Observation → proposing the hypothesis and gathering empirical data through observation.
- Induction → inductive reasoning to form a general conclusion based on the data that is gathered by observation.
- Deduction → deducting a conclusion that is based on logic and rationality (so that the conclusion is specific and unbiased).
- Testing → returning to empirical methods to test hypothesis (e.g. with the use of statistical methods).
- Evaluation → putting forth the data, including a conclusion and arguments, stating limitations and tips for follow-up study.
What is the difference between quantitative and qualitative empirical research?
- Quantitative empirical research → used to gather information using numerical data (can be different types of study designs)
- Qualitative empirical research → used to gather information via non-numerical data such often semi- or unstructured methods (e.g. via surveys, interviews, etc.)
In regard to Randomized Controlled Trials (RCTs), what is:
- Simple randomisation
- Restricted randomisation
- Simple randomisation → randomisation based on a single sequence of random assignments (e.g. flipping a coin)
- Restricted randomisation → randomisation where balance between study groups is achieved in size or baseline characteristics.
For restricted randomisation, participants can be randomised via block randomisation or via stratified randomisation. For what are the two types of restricted randomisation used?
- Block randomisation → used to ensure that the groups will be of approximately the same size
- Stratified randomisation → used to ensure that equal numbers of participants with one or more baseline characteristics are balanced within each stratum (e.g. age).
What is a placebo controlled trial?
A trial in which there are two (or more) groups. One groups gets the active treatment, the other gets the placebo. Everything else is held the same between the two groups, so that any difference in their outcome can be attributed to the active treatment.
Describe what blinding means during a clinical trial.
It is a procedure where one or more parties in a trial are kept unaware of which treatment participants have received. The parties that are kept ‘blind’ can be: patients, investigators, observer, analyst or a combination of one or more of the above.
Regarding experimental designs, what is a parallel design?
Here, two or more treatments are compared and the participants are randomised to the treatments that are being compared. A parallel design is seen as the golden standard for Phase 3 medicine trials.
Parallel designs can be:
- with the use of a control group
- Standard-of-care vs new treatment
- Treatment vs sham
- Two active treatment groups
- Treatment vs no-intervention
Regarding experimental designs, what is a crossover trial?
A crossover trial has a repeated measures design in which each patient is assigned to a sequence of two or more treatments (each patient receives these treatments in a different order), of which one may be a standard treatment or a placebo.
Name advantages and disadvantages of a experimental crossover trial.
- Advantages → higher statistical power
- Disadvantages → it is needed to take into account the washout period (the length of time that someone is enrolled in a trial and must not receive any treatment before receiving the trial’s experimental therapy) and the carry-over effect (the treatment in the first period has an effect on the treatment in the following second period)
The analysis of a crossover trial is complex. What can be tested?
- Period effect → implies that the effect of the same treatment that is received at two different periods is different for each period.
- Treatment effect → the causal effect of a given treatment or intervention on an outcome variable of interest.
Note: you can test for the period effect → if there is no difference in period effect → test for treatment effect → if there is a difference in treatment effect, only period 1 is usable.
Regarding experimental design, what is a cluster trial?
It is a type of a randomised controlled trial, where groups of subjects are randomised (groups can be wards, hospitals, provinces, educational levels). With this type of trial, a novel therapy is evaluated in a particular group.
What is an advantage of a cluster trial?
It helps to prevent contamination (=aspects of the intervention are applied to the non-intervention group).
Regarding experimental designs, what is a factorial trial?
In a factorial trial, two (or more) intervention comparisons are carried out simultaneously. E.g. participants may be randomised to receive aspirin or a placebo, and are also randomised to receive a behavioural intervention or standard care.
There are 3 main types of factorial trial designs:
- Within subject factorial design
- Between subject factorial design
- Mixed factorial design
Describe their characteristics.
- Within subject factorial design → all of the independent variables (factors) are manipulated within subjects (e.g. a participant is tested both while using a cell phone and while not using a cell phone and both during day and during night)
- Between subject factorial design → the subjects are assigned to different conditions and each subject only experiences one of the experimental conditions.
- Mixed factorial design → at least one within subject variable and one between subject variable.
What are advantages of RCTs?
What are disadvantages of RCTs?
What are the disadvatages of per-protocol analysis?
Overestimation, misinterpretation and inaccurate (biased) assessment of the effectiveness of the intervention.
Describe the characteristics of an observational case control study.
- Researchers identify people with an existing health problem (i.e. ‘cases’) and a similar group without the problem (‘controls’) and then compare them with respect to the exposure(s).
- The exposure is in the past, then comes the disease.
Describe the characteristics of an observational cross-sectional study.
- These studies look at a population at a single point in time, like taking a slice or cross-section of a group, and variables are recorded for each participant.
What does the hierarchical evidence pyramid contain?
- At the top of the pyramid is filtered evidence including systematic reviews, meta-analyses, and critical appraisals. These studies evaluate and synthesize the literature. The top of the pyramid represent the strongest evidence (→ systematic review).
- At the base of the pyramid is unfiltered evidence including RCTs, cohort studies and case-controls (i.e. primary literature).
Name every type of study that is included in the hierarchical evidence pyramid from weakest to strongest evidence.
- Background information or expert opinion.
- Case-control studies
- Cohort studies
- RCTs
- Critical-appraised individual articles (articles that are selected and rated for clinical relevance by physicians)
- Critical-appraised topics (short summary of the best available evidence on a focused question)
- Systematic reviews
What is a reporting guideline?
It is a simple, structured tool for health researchers to use while writing manuscripts. A reporting guideline provides a minimum list of information.
For observational studies, the reporting guideline that is used is the STROBE reporting guideline. Think of how an article looks like according to a reporting guideline.
For RCT studies, the reporting guideline that is used is the CONSORT reporting guideline. What (among others) is included in the article according to the CONSORT reporting guideline?
A flowchart.
