Lecture 2 - Research Design Flashcards
What is a research aim?
• A statement describing the purpose of the study
• A short, concise “catchy” re-statement of the research topic or theme
What is a research question?
• What the study aims to answer after data analysis and interpretation
• Must be specific and concise to provide a clear foundation for the hypotheses
What are the three types of research questions?
• Descriptive
• Comparative
• [UNFINISHED]
Describe and give an example of a descriptive research question
• Present variables to measure or analyse
• “What is the proportion of elite triathletes who take vitamin supplements?”
Describe and give an example of comparative research questions
• Clarifies the difference in an outcome variable between two or more groups
• Compares the effects of variables or interventions
• “Is there a difference between males in females in reaction time?”
• “Is quadriceps power improved more by resistance training or sprint training?”
Describe and give an example of [?] research questions
• Defines trends, associations or interactions between variables
• “Is there a relationship between number of exams and perceived stress levels among undergraduate students?”
What are the two acronyms for criteria for good research questions?
• FINER
• PICOT
What does the FINER acronym stand for?
• Feasible - appropriate number of participants and technical expertise, affordable in time and money, and manageable in scope
• Interesting - getting the answer intrigues the investigator, peers and the community
• Novel - confirms, refutes or extends previous findings (or discovers something new)
• Ethical - consistent with the standards that will lead to ethical approval
• Relevant - to scientific knowledge and/or applied practice
What does the PICOT acronym stand for?
• Population - what specific group are you interested in?
• Intervention - what is the intervention you are investigating (if it’s an intervention study)
• Comparison - what is the alternative to compare with the intervention?
• Outcome - what do you intend to accomplish, measure, improve or affect?
• Time - what is the appropriate follow up time to assess the outcome?
What is a research objective?
• Objectives define specific aims and explain how the overall aim is achieved
• Translates the overall aim into operational statements, tells the reader how the overall aim will be achieved, and may even specify how a variable is measured if this is important
What is a research hypothesis?
A specific, clear and testable statement about the possible outcome of a study
What are the two traditional hypotheses?
• Null hypothesis
• Alternate hypothesis
What are the six types of research hypotheses?
• Simple
• Complex
• Directional
• Non-directional
• Associative
• Causal
What is a simple hypothesis?
It predicts the relationship between a single dependent variable and a single independent variable
What is a complex hypothesis?
Predicts the relationship between multiple dependent and independent variables
What is a directional hypothesis?
Identifies the direction of the relationship between variables
What is a non-directional hypothesis?
Does not specify a direction of effect (intentionally)
What is an associative hypothesis?
Change in one variable is associated with change in another (but may not be causal)
What is a causal hypothesis?
Manipulating the independent variable will change the dependent variable
Why is research design important?
• To provide a framework of methods and techniques
• Enables research to be conducted in a structured way
• Research design tells us “how” a study is conducted
What is the difference between an observational study and an experimental study?
• Observational studies have no intervention whereas experimental studies involve an intervention
• Observational studies provide information on associations between an ‘exposure’ and an ‘event’ or ‘characteristics’ of the population
• Experimental studies allow for the determination of cause and effect if the study is designed appropriately
What are the differences between cross-sectional and longitudinal studies?
• Cross-sectional studies assess a phenomenon at one point in time, whereas longitudinal studies assess a phenomenon at several points in time
• CS measure different samples/populations whereas longitudinal ensure the same sample/population over time
• CS provides a snapshot of a given point in time, whereas longitudinal provides information on change at an individual level
• CS examines trends and changes at a societal/national/international level whereas longitudinal examines changes in the same people over time
What is retrospective design?
Backwards looking - examine data that already exists
What is prospective design?
Forward looking - collects new data, then sometimes watch/wait
What is a case control study?
• Usually retrospective (but not exclusively)
• “Cases” have the outcome (e.g. heart disease)
• “Controls” do not have the outcome
What is a randomised controlled trial?
• Forward looking (prospective)
• Participants are randomised into a “control” or “intervention” group
• Follows group over time to determine a difference in outcomes
What are the three criteria for randomised controlled trials to prove cause and effect?
• The cause must precede the effect
• The cause and effect must be correlated
• The correlation between cause and effect cannot be explained by another variable
What are the threats to internal validity in a randomised controlled trial?
• History - an unintended event during the study has an influence
• Maturation - processes within participants that happen due to time
• Testing - learning effect of multiple testing (e.g. scoring higher in cognitive function test second time)
• Instrumentation - changes in instrument calibration, staff performing procedures differently
• Statistical regression - non-random selection (e.g. self report or measurements) lead to values that are less extreme
• Selection bias - non-random selection or a specific condition biases the study and may explain the results
• Experimental mortality - loss of participants
• Selection-maturation interaction - if key variables interact with time and if one group is fated more than another
• Expectancy - experimenters expect one group to perform better
What are the threats to external validity in a randomised controlled trial?
• Reactive or interactive effects of testing - any pre testing makes the intervention more effective
• Interaction of selection bias and treatment/intervention - when a group is selected for a particular characteristic and the treatment only works with these participants
• Reactive effects of experimental treatment - the treatment/intervention only works under controlled laboratory conditions and may not be effective in the real world
• Multiple treatment interference - when participants receive more than one treatment/intervention and the effects of the other treatment affects the results
What is internal validity?
Whether the study design, conduct and analysis answer the research question without bias
What is external validity?
Whether the study findings can be generalised to other contexts
What can you do to control threats to internal validity?
• Randomisation
• Placebo controls
• Blinding
What can you do to control threats to external validity?
• Randomisation
• Selecting tests, interventions and participants that are representative and generalisable
What is a quasi-experimental design?
An intervention study that does not randomise participants and/or have a control group
What are feasibility studies?
• Research done before a main study to determine whether the methods are feasible and acceptable
What are pilot studies?
A small-scale study conducted prior to a large-scale experiment to test and refine procedures
What is epidemiology?
The study of the occurrence and distribution of health-related events, states, and processes in specified populations, including the study of the determinants influencing such processes, and the application of this knowledge to control relevant health problems
What are the ways to assign people to groups?
• Random assignment - randomly allocate participants to different study groups
• Random matched assignment - considering a participant characteristic
• Intact groups - not randomising (if it’s not possible - there are post-hoc justifications)