Experimental Design Flashcards
Type of variables
Independent and dependent variable
Which variable does the experimenter manipulate and why?
Independent to observe how the dependent variable responds
Where do the variables go on the axes of a graph?
X= independent Y= dependent
Confounding variables
third variable that affects both the independent and dependent variable. can obscure results.
Give an example of a confounding variable
Coffee consumption and cardiovascular disease. Cigarette smokers have a direct correlation with cardiovascular disease and it just so happened people who smoke drank coffee more than people who didn’t.
the confounding variable: cigarette smoking
Mediating variable
Provides mechanistic link between observed relationship between two different variables. can be seen between independent and dependent variable.
Give an Example of a mediating variable
A strong relationship was found that families with lower level income tend to die from cancer more. Cancer does not check a families financial situation before infecting them. so researchers found that poverty is associated with lack of good health care.
Moderating variable
a variable that modulates intensity of a certain relationship
Give an example of a moderating variable
A stressful work place can lead to anxiety/depression. Exercise and mediation may reduce strength of that relationship.
moderating variable: coping strategies
Operationalization
The process of defining variables in measurable, practical ways
Types of studies
Experimental studies and observational studies
Experimental studies
Researchers manipulate the world in some way
observational studies
researchers analyze pre-existing patterns of variation of to determine possible relationships
Quantitative studies
Numerical meausres
Qualitative studies
Verbal or open-ended measurements of variables
Experimental controls
Negative: do not receive treatment or get a placebo
Give an example of an experimental design with a negative control
A researcher comes up with a drug that can cure the flu. Each patient that has flu takes drug and gets better. Issue is that patients over time will get better anyway. So some patients cannot take drug and we can observe how fast they get better in comparison to those that take the drug.
Positive controls
Receive treatment known to influence outcome of interest
What does the positive control do?
Confirms the adequacy and competency of the experimental procedure. They provide a way to compare the effectiveness of the treatment being studied.
Randomization
samples are randomly given to control or treatment groups
Blinding
Researchers and/or participants do not know which sample is control/experimental when doing maintenance
Correlational studies
Measuring various stats among a set of people, looking for correlations
Cross-sectional studies
Data taken from a single point in time
EX) opinion poll
What are some cross sectional studies limitations?
Cannot provide information about causality. Provides no information about how things change over time.
Longitudinal studies
Multiple measurements are made over time
Cohort studies
a group of subjects is assembled based on an organizing principle and followed up on over time
Prospective stuides
Data gathered moving forward
Retrospective studies
Data gathered from the past
Case-control studies
Compare “cases” to “control”/ people who do not have outcome of interest
Meta-analyzses
Data from multiple studies are combined and reanalyzed
Systematic reviews
Critically assess the outcome of various studies
The 7 principles for research ethic written by NIH
1) must have social and clinical value
2) must have scientific validity
3) fair subject selection
* 4) risk-benefit ratio
5) independent review
* 6) informed consent
7) respect for potential and enrolled participants
Nature vs. Nurture
nature= genetic factors nurture= environmental factors
Genetic screening
Used to try to identify genetic predictors of various health outcomes
Causation
Unidirectional where changes in one variable result in changes in the other
Correlation
Can have a bidirectional or reciprocal relationship
