Research Flashcards
cross sectional studies
investigate a population at a single point in time, look for predictive relationships
limitations: can show correlations but not causation
What do you need for a cause and effect relationship?
to see the changes over time
experimental design
involves manipulating the independent variable to measure the dependent variable
positive control
have certain effects which can assess whether methodology was sound
negative control
has no effect
confounding variables
external variables that affect both independent and dependent variables
moderating variables
attenuate (reduce in force) or strengthen a given relationship
specifies conditions under which a given predictor is related to an outcome
can include: location, gender, religion, race, etc.
mediating variables
explains why two things are related
Example: “Whenever I watch tv at night, I feel tired the next morning”
The mediating variable in this case would be that the person didn’t get enough sleep because they were watching tv last night. So not getting enough sleep mediates the relationship between tv watching and being tired.
retrospective study
study that enrolls patients who have already had the disease; all cases have happened before the study
prospective study
reviewing new, incoming data
embedded field study
when researchers pose as patients
incidence
number of new cases during a specific time interval
risk ratio
risk of one disease compared to another
prevalence
how common the disease is; how many people have the disease at a specific time interval
mortality
number of deaths caused by specific disease
risk factors
anything that can increase the likelihood of having/developing a certain illness
case studies
extended interviews and reports about the experiences of a specific person
qualitiative research
uses focus groups and small groups to gather in-depth information about a specific research question
identifies patterns and themes
longitudinal studies
data gathered at multiple time points
randomized controlled trial
random assignment of treatment and control groups
causal conclusions can be drawn
non-randomized design
non-random allocations into treatment and placebo group
case-control
data gathered about individuals with conditions of interest and compared to individuals without the condition
experimenter bias
aka research bias
when the researcher unconsciously affects results, data, or a participant in an experiment due to subjective bias
counterbalancing
a technique used to deal with order effects when using a repeated measures design.
With counterbalancing, the participant sample is divided in half, with one half completing the two conditions in one order and the other half completing the conditions in the reverse order.
quantitative study
measurement focused, yields numeric data, and deductively tests a specific hypothesis (ex: the more hours you study for the MCAT will result in a higher score)
used to determine relationships between measurable variables
mixed methods study
integrates both quantitative and qualitative approaches
most comprehensive way to understand a phenomenon
ex: looking at both quantitative (height and weight) and also qualitative (types of relationships)
