Unit 4 Flashcards
what are observational studies
seek to quantify relationships bw 2 or more variables
Data is collected as they naturally exist
NO manipulation of variables
They measure how variables are associated with one another
2 types of observational studies
exploratory & descriptive
what are exploratory studies
link risk factors to conditions
what is correlation
describe relationship bw 2 variables; doesn’t imply causation
How change in 1 variable is or isn’t associated with change in another
how is correlation described
direction & strength of the relationship
describe direction of correlation
positive (as one goes in one direction the other goes in the same), negative (as one goes in one direction the other goes in the opposite direction) or no correlation (change in one variable doesn’t influence change in another)
+ meaning
means as one variable increases/decreases the other does the same
- meaning
means as one variable increases the other decreases
describe strength of correlation
magnitude of change (strong positive correlation - as one variable changes in one direction the other changes in the same direction & magnitude
Higher (absolute) number means
greater degree of association
Closer to 1 or -1
stronger relationship
Closer to 0
weaker relationshi
identify potential risk factors for condition from data at 1 point in time
cross sectional
identify potential risk factors for condition using retrospective approach
case control
identify potential risk factors for condition using prospective approach
cohort
2 differences bw exploratory designs
how data is collected and compared bw groups
control and/or manipulate variables to identify cause and effect relationships
experimental
collect data as they naturally occur & identify relationships bw 2 or more variables
observational design
2 types of observational design
exploratory & descriptive
4 exploratory designs
correlation, cross sectional, case control., cohort
Exploratory designs are also classified as:
Observational studies
When describing a correlation, one should indicate:
directio & strength of relationship
When choosing a control participant for your case-control study, you should consider:
How long you will track the participants
How you intend to prospectively study the participants
How many control participants to include per case participant
Which characteristics to match the control to the case
How many control participants to include per case participant
Which characteristics to match the control to the case
Snapshot in time
cross sectional design
Prospective approach
cohort design
Retrospective approach
case control design
identify a relationship between two variables
correlation
condition already happened & data represents 1 moment in time (snapshot)
cross sectional
