module 1 - RAMBOMAN Flashcards
What is random error?
An error that occurs by chance
What is a non-random error?
Errors due to poor study design, poor study processes or poor study measurement
What is a valid study?
A study with only a small amount of random or non-random error
Cohort study aka ____
Longitudinal study
What is a cohort study? Give an example
People are asked about their exposure (e.g. alcohol consumption) at the beginning of the study and they are then followed up over time (e.g. to find out their death rate)
Example: Swedish conscripts study
What is a cross-sectional study?
The exposure (e.g. male or female?) & outcomes (e.g. whether they have asthma) are measured at approximately the same time. Example: Asthma prevalence study in pop health 111 students
How does allocation (= the ‘A’ in RAMBOMAN) occur in a cohort study?
By measurement
What is a confounding error?
NOTE: confounding is an important concept in epidemiology
When the exposure (e.g. high alcohol use) is mixed with another (e.g. high solvent use) factor that also affects the outcome (e.g. death rate).
Confounding error/bias aka ______.
Why is this?
Allocation adjustment error, b/c this where the error occurs -> while moving (allocating) people from the pop. into their groups (EG or CG).
i.e. while moving from the triangle to the circle
How can we deal with confounding in a study?
Divide (stratify) the study into sub-studies (strata) based on the confounder factor (e.g. solvent use)
Example: separate the solvent users & non-users. Then look at heavy and moderate alcohol consumption in each group separately. Look at results - if inc. in death rate in each strata is similar then we put the groups back together (as before).
What is the benefit of using a randomised control trial? What is a disadvantage?
Adv. - It results in the people in EG & CG being very similar (at the beginning of a study) and so this is the best way to reduce confounding.
Disadv. - it is not always possible to do a RCT, e.g. in alcohol studies -> can’t tell to drink heavily or moderately.
How are people allocated to EG & CG in a randomised control trial?
randomly - i.e. “by flipping a coin”
What does the M in RAMBOMAN stand for? Describe what it means.
MAINTENANCE - were participants maintained in EG & CG during the study?
i. e.
- did people remain in the groups they were initially allocated to?
- were participants lost to follow-up (i.e. couldn’t couldn’t count their events) - were more lost from EG or CG?
