Exam 2 Flashcards
When values are taken from acutal data and the relative frequencies of occurence are calculated this is reffered to as?
Empirical Distribution
How is statistical distribution defined?
- Defined by some theoretical probability distribution
Normal distribution is also called?
- A normal distibution is any such distribution with a known? 2 things
- What two distribution can be known?
What is a z-distribution
- A special type of normal distribution where the data have been transformed to follow the normal distibution with a mean of 0 and a standard deviation of 1.
- z=Xi-_X/_SD
Xi+?
X line above?
Xi= observed value
Xline= mean value
T-distribution
When is it used and why?
When does t approach Z?
- Usefule when the sample size is relatively small.
- T approaches Z when sample size increases
T and Z are useful for continuous variables but not for?
Dont allow for categorical measurements
Binomial distribution describes?
Described a random variable with two possible outcomes.
Which represents the simplest of a categorical outcome (success vs. failure)
Event happening p and event not happening q
Calcualtion for risk of outcome?
Among Exposed= A/(A+B)
Unexposed= C/(C+D)
Risk Ratio (Relative Risk)
Only done for what type of study?
What is the equation?
If =1?
If >1?
<1?
RR= Risk of Exposed/ Risk Unexposed
Fixed Cohort Study
If = 1 no difference between groups
>1 Then exposure increases risk of outcome occurence
<1 then exposure decreases risk of occurence
Rate Ratio Calculation for what type of study?
equation
Incidence rate among exposed (A/person-time in exposed group)
/
The incidence rate among unexposed (C/person-time in unexposed group)
ARR ( Risk difference or attributable risk)
Equation
Risk of exposed - risk of unexposed
RD= A/(A+B) - C/(C+D)
Attributable fraction in the exposed (ARP)
ARP= [A/(A+B)-C/(C+D)]/(A/A+B) x 100
Prevented Fraction In Exposed
?
PFE= 1-Risk Ratio
Odds of being eposed In Case-Controlled Studies
Odds of being exposed A/C for cases
For controls B/D
Odds ratio?
AxD/BxC
>1 exposure and outcome positively assoicated
<1 indicates exposure and outcome is negatively associated
When is relative risk preferred over OR?
Preferred in prospective cohort studies and RCTs only.
Information Bias
Occurs when errors in measurement