Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
39 Cards in this Set
- Front
- Back
|
What are common statistical tests?
|
T Test
Anova Chi Square Linear Correlation Chi Square Test Meta Analysis |
|
|
T Test
Defined |
Tests difference between the MEANS of two samples
|
|
|
T Test
What are two types? |
Nonpaired/Independent:
mean of variable x in Group A and Group B sampled at time 2 Paired/Dependent: mean variable of x in Group A sampled a time 1 and time 2 |
|
|
ANOVA (analysis of variance)
Defined |
Tests difference between the MEANS of three or more samples
|
|
|
ANOVA
What are two types? |
One way analysis:
Compares MEANS of one variable (eg: group A: groupB: groupC) Two way analysis: two variables (group A,B,C and male v. female) |
|
|
Linear Correlation
Defined |
Tests relationships between two variables
correlation efficient (r) between -1 and1 |
|
|
Chi Square Test
Defined |
Tests difference among PERCENTAGES/FREEQUENCIES in a sample NOT MEAN
eg: What are differences in subjects below x level? |
|
|
Meta-Analysis
|
STATISTICAL procedure that integrates results of several independent studies considered combinable
|
|
|
What is p value?
|
probability or chance of committing a Type I error
|
|
|
How is p value evaluated?
|
a p value equal to or less than .05 is significant
(Null hypothesis must be rejected) |
|
|
Type I error
|
Null hypothesis rejected though its true
|
|
|
Type II error
|
Null hypothesis accepted though its false
|
|
|
What is Null hypothesis?
|
The asssumption that there is no difference between to variables
|
|
|
What are mean, median and modes?
|
mean-average
median- middle number when variables ordered sequentially mode- value that appears most often |
|
|
What are positive and negative skew?
|
Positive skew- modal peak to the left of the mean
Negative skew- modal peak to the right of the mean |
|
|
How to calculate clinical probability that PT A and B will experience an event?
|
Prob sickness A x Prob sickness B
|
|
|
How to calculate clinical probability that at least one PT of either A and B will experience an event?
|
(Prob sickness A+Prob sickness B) - (Prob sickness A * Prob sickness B)
|
|
|
How to calculate clinical probablity that Neither PT will experience an event?
|
Prob wellness A * Prob of Wellness B
|
|
|
What is PPV?
|
Probability that person with positive test result is truly sick
|
|
|
What is NPV
|
Probability that person with negative test is truly well.
|
|
|
What does Reliability indicate?
|
REPRODUCIBILITY of Test
|
|
|
What is infant mortality?
|
Number of deaths occurring to until first birth date DIVIDED BY
Number of Lives Birth |
|
|
What is proficiency bias?
|
Intervention studied delivered with great skill
(eg: coronary surgery given at Mayo) |
|
|
How many more samples needed in a study to decrease the confidence interval by 50% ?
|
Sample size must be increased FOUR FOLD
|
|
|
What does sensitivity and specificity?
|
Sensitivity- How well a test identifies truly sick people
Specificity- How well a test identifies truly well people |
|
|
What is crossover study?
|
Group A gets drug Group B gets placebo then
Group B gets drug and Group A gets placebo |
|
|
Define selection bias
|
subjects or investigators choose the placebo or group participants rather than them being assigned randomly
|
|
|
Sampling bias defined
|
study subjects different than normal population
(eg: choosing college students or only nursing home residents) |
|
|
What can cohort studies measure?
|
absolute risk
relative risk attributable risk |
|
|
What can case control studies measure?
|
odds ratio
|
|
|
What is number needed to harm?
|
NNH is number of persons needed to be exposed to risk for one person to be harmed.
|
|
|
What is Number Needed to Harm (NNH) Formula?
|
1/attributable risk
|
|
|
What is Number Needed to Treat (NNT)?
|
Number of persons needed to take a treatment for one person to benefit.
NNT allows comparison of effectiveness of different treatments. |
|
|
What is Number Needed to Treat Formula?
|
1/absolute risk reduction
|
|
|
Formula for Relative Risk
|
INCIDENCE of Disease in Exposed / INCIDENCE OF DISEASE IN UNEXPOSED
|
|
|
Formula for Absolute Risk
|
INCIDENCE of DISEASE
|
|
|
What is absolute risk reduction and what is formula?
|
Difference in incidence rates for exposed and unexposed (EIR-UIR)
|
|
|
Late look bias
|
In severe disease many dead before study begins
|
|
|
Lead time bias
|
early detection confused with increased survival
|
