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;
108 Cards in this Set
- Front
- Back
|
Definition: Range
|
(Largest variable) - (Smallest variable)
|
|
|
Definition: Variance
|
The spread of the variables in the distribution
|
|
|
Definition: Standard Deviation
|
The square root of the variance.
|
|
|
Normal distribution: % of variables in the distribution one standard deviation above and below the mean.
|
68%
|
|
|
Normal distribution: % of variables in the distribution two standard deviations above and below the mean.
|
95%
|
|
|
Normal distribution: % of variables in the distribution three standard deviations above and below the mean.
|
99%
|
|
|
Definition: Z-score
|
The number of standard deviations above a mean where a variable lies.
|
|
|
Definition: Percentile
|
# of variables at or below that level.
|
|
|
Definition: Interquartile range
|
The range of values between the 25th and 75th percentile.
|
|
|
Definition: Left skewed distribution
|
Tail on left. Mode on right.
|
|
|
Defintion: Right skewed distribution
|
Tail on right. Mode on left.
|
|
|
Definition: T-distribution
|
Different distribution for each sample size
|
|
|
Central limit theorem
|
Given a sufficiently large sample size, a sample drawn from the population will be normally distributed regardless of the shape of the original population distribution.
|
|
|
Definitions: Standard Error of the Mean (descriptive and mathematical definitions)
|
Descriptive: Variability in a distribution of sample means
|
|
|
Which is larger: standard error the mean OR standard deviation
|
standard deviation
|
|
|
Definition: 95% confidence interval
|
the interval such that the true value has a 95% probability of lying within
|
|
|
Definition: P-value
|
Given the null hypothesis is true, the probability off getting a result as extreme or more extreme than the observed outcome by chance alone.
|
|
|
Definition: Alpha value
|
The value used to compare the p-value to to determine if it is significant.
|
|
|
P-value > Alpha value: Conclusion?
|
Results are likely due to chance alone and are not statistically significant.
|
|
|
P-value <= Alpha value
|
Results are unlikely due to chance alone and are statistically significant.
|
|
|
Definition: Null hypothesis
|
There is no difference between the groups being assessed.
|
|
|
Definition: Alternative hypothesis
|
There is a difference bbetween the two groups being assessed.
|
|
|
Definition: Two-tailed alternative hypothesis
|
The two groups' values are not equal. (no more specific than that)
|
|
|
Definition: One-tailed alternative hypothesis
|
One groups' values is less than those of the other group
|
|
|
Definition: Type 1 (Alpha) Error
|
Incorrectly rejecting the null hypothesis ( a false positive conclusion)
|
|
|
Definition: Type 2 (Beta) Error
|
Failure to reject the null hypothesis when the alternative hypothesis is correct (a false negative conclusion)
|
|
|
Definition: Power
|
The probability of finding a specified difference, or larger, when a true difference exists
|
|
|
The difference between statistical significance and clinical significance
|
Statistical significance only comments on chance, not benefit to patient.
|
|
|
Generally, minimum power for a clinically relevant difference
|
80% or more
|
|
|
Problem with performing multiple tests without adjusting error
|
Increases probability of a Type 1 error.
|
|
|
Bonferroni correction
|
Divide alpha by number of statistical tests to perform to yield a new alpha.
|
|
|
The name for the type of correction where alpha is divided by the number of statistical tests to perform.
|
Bonferroni Correction
|
|
|
Difference between parametric and non-parametric tests
|
Parametric: performed on data that are normally distributed
|
|
|
Chi-square test
|
Performed on discrete data
|
|
|
Fisher's exact test
|
Performed on discrete data
|
|
|
The other name for the Student t-test
|
two sample t-test
|
|
|
The other name for the two sample t-test
|
Student t-test
|
|
|
Student t-test/Two sample t-test
|
Performed on normally distributed data
|
|
|
One sample t-test
|
Performed on normally distributed data
|
|
|
What does ANOVA stand for?
|
ANalysis
|
|
|
ANOVA
|
Performed on normally distributed data
|
|
|
Paired t-test
|
Performed on normally distributed data
|
|
|
Correlation
|
Performed on normally distributed data
|
|
|
Pearson's Correlation coefficient
|
aka rho
|
|
|
Rho
|
aka Pearson's Correlation coefficient
|
|
|
Spearman Rank Correlation
|
Performed on non-normally distributed data
|
|
|
R-squared
|
aka Coefficient of determination
|
|
|
Coefficient of Determination
|
aka R-squared
|
|
|
Multiple regression analysis
|
Has two or more independent variables
|
|
|
Logistic regression analysis
|
Performed when the dependant variable is binary
|
|
|
Primary screening
|
Peformed to prevent a disease
|
|
|
Secondary screening
|
Peformed to reduce the impact of a disease
|
|
|
Sensitivity
|
Given disease is present, the probability of testing positive
|
|
|
Specificity
|
Given disease is absent, the probability of testing negative
|
|
|
Predictive value positive
|
Given the test is positive, the probability that the disease is present
|
|
|
Predictive value negative
|
Given the test is negative, the probability disease is absent
|
|
|
Types of descriptive studies
|
Case Report
|
|
|
Difference between a case report and a case series
|
One versus multiple interesting observations
|
|
|
Correlation study
|
Descriptive study done on large populations (the per capita meat consumption of a town and the prevalance of CAD there)
|
|
|
Prevalance study
|
aka Cross-Sectional study
|
|
|
Stratified Random Sample
|
Sample is first stratified into groups, and then subjects are tne randomly drawn from each group.
|
|
|
Types of analytic studies and how their subjects are selected
|
Case Control (Outcome, then exposure is assessed)
|
|
|
Randomized Controlled Trial: Pros
|
gold-standard
|
|
|
Randomized Controlled Trial: Cons
|
May take time = $ and loss-to-follow-up of subjects
|
|
|
Prospective Cohort Study: Pros
|
generates incidence data
|
|
|
Prospective Cohort Study: Cons
|
bad for rare outcomes
|
|
|
Retrospective Cohort Study: Pros
|
easy to do, so less $
|
|
|
Retrospective Cohort Study: Cons
|
bad for rare outcomes
|
|
|
Case Control: Pros
|
good for rare disease
|
|
|
Case Control: Cons
|
subject to several biases as recall bias, interview bias, selection bias
|
|
|
Meta-analysis: Pros
|
increases power by combining study results
|
|
|
Meta-analysis: Cons
|
Studies not done exactly the same way
|
|
|
Difference between intention-to-treat analysis and efficacy analysis, and which is prefered by the FDA
|
Intention to treat: All subjects analyzed by study arm, regardless of compliance
|
|
|
Phase One Study
|
Safety
|
|
|
Phase Two Study
|
Efficacy
|
|
|
Phase Three Study
|
Compares new treatment to standard therapy
|
|
|
Prevalance
|
The proportion of subjects in a group with a certain disease, includes new and old cases (a snapshot in time)
|
|
|
Incidence
|
New cases occurring over a defined period of time
|
|
|
Cumulative Incidence
|
aka Attack Rate
|
|
|
Attack Rate
|
aka Cumulative Incidence
|
|
|
Incidence Rate
|
aka Density
|
|
|
Density
|
aka Incidence Rate
|
|
|
Relationship of Prevalence to Incidence Rate
|
Prevalence = (Incidence Rate)(Average Duration of Disease)
|
|
|
Absolute Risk
|
The probability of going from a healthy state to an ill state (eg the probability that person x will develop condition y over the next z years when they don't have it already)
|
|
|
Relative Risk
|
The strength of association between an exposure and outcome.
|
|
|
(xy): x: exposure, y: disease
|
0=negative, 1=positive
|
|
|
Relative Risk (calculated from prevalances)
|
Relative Risk = Prevalence of disease in population A divided by Prevalence of disease in population B
|
|
|
Odds Ratio
|
An approximation of relative risk used in case control studies.
|
|
|
(xy): x: exposure, y: disease
|
0=negative, 1=positive
|
|
|
Attributable Risk
|
Excess disease in the exposed population that can be attributed to the exposure
|
|
|
(xy): x: exposure, y: disease
|
0=negative, 1=positive
|
|
|
Confounding
|
An erroneous study conclusion when a factor is associated with an exposure and is itself an independent risk factor for the outcome
|
|
|
Ways to control for confounding
|
Restriction (subjects with known risk factors are excluded)
|
|
|
Effect Modification
|
aka Interaction
|
|
|
Surveillance Bias
|
One cohort is followed more closely than another and is thus more likely to be diagnosed with the disease, leading to a difference in risk estimation between the two cohorts
|
|
|
Interaction
|
aka Effect Modification
|
|
|
Herd immunity
|
Disease protection of an unimmunized individual because the population is immunized
|
|
|
Epidemic curve
|
Plots an epidemic by:
|
|
|
Kaplan Meier survival cuves
|
Shows survival of different groups over time.
|
|
|
Bias
|
A systematic error with a study leading to an erroneous estimation of the association between the exposure and the outcome
|
|
|
4 major types of bias
|
-Recall bias (differential recall of exposure status by individuals based on their health status)
|
|
|
Criteria for cause-effect relationships
|
-Strength of association (how large or small is the relative risk/odds ratio
|
|
|
Definition: Epidemic
|
Greater than expected disease frequency in a defined population
|
|
|
Definition: Pandemic
|
Greater than expected disease frequency in a large defined area
|
|
|
Definition: Endemic
|
Constant presence of a disease in a defined population
|
|
|
Case Fatality Rate
|
Proportion of patients with a disease who die of that disease
|
|
|
Mortality Rate
|
Proportion of patients who die of a disease in a defined population
|
|
|
Key difference between Case Fatality rate and Mortality Rate
|
The denominator:
|
