Lecture 8 and 9 (Correlation, Regression, CIs)

Question 1

The dependent variable is on what axis?

Answer 1

Y-axis

Question 2

The independent variable is on what axis?

Answer 2

X-axis

Question 3

When should you use a correlation analysis?

Answer 3

• examine relationship between variables
• estimate strength of association between variables
• when independent and dependent variables are not clearly different
• when regression requirements not met

Question 4

A correlation coefficient of 0 means:

Answer 4

• there is no association between the two variables

Question 5

A regression is:

Answer 5

• how well data fits a line
• r-value close to 0 = no correlation
• r-value closer to 1 or -1 = high correlation
• r-squared tells you the amount of variation in Y that is contributed by variation in X.

Question 6

When should you use regression analysis?

Answer 6

• look for a trend in data between variables
• more than one X (independent) variable = multiple regression
• predict a dependent variable
• adjust for confounding variables
• curve fitting (pharmacokinetics)
• calibration and laboratory assays
• detect patterns in microarray data

Question 7

Regression r-value close to 0:

Answer 7

no association

Question 8

Regression r-value close to 1:

Answer 8

strong association

Question 9

Regression r-squared value tells you:

Answer 9

• the amount of variation in Y that is contributed by variation in X.

Question 10

Parametric test characteristics:

Answer 10

• assume variables are normally distributed with equal variances
• dependent on mean and variance
• susceptible to outliers
• requires continuous variables

Question 11

Non-parametric test characteristics:

Answer 11

• based on ranks
• distribution, variance, mean does not matter

Question 12

You can transform non-linear data to linear data by:

Answer 12

• taking logs

Question 13

Three ways you can control for outliers:

Answer 13

1. using non-parametric test
2. dropping the outlier(s)
3. log transformation

Question 14

Multivariate regression:

Answer 14

• more than one X (independent) variable
• allows adjustment for confounders
• controls for variable interactions by multiplying variables together

Question 15

Stepwise regression:

Answer 15

• finds the top contributing variable, then the second, then the third, etc. until a point of diminishing returns is reached.
  
  • a.k.a finds the group of variables that has the largest collective r-squared value.

Question 16

Multiple logistic regression:

Answer 16

• a multivariate analysis
• adjusts for confounding
• useful when outcome is dichotomous
• provides a direct estimate of the ODDS RATIO for each independent variable

Question 17

When the distribution of your data is not normal, what type of test should you use?

Answer 17

non-parametric

Question 18

If you are analyzing more than one type of independent (X) variable, what type of analysis should you use?

Answer 18

multivariate regression

Question 19

Principal Component Analysis (PCA):

Answer 19

• takes many variables and reduces them by regression
• gives you groups of variables that best explain variation

Question 20

Risk factors are modfiable through:

Answer 20

primary prevention

Question 21

Prognostic factors are modifiable through:

Answer 21

secondary prevention

Question 22

Common prognosis endpoints:

Answer 22

1. case fatality (patients with disease who die of it)
2. disease-specific mortality (people per 10,000 who are dying of specific disease)
3. response
4. remission
5. recurrence

Question 23

Equipoise:

Answer 23

• a genuine lack of consensus in the medical community about a treatment or prognosis, and how to treat.
• allows for RCTs

Question 24

Kaplan-Meier Analysis:

Answer 24

• most widely used survival analysis:
• a graph of time to event
• every horizontal segment is a time period
• every vertical drop is an event (death) or a dropout
• larger the sample size, smoother the curve

Question 25

Kaplan-Meier analysis truncation:

Answer 25

when a patient enters the study after it has already started

Question 26

Kaplan-Meier analysis censoring:

Answer 26

when a patient drops out of a study after it has started

Question 27

Can a Kaplan-Meier analysis handle covariates?

Answer 27

No.

• use a Cox regression for this

Question 28

Cox regression:

Answer 28

• multivariate survival analysis
• can control for other factors
• calculates hazard ratio (same as relative risk)

Question 29

Equipoise allows for:

Answer 29

• randomized control trials to occur
• equipoise = uncertainty in the medical community

Question 30

Variance =

Answer 30

measure of the spread/dispersion of values around the mean.

Question 31

Standard deviation =

Answer 31

√v; (v = variance)

• decreases as sample size increases

Question 32

Standard error of the mean (SEM) =

Answer 32

SD/ √n

Question 33

Central limit theorem posits:

Answer 33

• larger the sample size, the closer the study mean is to the population mean
  
  • i.e. narrower confidence interval

Question 34

Interquartile range:

Answer 34

• IQR contains 50% of the observations
  
  • (from the 25th - 75th percentile)

Question 35

Confidence intervals describe:

Answer 35

• the uncertainty that surrounds a particular observation
• larger the sample size, narrower the CI = MORE PRECISE STUDY

Question 36

Equation for 95% CI:

Answer 36

95% CI = mean +/- 1.96(SD/ √n)

• SD = standard deviation
• n = sample size

Question 37

For correlation analyses, the confidence interval cannot contain:

Answer 37

0

0 = no correlation

Question 38

For relative risk, hazard ratios, and odds ratios, the confidence interval cannot contain:

Answer 38

1