Binary Logistic Regression Flashcards

1
Q

Continuous Variable for Outcome use what test?

A

Linear Regression

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Discrete or Binary Variable for Outcome use what test?

A

Logistic Regression

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What is the Purpose of Binary Logistic Regression?

A
  1. Predict the outcome of a BINART dependent outcome variable based on one or more independent predictor variables, which may be discrete or continuous
  2. Interpret significant odds ratios
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Researchers can transform continuous outcome variables into what?

A

BINARY variable if assumptions of linear regression are not met

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Residual Homoscedasticity

A

What you WANT for linear regression

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Heteroscedastic

A

VARIANCE is NOT equal at all points = MUST do LOGISTIC regression

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Logit is what?

A

LOG-ODDS
Natural log of the odds of success

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Maximum Likelihood Estimation MLE

A

Used to maximize the predictive capabilities of the regression

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Overall M

A

Model of Significance

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Overall Model Significance (step 1)

A

No R^2 or Adjusted R^2 to interpret the proportion of total variation
USE a Chi-Squaree goodness of fit statistic for the overall model

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Significance of Predictor Variables (step 2)

A

Hypothesis tests are conducted using a WALD-CHI-SQUARE TEST

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Since logistic regression is a non-linear model, the coefficient estimate B is transformed into ODDS RATIO

A

OR or expB or adjusted OR
Adjusted OR = allows you to look at one predictor value by controlling for the others

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Definition of Odds Ratio OR

A

The odds that an outcome Y=1 will occur given a particular exposure X, compared to the odds of the outcome occurring in the absence of that exposure

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

OR Meanings

A

OR = 1 = X does not affect odds of Y
OR >1 = X is associated with higher odds of Y
OR <1 = X is associated with lower odds of Y

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

For OR the 95% CI is significant when it what?

A

Does NOT include 1

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

For Linear Regression the 95% CI is significant when it what?

A

Does NOT include 0

17
Q

BINARY LOGISITIC Contrast

A
  1. Dichotomous outcome variable
  2. MLE to maximize the predictive capabilities of the regression
  3. X^2 statistic for overall model
  4. WALD test for significance
18
Q

Simple/Multiple LINEAR Regression Contrast

A
  1. Continuous outcome variable
  2. OLS estimation to minimize the residuals that relate each data point to the trend line
  3. F-test for overall model
  4. T-test for significance