Biostats 5- RCT

Question 1

What is the test that compares the differences btwn mean of 2 groups?

Answer 1

T test

Question 2

When do u use independent t tests?

Answer 2

When there are unpaired measurements

Question 3

When do u use a paired t test?

Answer 3

Paired values

Question 4

What are the 3 assumpotions in the t test?

Answer 4

independent variable with 2 gorups
independence
1 continuous dependnt variable with normal distribution

Question 5

What is the t test formula?

Answer 5

signal/noise = (difference btwn 2 gorups)/(variability of the groups)

Question 6

What type of study can u use to determine if atenolol reduces the length of stay for alcohol withdrawl?

Answer 6

RCT

(use atenolol and placebo in 2 withdrawal pts for Tx and control groups, and then measure LOS and use stats to see if there are differences in LOS)

Question 7

Is lenght of stay outcome continuous or categorical?

Answer 7

Continuous

Question 8

This is the test when u comapre 2 means where study subjects are tracked longitudinally. like before-after treatments.

Answer 8

Paired T test

Question 9

This comapres differences between means of 3 or more groups.

Answer 9

Analysis of variance

Question 10

How many groups can u only have for the t test?

Answer 10

2

Question 11

What are the 2 questions u ask for analysis of variance?

Answer 11

is there a significant difference between the menas of the groups?
if so, which gorups are significant;y difference form which others?

Question 12

After u accept or reject Ho, what is always the next thing u have to do?

Answer 12

find the power

Question 13

What is the test to find if there is statistical significant between teh 3 groups in ANOVA?

Answer 13

post hoc tests

Question 14

This si teh tes that termines which groups are significant different from each other in ANOVA.

Answer 14

Tukey’s HSD

Question 15

What is the Q quotient in Tukeys HSD?

Answer 15

its a value that is compared between 2 group means to determine statistical significance

Question 16

Do the difference of means in an ANOVA ebtween 2 values have to higher or lower than the HSD to be of statistical significant?

Answer 16

Higher

Question 17

This is the test that determines the strength and direction of a linear assocaition btwn 2 CONTINUOUS variables.

Answer 17

Correlation analysis

Question 18

What is the signficance of the correlation (r) = 0?

Answer 18

There is no correlation between the 2 variables

Question 19

What is the signficance of the correlation (r) = 1 or -1?

Answer 19

There is a perfect postiive or negative relationship between the 2 variables.

Question 20

Does correlation mean causation?

Answer 20

yes.
wait.

yes.

jk. no
yeah. no.

Question 21

What are the 4 things that must be checked off for causation?

Answer 21

1. an association must exist
2. be nonspuripous (no random chance)
3. confounding factors must be accoutned for
4. biologically plausible

Question 22

This is the test to analyze the assocaition ebtwene 2 or more variables to determine if a relationship exists, predict values of the dependent variable based on the known vlues of the independent variable, and quantifies the amt of variation in dep var explained by ind var.

Answer 22

Regression analysis

Question 23

What does a regression analysis used that minimizes the overall vertical distances between the points and the line in the scatter plot.

Answer 23

Best fit straight line

Question 24

What is the sum of differences of a best fit line calles?

Answer 24

residuals

Question 25

This type of regression uses 1 independent variable to prdeict values of 1 dependent variable.

Answer 25

Simple linear regression

Question 26

This type of regression simultaneously uses the effect of 2 or more independnt variable on 1 dep var

Answer 26

Multiple regression

Question 27

This type of regression is when at lease one of the variables is categorical.

Answer 27

Logistic regression

Question 28

What does the slope of the regresison tell?

Answer 28

if a linear relaitonship exists and the direction of that relationsjop

Question 29

What does the R^2 of the regresison tell?

Answer 29

amt of variation in y that can be explained by x

like R2 - 0.8 means that 80% of the variation in the # of injuries can be explained by cahnges in the players body fat

Question 30

This is the test used to compare difference when in the data are categorical or if a cintinuous variable is not normal.

Answer 30

Non-parametric analysis

Question 31

What is the problem with non-parametric analysis?

Answer 31

less power
uses medians (cuz of skewed distributions)