Biostats

Question 1

what are the 2 ways to increase statistical power?

Answer 1

1. increase sample size

1. decreased Beta/Type 2 error

Question 2

define a QUALITATIVE study

Answer 2

study collects data with questionaires/moderators/discussions/interviews to understand social phenomenon (adherence to meds, barriers to tx, how pts FEEL about disease/stigma)

Question 3

1 SD=

2 SD=

3 SD=

Answer 3

1 = 68  +/-10
2= 95. +/-20
3= 99.7. +/-30

Question 4

differentiate between a Pre-protocol study and an Intention to Treat study

Answer 4

Pre-protocol = only include participants who adhered to study

Intention to treat= will include ALL participants (even the ones who dont adhere/drop out)

Question 5

how does a pre-protocol study effect:

strength of association (effect size)
external validity (generalizability)
risk of biases

Answer 5

increases effect size (strength of association); causes overestimation of the new intervention’s effect

decreases external validity (generalization); bc its no longer randomized

increases risk of biases; bc the original randomization of the trial is lost

Question 6

how does an Intention to Treat study effect:

strength of association (effect size)
external validity (generalizability)
risk of biases

Answer 6

decreases strength of association (effect size); bc it avoids overestimation of the new intervention’s effect

increases external validity (generalizability); maintains randomization

decreases risk of biases; maintains randomization

Question 7

what are the criteria for MediCARE?

Answer 7

> 65yo (I CARE for OLD ppl)
permanent disability
ESRD

Question 8

what is the single criteria for medicAID?

Answer 8

temporary AID for pp <65yo (housing, pregnant, children, stimulus, etc)

Question 9

what does Chi-square test measure?

Answer 9

measures significance between 2 CATEGORICAL variables (ie BMI and gender)

*Chi-tegorical :) *

Question 10

what does a paired t test meausre?

Answer 10

measures the MEAN/AVERAGE of variables @ 2 DIFF POINTS in time!!
ex: if the researchers were comparing the Weight at baseline and 3 months after surgery.

paired for a life-TIME

Question 11

a study that states “On average” will have what 2 options as an answer choice?

why?

Answer 11

paired t test OR unpaired test OR anova (measures multiple variables)

bc they all measure the mean/average of continuous variables

Question 12

what does an UNpaired t test measure?

Answer 12

the mean/average of continuous variables in a single point in time (not )

Question 12

what does an UNpaired t test measure?

Answer 12

the mean/average of continuous variables in a single point in time (NOT TIME dependent)

UNpaired = not time related

Question 13

what does Kaplan-Meier analysis measure?

Answer 13

the incidence of an event occurring over time (incidence of MI, death, etc)

Question 14

what does the pearson correlation coefficient (r) measure?

Answer 14

whether there is a linear association between two continuous variables

example: to assess the relationship between BMI and serum cholesterol levels on a line graph

Question 15

what are the 2 most common risk factors of in-hospital mortality?

Answer 15

SHOCK
OLD AGE

high BMI

Question 16

Adding outliers causes a change in mean/median/mode?

Answer 16

Median will always change when an outlier is added;

large outlier = positive skew/inc in SD = higher median
small outlier = negative skew/dec in SD = smaller median