ECBM Lecture 1--stats review

Question 1

scientific method

Answer 1

prove or disprove hypothesis

Question 2

precision

Answer 2

Target: clustered in same spot

Question 3

accuracy

Answer 3

Target: hit bullseye

Question 4

necessary characteristics of research

Answer 4

Objectivity
precision 
verification
economic 
reasonable

Question 5

Research: Observational

Answer 5

Exploratory –> often when thing under observation would be unethical

Question 6

RCT

Answer 6

randomized control trial

Question 7

Comparitive analysis

Answer 7

research on drug A vs placibo
research on drug B vs placibo
comparison

Question 8

Strongest evidence in clinical research?

Answer 8

Randomized control trial RCT

Question 9

Observational study divisions

Answer 9

Cohort– groups with/ without exposure –>outcome

Case Control– grps exposure with/without disease–> outcome

Question 10

N

Answer 10

population–compare pt to population

Question 11

n

Answer 11

sample–characterized by statistics

Question 12

n statistics

Answer 12

draw conclusions and broadly apply

Question 13

variable being manipulated (the intervention)

Answer 13

Independent variable

Question 14

any variable being measured (the outcome)

Answer 14

dependent variable

Question 15

whole units of data Qualitative

Answer 15

discrete –

Question 16

data with range

Answer 16

continuous

Question 17

ranked data (1st, 2nd, 3rd)

Answer 17

ordinal scale

Question 18

ordered data w/out a meaningful zero (water temp)

Answer 18

interval scale–0 temp is an actual temp

Question 19

bar graph with no spaces–for trend

Answer 19

histogram

Question 20

most frequent #

Answer 20

mode

Question 21

value in middle

Answer 21

median

Question 22

standard deviation

Answer 22

68% of data to either side of mean on line graph = confidence interval

Question 23

confidence interval most often used

Answer 23

95% confidence – related to P needing to be less than 0.05

Question 24

trustworthy graph

Answer 24

tall and narrow–wide curve not precise

Question 25

when asked "what is the confidence interval?"

Answer 25

compare upper and lower limits--not 95%

Question 26

empirical rule

Answer 26

when you have a normalized bell curve--two standard deviations from mean will give 95% confidence interval

Question 27

null hypothesis is rejected when true

Answer 27

Type I error--ex. assume you have enough $ for groceries, but you don't

Question 28

the null hypothesis is kept when it is fake

Answer 28

Type II error-- you assume you don't have enough $, but you don't

Question 29

Better error to make in medicine?

Answer 29

Type II error--underestimate drugs efficacy

Question 30

Regression

Answer 30

predict one factor from existing graph/data

Question 31

how benefitial is drug

Answer 31

treatment effect

Question 32

which numbers for 2X2 square?

Answer 32

think of scenario | read abstract

Question 33

2X2 -- the scenario

Answer 33

always deal with PEOPLE--may need to add/subtract groups

Question 34

2X2 -- the abstract

Answer 34

avoid scores--look for PEOPLE--may be in percentages

Question 35

For 2X2-- what is the % of people who something happened to

Answer 35

Event Rate (EER)--incidence of some thing--may be in abstract

Question 36

First step

Answer 36

calculate EER (experimental event rate) and CER (control event rate)

Question 37

second step

Answer 37

compare EER and CER (risk ratio, relative risk)

Question 38

see ratio think...

Answer 38

divide RR = EER/CER

Question 39

100 minus risk ratio-- 100-X= Y

Answer 39

relative risk reduction RRR

Question 40

how many people would I need to treat with drugX to save one life

Answer 40

number needed to treat NNT

Question 41

NNT

Answer 41

100%/ARR (in %)

Question 42

apply NNT math to negative outcomes

Answer 42

number needed to harm NNH

Question 43

odds for treatment/ odds for control

Answer 43

odds ratio (OR) -- Odds vs risk

Question 44

odds

Answer 44

"yes" column divided by "no"

Question 45

positive test has diseas

Answer 45

true positive TP

Question 46

negative test has disease

Answer 46

False negative FN

Question 47

positive test doesn't have disease

Answer 47

False positive FP

Question 48

negative test doesn't have disease

Answer 48

true negative

Question 49

sensitivity=TP/(TP+FN) FN will impact sensitivity most--only on one side of equation

Answer 49

true positive rate (% positive test results in pt who have the disease

Question 50

SnNOUT

Answer 50

snesitivity--"SeNsitivity means Negative rules ""it"" OUT"

Question 51

SpPin

Answer 51

"Specificity means Positive rules ""it"" in"

Question 52

SpPin

Answer 52

low false positivity-- | true negative/true negative + false positive

Question 53

SnNout

Answer 53

low false negativity= | true positive/true positive + false negative

Question 54

sensitivity and specificity equation key

Answer 54

look for the data point on only one side of equation--this factor will play in more to results

Question 55

sensitivity and specificity combined

Answer 55

Likelihood Ratios

Question 56

TPR + FNR = | TNR + FPR=

Answer 56

100% -- always

Question 57

default hypothesis that A and B have no correlation i.e. drug has no effect

Answer 57

null hypothesis