Epidemiology And Biostats 1 And 2 9/14/05

Front 1

Hugh levels 3 levels:
-___keeps disease from becoming established(immunization,fluoridation,seatbelt use)
-interrupts disease before symptomatic (cancer screening)
-limits physical social consequences of symptomatic disease (med or surgery treat, rehab)

Back 1

Primary
Secondary
Tertiary

Front 2

1900's disease due to_______
1990's due to________

Back 2

infxs (pneu,TB,Diarr,Heart disease)
All lifestyle choices

Front 3

Epidemiology uses ____reasoning- specific to general , John Snow-cholera

Back 3

Inductive reasoning

Front 4

BEINGS model

VECTORS btw agent,host,enviro

Back 4

Biological Behavioral
Environmental
Immunologic
Nutrition
Genetic
Services, Social, Spiritual

Front 5

___ biological spectrum of disease, few deaths, litle more clinical illness, lot more asymptomatic cases

Back 5

The iceberg phenomenon

Front 6

___ number new cases in pop during study period
____ preexisting who have disease at certain point in time

Back 6

Incidence

Prevalence

Front 7

Present Absent

Present a b

Absent c d

Relative risk =
Odds ratio=

Back 7

[a/(a+b)]/[c/(c+d)]

ad/bc

Front 8

___data indicate diff categ. cant be ranked (N,S,E,W, Blood O A AB B)
____type nominal only 2 levels (well/sick,live dead,M/F)
___grouped and ranked but not measure rank (none, mild, moderate, severe symp class low mid high)
____not ordered, but measurable(BP, temp, weight,height, blood sugar)
_______ 200lb is 2x 100lb

Back 8

Nominal
Dichotomous(binary)
Ordinal
Continuous (interval)
Ratio

Front 9

Variance formula _____
squared, not same units

Standard devi form_____
square variance
small SD=

Back 9

sum(X-mean)2/N

sq root of variance (how much varies from mean)

values dont vary much from mean

Front 10

Normal dist

w/in 1 SD:
w/in 2 SD:
w/in 3 SD:

Back 10

68.2%
95.4
99.6%

Front 11

_____ hypoth there is no true correlation btw 2 things or no difference in means

_____ max probablity there is an association (when there isnt) 0.05

Back 11

Null

Alpha

Front 12

____hypoth that there is a true diff btw means or prop or is an association btw 2 var

____ probab the observed diff could be due to chance

Back 12

Alternative

P-value

Front 13

____ error false neg fail to reject null that is false

_____ error (false positive reject null that is really true

Back 13

Type II

Type I

Front 14

____ test to compare means of continuous variable to determine if the diff btw 2 means is more than expected due to chance (t distr flatter than normal)
_____used in before and after exp.sub followed around to see if change in value of contin var (BP) any varia attributable to intervention or changes over time in same subject

Back 14

Student's t test

Paired t-test

Front 15

_____ test determine prob of seeing specific calculated value, larger = less likely due to chance uses for 2x2

Back 15

Chi square test independence

Front 16

___a stat test strength of linear relationship btw var which are continuos (weight vs height) correl coeff r r close to 1_______

Back 16

simple linear regression

more closely related variables are

Front 17

sensitivity form:
Specificity form:
positive predic value form:

Back 17

a/(a+c)

d/(b+d)

a/(a+b)