Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
57 Cards in this Set
- Front
- Back
|
equation for sensitivity using a 2x2 table
|
A/ A+C
|
|
|
equation for specificity using a 2x2 table
|
D/ D+B
|
|
|
equation for PPV using a 2x2 table
|
A/ A+B
|
|
|
equation for NPV using a 2x2 table
|
D/ C+D
|
|
|
equation for Odds ratio using a 2x2 table
|
(A/B) / (C/D)
|
|
|
equation for relative risk using a 2x2 table
|
A/(A+B) / C/(C+D)
|
|
|
from which study can you derive the odds ratio
|
a case control study
|
|
|
from which study can you derive the relative risk
|
cohort study
|
|
|
what effect will high prevelance of a disease in a given population have on the PPV and NPV have?
|
true positives and false negatives will be increased...
this will increase the PPV this will decrease the NPV |
|
|
what effect will low prevelance of a disease in a given population have on the PPV and NPV have?
|
true positives and false negatives will be decreased...
this will decrease the PPV this will increase the NPV |
|
|
why are screening programs more effective when targeting populations at high risk of a disease?
|
screening programs are more effective when targeting populations at high risk of a disease because the PREVELANCE of a disease is higher in "high risk" populations, and a high prevelance will lead to a high PPV
|
|
|
point prevelance
|
total cases/ total population
(total cases in a population at a given time divided by the population at a given time) |
|
|
incidence
|
new cases in a given time/ total pop at risk during that time
(new cases in a population over a given time period divided by the total population at risk during that time) |
|
|
rough equation for prevelance?
|
Prevelance = (incidence) (duration)
|
|
|
in what way are odds ratio and relative risk similar in the event that the disease prevelance is quite low?
|
If the disease prevelance is quite low, OR appoximates RR
|
|
|
attributable risk
|
A/(A+B) / C/C+D
|
|
|
give a sexy example of attributable risk
|
the percentage sexually active people with HPV is 30%
the percentage of non-sexually active people with HPV is 5% Therefore, being sexually active makes you 25% more likely to get HPV |
|
|
Absolute risk reduction
|
C/(C+D) / A/(A+B)
(reverse of attributable risk... this is an equation that will show the risk reduction with a given intervention) |
|
|
give an example of absolute risk reduction (ARR)
|
50% of people who are at risk of disease X get will actually get disease X without intervention
10% of people at risk of disease X will actually get disease X if they use a given intervention (drug or avoiding an action) the ARR in this case is 40%... an at risk person is 40% less likey to get disease X if the intervention is taken |
|
|
Number needed to treat
|
NNT = 1/ ARR
Number needed to treat = 1/ absolute risk reduction how many people do i need to treat to get the desired effect of the drug in 1 person |
|
|
number needed to harm (NNH)
|
1- AR
(AR = atributable risk) (AR= A/(A+B) - C/C+D) |
|
|
mean
|
average of all the avalues
|
|
|
median
|
the value that falls in the middle of all the values
|
|
|
mode
|
the valuea that occurs most commonly
|
|
|
when you're talking about a normal distribution, what is similar about the mean, median and mode?
|
mean = median = mode
all three are equal in a normal or Gaussian distribution (a perfect bell curve) |
|
|
this type of distribution is centered around two different modes wherein the participants are falling into 2 different group
|
bimodal distribution
|
|
|
describe a positively skewed distribution
|
tail to the right
Mean > median > mode mode is leas affected by outlier in this sample |
|
|
describe a negatively skewed distribution
|
asymmetry with tail to the left
Mean < median < mode |
|
|
this type of bias occurs when the referral centers for a trial of a new anticancer drug have more pateint swith end stage disease than early stage, so more patients with end stage disease are referred for the trial
|
Selection bias: non random assignment of a study group
|
|
|
this type of bias occurs when the study is performed on patients that have been hospitalized
|
Berkson's bias (a type of selection bias)
|
|
|
parents of a child with autism have a more detailed recall of events and illnesses in their chil's first two years of life compared to parents of healthy controls
|
recall bias: knowledge of the pressence of a desease alters the recall of participants
|
|
|
a study performed in china may not be generalizable to the US population
|
sampling bias: subject are not representative relative to general population; therefore, results are not generalizable.
|
|
|
sending a surve out to people diagnosed with a fatal illness 5 years after diagnosis will preferentially sample those wiht a low grade disease (or fewer comorbidities)
|
late-Look bias: information gathered at an innapropriate time
|
|
|
the positive benefit of a new drug during a study simply may have been due to the fact that study participants were required to attend clinic monthly, where they recieved extra disease education and counseling compared to with the controles
|
procedure bias: subjects in different groups are not treated the same
|
|
|
are asbestos miners more likely to have cancer because they mine asbestos or bc they are more likely to smoke
|
confounding bias: occurs with 2 closely associated factors, the effect of 1 factor distorts or confuses the other
|
|
|
explain lead time bias
|
Lead time bias: it appears that the course of a disease has been changed bc we can detect it earlier
(i.e. normally you've got 15 years to live after detecting prostate cancer... then we make a test that discovers the cancer 5 years earlier... so it looks like you live 20 years after getting diagnosed, but really its that same amount of time) |
|
|
explain the pygmalion effect
|
occurs when a researcher's belief in the efficacy of a treatment changes the outcome of the treatment... (my version: home team effect)
|
|
|
when studying the effects that infection control eductation has on physicians, the incestigator notes that both the experimental and the control groups improve their hand hygeine
|
Hawthorn effect: ocurs when groups being studied changes its behavior owing to the knowledge of being studied
|
|
|
what does the P value represent
|
the P value suggests the probability that the given result occured due to chance alone...
a very low P value is good for proving a point |
|
|
what is the normal P value set at? what does it mean if the results score a P value less than this set numeber?
|
Normally, P=.05
if P <.05 it is agreed that a given event did not happen by chance |
|
|
describe a type I error (aka alpha error)
|
false positive error
stating that there is an effect or difference when non exists |
|
|
describe a type II error (aka beta error)
|
false negative error
stating that there is not an effect or difference when one exists |
|
|
what is the POWER of a test
|
(1- beta)
or 1 - the probability of a type II error 1 - (the probability that you said there was not an effect when there was) 1- false negative |
|
|
what can be said about a P value > .05
|
if the P value is > 0.05 then the null hypothesis is accepted and the alternative is rejected.
It is aggreed that the results of the treatment are designated not statistically significant |
|
|
what can be said about a P value < .05
|
if the P value is <0.05 then the null hypothesis is rejected and the alternative is accepted.
It is aggreed that the results of the treatment are designated statistically significant |
|
|
what percent of the population studied is contained within 1 standard deviation of the mean? 2? 3?
|
1 = 68%
2 = 95% 3 = 99% |
|
|
standard deviation of the mean
|
SEM = standard deviation/ the square root of the sample size
|
|
|
explain a qualitative value change in the SEM
|
standard deviation of the mean is inversely proportional to the sample size
|
|
|
what is the effect of increasing you confidency interval frok 95% to 99%?
|
you would get a broader range of numbers
|
|
|
what is the relation of sample size to confidence interval
|
the sample size is inversely to the broadness of the interval
i.e. the larger the sample size, the smaller the interval... the smaller the sample size, the broader the interval |
|
|
give a very brief description of the difference between a T-test and an ANOVA test
|
T-test compares the difference between the means of two different groups
ANOVA compares the means of several different groups |
|
|
this test compares the differences between the percentages or proportions of categorical outcomes
|
chi-square
|
|
|
describe the correlation coefficient
|
r = correlation coefficient
r is always between - and +1 the closer the absolute vlue of r is to 1, the stronger the correlation |
|
|
the best way to test or perform analysis of variance checks the difference in means between 3 or more groups
|
anova
|
|
|
examins categorical outcomes and looks for differences between categorical outcomes or proportions
|
chi squared
|
|
|
best test to compare the difference in the means of the two study groups are examined.
|
t-test
|
|
|
measures the association that exists among two variables
|
linear regression analysis
|
