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;
53 Cards in this Set
- Front
- Back
define variable
what does it generate? |
measures of a single characteristic that can vary
Characteristics of a population or person vary (e.g., blood pressure) These characteristic measures generate study data |
|
if you are measuring the amount of caffiene a person is exposed to, and the outcome of pancreatic cancer, what are caffinene and cancer as far as variables?
|
Caf: Independent variable
Cancer: dependent variable |
|
what is an independent variable?
|
Set/determined by Investigators
exposure/intervention |
|
what is a dependent variable?
|
The effects that depend upon the independent variable(s)
the outcome |
|
what re the 2 main types of variables (not dep/indep)
|
Qualitative: descriptions (words, not numbers) of variable features
Quantitative: uses rigid, continuous measurement scale |
|
what is a qualitative variable
|
descriptions (words, not numbers) of variable features
|
|
what is a Quantitative variable
|
uses rigid, continuous measurement scale
|
|
Nominal variable types have low or high information content?
|
LOW
|
|
example of a nominal variable
|
Blood type; color = cyanotic or jaundiced; taste = bitter or sweet
|
|
binary variables are what? (example plz)
|
Sex; heart murmur = present/absent
|
|
please give an example of continuous variable data and tell if it has high or low information content.
|
Temperature (Fahrenheit)
second highest info content |
|
please give an example of ratio variable data and tell if it has high or low information content.
|
Temperature (Kelvin); blood pressure
Highest info content |
|
rank the following in order from lowest to highest information content:
nominal, ordinal, binary, ratio, continuous |
nominal
binary ordinal continuous ratio |
|
what is the goal of statistical methods?
|
explain variation
|
|
what is the mean (mu)
|
AVERAGE
would be the middle of the bell curve |
|
what is the mode?
|
measurement value that occurs most frequently
|
|
median
|
equal number of variables above and below this point
|
|
what is central tendency?
|
central feature of the data
mean, mode, median |
|
Mean
|
the average value
|
|
Mode
|
most commonly observed value
|
|
Percentiles (quantiles)
|
percentage of observations below the indicated point when all the observations are ranked in descending order
Median = 50th percentile |
|
Range
|
lowest to highest value
|
|
Variance ** ON TEST
|
measure of variability of data about the mean (sum of squared deviations from the mean, or s2)
|
|
Skewness
|
has to do with horizontal distribution of data (wider than it should be)
Horizontal stretching of a frequency distribution to one side or the other creates one long, “thin” tail to the data distribution |
|
Kurtosis
|
has to do with vertical distribution of data (too tall)
Vertical stretching or flattening of the frequency distribution |
|
when looking at a bell curve and determining skew, what do you look at
THIS IS ON THE TEST |
which side below the highest point has more values
SLIDE 17 is SKEWED LEFT |
|
Deductive
|
info known from something general and then applying it to something specific
proceed from the general to the specific Features of a larger population or general scenario are applied to a specific person or situation |
|
Inductive
|
take specific info general population (this is what epidemiology does)
proceeds from the specific to the general Seeks to find general principles and valid generalizations from specific data |
|
What type of reasoning is the following?
All Americans Love Democracy People who Love Democracy Vote John Doe is an American SO John Doe Loves Democracy and Votes |
Deductive Reasoning
|
|
What type of reasoning is the following?
Ben, a 78 yo man, has a systolic BP of 140 mm Hg Hank, a 44 yo man, has a systolic BP of 128 mm Hg Bob, a 21 yo man, has a systolic BP of 116 mm Hg SO Systolic BP in men increases with age |
Inductive Reasoning
|
|
what is a hypothesis?
|
predictions about what an examination of appropriately collected data will show
|
|
caffeine consumption increases the risk of pancreatic cancer..
this is an example of what? |
a hypothesis
|
|
what is a Null hypothesis (H0)
|
there is no statistically significant difference between two groups
|
|
People who consume more caffeine are not at greater risk of developing pancreatic cancer than those who consume less caffeine
this is an example of what? |
a null hypothesis
|
|
what is an alternative hypothesis? (H alpha)
|
there is a statistically significant difference between two groups
|
|
People who consume caffeine are at greater risk of developing pancreatic cancer than people who do not
this is example of a what? |
Alternative hypothesis (Ha):
|
|
Statistical analyses test what?
|
the null hypothesis
|
|
using a 2x2 table please fill it in according to hypothesis testing
|
see slide 27
TOP: H0 is true | H0 is false Vertical : Fail to reject H0 Reject H0 A=correct B=type II error (beta) C=Type I error (alpha) D=correct |
|
what is alpha? usually set at?
|
= Type I error = the probability of incorrectly rejecting the H0
alpha is usually set at 0.05 |
|
what is Beta?
|
= Type II error = probability of failing to reject the null hypothesis when it is false
|
|
type II error is =?
|
Beta
= probability of failing to reject the null hypothesis when it is false |
|
type I error is =?
|
alpha
the probability of incorrectly rejecting the H0 alpha is usually set at 0.05 |
|
what is the p value
|
probability that the 2 groups ARE NOT DIFFERENT
anything less than .05 then you reject null hypothesis |
|
what is 95% Confidence Interval (CI)
|
For normally distributed data, the mean ± 1.96 SE (standard error) estimates the range within which 95% of means of repeated measures would be expected to fall --> this range is the 95% CI
|
|
If RR= 1 what does this mean?
|
there is no differentiation in having the disease and not
rr of disease/ rr of not disease with both exposed no comparison, not significant |
|
if RR= below 1 what does this mean
|
greater chance of being exposed and NOT having disease
|
|
if RR= is greater than one
|
you have greater chance of being diseased with exposure
disease/no disease |
|
REVIEW SLIDE 30
|
it was confusing
|
|
have RR= 1.8 +/- .92
what is your mean? |
1.8
|
|
have RR= 1.8 +/- .92
what is your 95% CI range? |
.88 - 2.72
basically add/subtract .92 from 1.8 |
|
if the value of 1 falls within the 95% range, what does this mean?
ON TEST |
your information is not statistically significant
the value of 1 has to exist in either the positive or negative tail so if you have have RR= 1.8 +/- .92 this means 1 falls within that range, and isn't significant |
|
if you have a 95% CI of 1.3-1.6 what does this mean?
|
you can reject the null hypothesis
so your variable does cause the outcome |
|
what is the definition of Number Needed to Treat (NNT)
|
Guess what, it doesn't matter cuz it is not on the test! BOOM
|