Graphs & Numerical measures
Testing Rules
Probability
Theories & Formulas
Important Stuff
100
These graphs look the same but one is used to show different categories and one is used to show distributions
What is a bar chart and a histogram?
100
These two hypotheses cover all possible values of the population mean. One is used when there is no difference in standard deviation and one is used when there is a difference
What is Null Hypothesis and Alternative Hypothesis
100
___________ is the probability that patient has condition given that test is positive and ___________ is the probability that patient does not have disease given that test is negative
What is sensitivity and specificity?
100
This theorem states that if n is large enough the distribution of sample means is approximately normal even if the underlying distribution is NOT
What is the Central Limit Theorem
100
The amount of numbers that are freely changed. Number of independent observations in a quantity. Penalty for using sample standard deviation instead of population standard deviation.
What is Degree of Freedom?
200
A description of the data including: 1. MINIMUM 2. 1ST QUARTILE 3. MEDIAN 4. 3RD QUARTILE 5. MAXIMUM
What is the 5 number summary? The middle 50% of data (between the 25th percentile and 75th percentile) are called “interquartile” range.
200
____ _ _____ is described as: alpha (a), rejecting a null hypothesis when it is true; rejection error: reject when you should not. ____ __ _____ is described as: beta (b), failing to reject a null hypothesis when it is false; acceptance error; accept when you shouldn’t.
What is type I error and type II error?
200
Events that cannot occur simultaneously are called ________ _________ ; occurrence of one event prevents occurrence of the other event. Use the additive rule P ( A U B ) = P(A) + P(B), since P that A intersects B = 0 and P ( A U B ) = prob A occurs + prb B occurs minus prob both a and b occur
What is mutually exclusive and not mutually exclusive?
200
This is used to determine characteristics of screening tests Characteristics = conditional probabilities Ex. Probability that a woman is actually pregnant given that her home pregnancy test is positive?
What is Bayes Theorem.
200
This type of test looks for homogeneity or independence and compares expected value (under null) with observed values. It also tells about probability but not MAGNITUDE of differences in odds ratios and relative risks for umatched designs.
What is Chi-Square test?
300
________ describes variability or Spread around the mean and the ________ _________ is found by taking the square root of that number.
What is Variance and Standard Deviation?
300
Use this test___________ when asking if greater than or less than a. Use this test___________ when asking if two things are unequal, looking at difference between two things
What is 1 sided t-test and 2-sided t-test
300
Probabilities of two or more mutually exclusive events sum to 1. P (A) + P (B)+ P(C) . . . =1; Example: Every individual must fall into one of the groups of currently employed, currently unemployed, or not in labor force.
What is Mutually Exhaustive?
300
a ________ distribution is the discrete probability distribution of the number of successes in a sequence of n independent yes/no experiments, each of which yields success with probability p.– parameters are n (number of times you tried) and p (probability of success or failure). Ex: Coin tossing and A ______ distribution is continuous and can be any value (ex: height or blood pressure).– parameters are u (population mean) and o (population st. deviation)- Symmetrical, bell shaped, central limit theorem
What is Binomial distribution and normal distribution
300
We use this test for comparing the means of two samples (or treatments), even if they have different numbers of replicates. This test compares the actual difference between two means in relation to the variation in the data.
What is a t-test?
400
This Rule states that when data is symmetirical and Unimodal; 68% observations within +/- 1 st. dev 95% observations within +/- 2 st. dev 99 + observations within +/- 3 st. dev
What is the Empirical Rule?
400
This word describes the probability of finding a statistically significant difference in a study, given that there really is a difference (or association)
What is Power?
400
Both of these values depend on prevalence of disease in a population. HINT: Screening for a disease which is rare in a population will generate many false positives.
What is positive and negative predictive value?
400
This type of distribution is similar to normal distribution BUT it has thicker ‘tails’ because it is less precise. Looks like Z score BUT uses estimates of population mean and estimate of population std deviation. When n>= 30, this distribution is very close to normal distribution. Most important to make the distinction with small sample sizes.
What is t-distribution?
400
Odds ratios can be used in these type of study designs: ___________, _____________, and _____ _________. Relative Risk only works with this type of study design: ___________
What is: Prospective (cohort studies), Retrospective (case control studies) and Cross sectional studies? and Prospective
500
The names of the following graphs, charts and plots are...
What is bar charts, histograms, one-way scatter plots, box plots, and line graphs
500
Calculate given the significance level (a), power (1-b), and desired difference to be detected Affected by the size of difference to be detected, power chosen(higher increases n), size of alpha (smaller alpha increases n), variability in population (more variable, larger n needed)
What is the method used to calculate sample size?
500
VOCABULARY: Match the word to the definition: Complement, intersection, conditioned probability, independent, union 1. events A and B both occur, “cap” 2. either A or B or both A and B occur. “cup” 3. complement of event A= A does not occur, “not a”, Ac 4. chance that one event will occur is the same whether or not the other event has occurred 5. Probability of A given B Will equal probability of A if events are independent
What is 1. Intersection 2. Union 3. Complement 4. Independent 5. Conditional Probability
500
This is a type of interval estimate of a population parameter and is used to indicate the reliability of an estimate.
What is confidence intervals? Assumes population mean and population std deviation are known. The greater the level of confidence, the more accurate the estimation. Larger n = smaller standard error > narrow CI Smaller population st. dev=smaller SE > narrow CI Usually for 95% CI the formula is x bar +/- 1.96*StdDev/squrt n
500
When using ____ sampling, for each observation in 1st group there is a corresponding observation in the 2nd group (Use 1-sample T –Test) In ___________ sampling, we collect continuous data from 2 independent samples and assume a normal distribution. (Use 2-sample T-Test)
What is Paired Sampling and Independent Sampling?
M
e
n
u