How can we help?

You can also find more resources in our Help Center.

probability=

4 steps of hypothesis testing

null hypothesis

random sampling variation

chance

1. making hypothesis... 2. set criteria for decision... 3. gather d…

there is no difference between your sample and the population…

all observed differences are due to chance

probability=

chance

4 steps of hypothesis testing

1. making hypothesis... 2. set criteria for decision... 3. gather d…

Selection bias

Measurement bias

Experimenter Expectancy bias

Lead time bias

sample not representative of population - pick more at random

info presented gives distorted picture - need control group

pygmalion effect - experimenters' expectations perceived by s…

disease discovered sooner, so seem to live longer - use life…

Selection bias

sample not representative of population - pick more at random

Measurement bias

info presented gives distorted picture - need control group

Sensitivity

Specificity

PPV

NPV

probability that a test detects dz when dz is present... propor…

- true-negative rate... - probability that a test indicates no d…

probability that the person actually has the dz when given a…

- proportion of negative test results that are TN... - probabili…

Sensitivity

probability that a test detects dz when dz is present... propor…

Specificity

- true-negative rate... - probability that a test indicates no d…

A contingency table is widely used to…

what is the R code to build a continge…

risk difference is defined as?

NNT or NNH can or cannot contain a dec…

the association between 2 categorical variables

table(data~independent,data~outcome)

a measure of the difference between two proportions... (a/(a+b))…

cannot (don't have half a person)... round UP

A contingency table is widely used to…

the association between 2 categorical variables

what is the R code to build a continge…

table(data~independent,data~outcome)

Nominal

Ordinal

Interval

Ratio

- Data are coded by a number, name, or letter that is assigne…

- categories are ordered by rank... - Examples: Pain level (e.g…

- Measurements are ordered (like ordinal data) and have equal…

- Ratio: Measurements have equal intervals & There is a true…

Nominal

- Data are coded by a number, name, or letter that is assigne…

Ordinal

- categories are ordered by rank... - Examples: Pain level (e.g…

95% CI

Relative risk/Risk ratio (RR)... Probabi…

Number needed to treat (NNT)

Odds ratio (OR)... Represents that proba…

-When 95% for estimated DIFFERENCE between groups (or within…

-Probability of an unfavorable event occurring in the tx grou…

NNT = 100 / % ARR... MUST ALWAYS around up this number! (e.g. 3.…

-ratio of two odds, or the ratio of the odds of an event occu…

95% CI

-When 95% for estimated DIFFERENCE between groups (or within…

Relative risk/Risk ratio (RR)... Probabi…

-Probability of an unfavorable event occurring in the tx grou…

alpha

beta

power

chi square test

- related to p-value... - chance that we will have a type 1 erro…

- chance that we will have a type II error - we couldn't reje…

- 1-beta

- use when you have categorical (as opposed to nominal/contin…

alpha

- related to p-value... - chance that we will have a type 1 erro…

beta

- chance that we will have a type II error - we couldn't reje…

Scientific Method

Descriptive Statistics is also called

Descriptive Statistics

Statistical Inference is also called

Testing and eliminating hypotheses whose predictions are inco…

Exploratory Data Analysis (EDA)

Organization and summarization of data; graphical display to…

Confirmatory Data Analysis (CDA)

Scientific Method

Testing and eliminating hypotheses whose predictions are inco…

Descriptive Statistics is also called

Exploratory Data Analysis (EDA)

Incidence

Z Score

Crude vs. cause specific mortality

Case-fatality rate

the number of new cases that develop in a population over a c…

Indicates how many standard deviations a given value is from…

Crude mortality rate: # of deaths/total population size... Caus…

# of deaths from a specific disease/# of people affected by t…

Incidence

the number of new cases that develop in a population over a c…

Z Score

Indicates how many standard deviations a given value is from…

Relative Risk

Odds ratio

Negative predictive value

Case-control studies

Typically used in *cohort studies*. Risk of a group developin…

Typically used in *case-control studies. * Odds that group wi…

True negatives divided by (true negatives and false negatives)

Observational & *retrospective*... Compares group of people with…

Relative Risk

Typically used in *cohort studies*. Risk of a group developin…

Odds ratio

Typically used in *case-control studies. * Odds that group wi…

biostatistics

evidence based medicine

anchoring

availability

application of statistics and mathematical testing in medicine

using the best information to make decisions about diagnosing…

fix on particular information, think in linear way

strongly influenced by dramatic or unusual cases that are pro…

biostatistics

application of statistics and mathematical testing in medicine

evidence based medicine

using the best information to make decisions about diagnosing…

simple random sampling

systemic sampling

stratified sampling

cluster sampling

each individual has equal chance of being selected

select at regular intervals through ordered list... -pick random…

separate pop into strata based on a characteristic... randomly…

pop already in strata... -randomly select groups of strata

simple random sampling

each individual has equal chance of being selected

systemic sampling

select at regular intervals through ordered list... -pick random…

cases

variable

categorical variable

quantitative variable

the subjects/objects we obtain information about (rows of a d…

any characteristic that is recorded for each case

places each case into exactly one of two of more categories

measures/records a numerical value for each case (makes sense…

cases

the subjects/objects we obtain information about (rows of a d…

variable

any characteristic that is recorded for each case

Sensitivity

Specificity

Positive predictive value

Negative predictive value

True positives / true positive + false negatives

True negatives / true negatives + false positives

True positives / true positives + false positives

True negatives / true negative plus false negatives

Sensitivity

True positives / true positive + false negatives

Specificity

True negatives / true negatives + false positives

Cross sectional study

Case- Control Study

Cohort Study

Phase 1

Data from a group of people to assess the frequency of a dise…

Compares a group with a disease to a group without the diseas…

Compares a group with a given exposure or risk factor to a gr…

Small # of healty volunteers- is it safe?

Cross sectional study

Data from a group of people to assess the frequency of a dise…

Case- Control Study

Compares a group with a disease to a group without the diseas…

Central Limit Theorem

Standard Errors (SE)

Standard deviation (s)

Standard error of the mean (SEM)

n must be > 60, if sample size is large, distribution of samp…

standard deviation of sampling distribution, measures precisi…

measures variability/spread amount individual observations in…

measure of the precision of the sample mean; SEM = s/root(n);…

Central Limit Theorem

n must be > 60, if sample size is large, distribution of samp…

Standard Errors (SE)

standard deviation of sampling distribution, measures precisi…

Cross Sectional

Case-control

Cohort

Twin concordance

Data from a group of people to assess frequency of dz at a pa…

Compares a group of people without a dz to a group with dz.…

Compares a group with a given exposure or risk factor to a gr…

Compares frequency with which both monozygotic twins or both…

Cross Sectional

Data from a group of people to assess frequency of dz at a pa…

Case-control

Compares a group of people without a dz to a group with dz.…

Cross-sectional Study

Case-control study

Cohort study

Phases of a clinical trial

Disease prevelance... Basically it's like looking at a bunch of…

Looks for prior exposure or risk factor... uses odds ratio ... Comp…

Compares a group w/ exposure to group w/o - looks to see if e…

1: yo we healthy, is this shit safe?... 2: Small sick group, is…

Cross-sectional Study

Disease prevelance... Basically it's like looking at a bunch of…

Case-control study

Looks for prior exposure or risk factor... uses odds ratio ... Comp…

Variation

Variation examples

Epidemiology

Statistics is the cornerstone of publi…

common feature of most organisms ... (may be environmentally inf…

Weight, height, reaction times, aptitudes, and disease resist…

study of the patterns, causes, and effects of health and dise…

risk factors for disease and targets for preventative medicine

Variation

common feature of most organisms ... (may be environmentally inf…

Variation examples

Weight, height, reaction times, aptitudes, and disease resist…

Cross-Sectional Study

Case-control study

Cohort study

Twin concordance study

Collects data from a group of people to assess the frequency…

Compares a group of people with disease to a group without di…

Compares a group with a given exposure or risk factor to a gr…

Compares the frequency with which both monozygotic and dizygo…

Cross-Sectional Study

Collects data from a group of people to assess the frequency…

Case-control study

Compares a group of people with disease to a group without di…

When can the mean be used?

What type of data can the mean be used…

Is the mean sensitive to outlying valu…

When should the median be used?

when data is not skewed

continuous data

yes

when data is skewed

When can the mean be used?

when data is not skewed

What type of data can the mean be used…

continuous data

mutually exclusive

independent

additive rule is for ___... multiplicativ…

ex. complete blood count: probability…

-either OR (NOT both)... ex) heads OR tails

-and (either event could occur)... ex) roll 6 AND 2

add="or"... multi="and"

(0.95)^9

mutually exclusive

-either OR (NOT both)... ex) heads OR tails

independent

-and (either event could occur)... ex) roll 6 AND 2

Multivariate analysis is used for:

Linear Regression Model

Results of a regression analysis are p…

How to read beta coefficients for line…

looking at more than 1 independent variable at the same time

Use when you have a:... - Continuous dependent variable (ie CD4…

Estimates of population parameters for each of your independe…

If one has an abdominal measurement that is 1 cm greater than…

Multivariate analysis is used for:

looking at more than 1 independent variable at the same time

Linear Regression Model

Use when you have a:... - Continuous dependent variable (ie CD4…

cross-sectional study

case-control study

cohort study

twin concordance study

Assess frequency of disease at a particular point in time. "W…

(retrospective) compares group of people with disease to grou…

(Prospective [who will develop the disease] or retrospective…

compares the frequency of disease development between monozyg…

cross-sectional study

Assess frequency of disease at a particular point in time. "W…

case-control study

(retrospective) compares group of people with disease to grou…

What is sensitivity?

What is specificity?

100 patient with disease and test dete…

What is the predictive value?

Proportion of subject with disease who test positive

Proportion of subject without disease who test negative

90 / (90 pos + 10 neg) = 90%

help decide whether to believe result for an individual patient

What is sensitivity?

Proportion of subject with disease who test positive

What is specificity?

Proportion of subject without disease who test negative

What is a variable?

What is an independent variable?

What is a dependent variable?

What are the four data types?

Measurement or description of a characteristic

value known of determined before the study starts... *some inde…

value determined during/after the study

1) numerical/interval **continuous distribution**... 2) nominal ca…

What is a variable?

Measurement or description of a characteristic

What is an independent variable?

value known of determined before the study starts... *some inde…

population

sampling

sampling methods

simple random sampling

a complete set of individuals, objects, or events, that share…

a portion of a population selected to represent the populatio…

1) simple random... 2) systematic... 3) stratified... 4) cluster... 5) co…

the simplest form o sampling methods where all members of a g…

population

a complete set of individuals, objects, or events, that share…

sampling

a portion of a population selected to represent the populatio…

What do the odds ratio and relative ri…

You need to use odds ratio in what kin…

What is one advantage of odds ratios?

The ____ for any event X, considers BO…

When the outcome being analyzed is common; they do approximat…

Case control

It doesn't change depending on whether we focus on the event'…

the Odds

What do the odds ratio and relative ri…

When the outcome being analyzed is common; they do approximat…

You need to use odds ratio in what kin…

Case control

Deviation score

nominal data

ordinal data

interval data

D=X-x... Distance a number is from the mean

data are placed in qualitative groups... ex. based on descriptio…

data are placed in ranking orders ... ex. 1st, 2nd, 3rd ... DOESNOT…

data are in a meaningful order **without** an absolute zero... oft…

Deviation score

D=X-x... Distance a number is from the mean

nominal data

data are placed in qualitative groups... ex. based on descriptio…

What is a type 1 error (alpha)

What is a type 2 error (beta)... What is…

When can a written advance directive b…

In what type of study to you use RR an…

You s**a**w a difference that didn't exist... You say a drug works…

Don't approve a drug when it actually works... **B**lind to a diff…

Probably never...... For sure not in **emergency situation**

Prospective cohort studies... Case-control studies

What is a type 1 error (alpha)

You s**a**w a difference that didn't exist... You say a drug works…

What is a type 2 error (beta)... What is…

Don't approve a drug when it actually works... **B**lind to a diff…

What are the 3 components of Evidence…

What is the goal of EBM?

What are the sources of evidence in or…

"Systematically developed statements t…

Individual clinical experience, best available clinical evide…

Improved patient outcomes

Systems, Synopses, Systematic Reviews, and Studies

Guidelines

What are the 3 components of Evidence…

Individual clinical experience, best available clinical evide…

What is the goal of EBM?

Improved patient outcomes

Statistics

Parameter

Estimate

Categorical

study of methods to describe and measure aspects of nature fr…

describes the population (averages, proportions, measures of…

statistic that is subject to error; calculation from a sample

not numbers

Statistics

study of methods to describe and measure aspects of nature fr…

Parameter

describes the population (averages, proportions, measures of…

Confound variables

Mensurative experiment... vs. Manipulativ…

Random Sampling vs Regular Sampling

3 variables

the experiment design should control for the effect of the co…

measure what exists ... you control the independent variable ... •…

Random: each individual has an equal and independant chance o…

Attribute Ranked Measurement

Confound variables

the experiment design should control for the effect of the co…

Mensurative experiment... vs. Manipulativ…

measure what exists ... you control the independent variable ... •…

Nominal

categorical

continuous

Dichotomous

a named category (i.e. names of cities)

ordered categories (presented with median and range)

variables that can take on any value in a given range (presen…

categorical variable that can fall into 2 categories only (i.…

Nominal

a named category (i.e. names of cities)

categorical

ordered categories (presented with median and range)

standard deviation equation

confidence interval

CI significant for difference if

CI significant for odds/risk/hazard ra…

square root of variance

1-alpha or 1- type I error AVOIDING TYPE I ERROR

doesnt include zero

doesnt include 1

standard deviation equation

square root of variance

confidence interval

1-alpha or 1- type I error AVOIDING TYPE I ERROR

mean

median

mode

range

average... sensitive to outliers... can be used when data is not sk…

middle of ranked list... arrange in order, count... less influenced…

most frequent number

difference between highest and lowest

mean

average... sensitive to outliers... can be used when data is not sk…

median

middle of ranked list... arrange in order, count... less influenced…

cross sectional

case control

cohort study

clinical trial

what is happening... assesses **frequency of disease** at particul…

what happened?... **retrospective**... compare people *with dz to tho…

can be prospective or retrospective... asks who will develop, wh…

compares therapeutic beneifts of 2 or more txs or tx vs placebo

cross sectional

what is happening... assesses **frequency of disease** at particul…

case control

what happened?... **retrospective**... compare people *with dz to tho…

Binomial Distribution

Population

Sample

Sampling frame

A random variable has a binomial distribution if it is the nu…

Complete set of entities or elements or units or subjects tha…

A subset of a population

List of items in a population from which a sample is drawn

Binomial Distribution

A random variable has a binomial distribution if it is the nu…

Population

Complete set of entities or elements or units or subjects tha…

OR

RR

AR

RRR

Case-Control: Odds group with disease was exposed to risk fac…

Cohort: Risk of developing disease in exposed group divided b…

Difference in risk between exposed and unexposed groups... AR =…

Proportion of Risk Reduction attributable to intervention... 2%…

OR

Case-Control: Odds group with disease was exposed to risk fac…

RR

Cohort: Risk of developing disease in exposed group divided b…