Ch 1 - Part 2: Sampling

What are the 2 main branches of statistics?

- Descriptive - inferential

What are some acronyms/terms from medical trials?

- IRB: Internal Review Board - IEC: Independent Ethics Committee - ERB: Ethical Review Board - Informed Consent: requires that the individual be informed and gives consent to participate

What are common observational studies?

- Retrospective (historical data - past) - cross sectional (data from a single point in time - present) - prospective aka longitudinal (data gathered over an extended period of time (future)

What are some pitfalls of sampling?

- Sampling error - hidden bias - survey error: leading questions - under-coverage: when some members of the population are inadequately represented in the sample - incorrect arithmetic - implausible numbers: processing errors (entering, etc) - inconsistencies - missing information

What are examples of cluster sampling?

- Selecting a region in a country - Selecting a zip code in a state

What is simple random sampling?

- each individual has an equally likely chance of being selected - all groups of size n have an equally likely chance of being selected

What is inferential statistics?

includes methods of using information from a sample to make inferences/conclusions about the population

What is descriptive statistics?

involves methods of organizing/summarizing information from a sample or population

What is an experimental study?

Designed to be observed with interference from the observer in that specific treatments are applied to the individuals, in an effort to measure differences in the subject response - treatment = any condition set forth that is applied to the individual or subject in an effort to determine differences compared to each other or a control group

What is an observational study?

Designed to observe without interference from the observer - every effort is made not to sway the subject response or lead a subject in their response

EXAMPLE ON RANDOM DIGIT CHART ON SLIDE 33

EXAMPLE OF RANDOM DIGIT CHART ON SLIDE 33

EXAMPLE OF SAMPLING/RESAMPLING ON SLIDE 34

EXAMPLE OF SAMPLING/RESAMPLING ON SLIDE 34

What is a control group?

Group created for the sake of comparison - can be a treatment group or a group that receives a false treatment (placebo)

What is a double blind experiment?

neither the individual subjects nor the researcher are aware of who received what treatment

What is "randomization"?

selected at random

What is "enough information"?

More important in inferential statistics and not so much in descriptive statistics

What are the 2 different types of cause-and-effect studies?

Observational & experimental

SEE SLIDE 40 FOR COMMON PARAMETERS

SEE SLIDE 40 FOR COMMON PARAMETERS

What is simulation?

The imitation of a natural process using general characteristics or behaviors in an effort to mimic or model the natural system

What is an example of systematic sampling?

The population has 27 people and every third person in the population is selected, making the sample size 9 people

What's an example of a simulation?

Using dice to instead of flipping a coin - even numbers are head - odd numbers are tails

What is an example of convince sampling?

Volunteer response

What is random sampling?

When a sample size of n is taken from a population size of N in such a way that each individual has an equally likely chance of being selected - literally just random

What is the placebo effect?

When a subject receives a false treatment or no treatment, but believes he or she is receiving a true treatment, and responds favorably

What is cluster sampling?

When groups are selected based on pre-existing clusters that is arbitrary to the individual and not based on any characteristic of the individual - select a cluster and then gather information of every individual in that cluster - using clusters of individuals that are pre-existing

What is a (single) blind experiment?

When individual subjects don't know the treatment they receive - the researcher is aware

What is stratified sampling?

When individuals are first grouped by specific characteristics such as gender, age, or race and then samples are taken from each group or strata - using "strata" of individuals selected by a specific characteristic

What is convenience sampling?

When individuals are selected based upon ease of access - Prone to bias - (volunteer response) using individuals who are conveniently surveyed

What is multistage sampling?

When more than one sampling technique is employed in the gathering of information - more than one stage of sampling done in succession

What is anonymity?

When no personal information is taken - nothing in a survey can identify a person

What is a block?

a group of individuals stratified based on a similar characteristic and given treatments - in experimental design

What is a parameter?

a numerical measure that describes the characteristic of the population, such as a central tendency, spread, and shape - when a specific parameter is not stated, the lowercase theta is used

What is a statistic?

a numerical measure that eyelids an estimate of a population parameter using sample data - a function of the sample that does not depend on the parameter

What are the 2 types of measure?

discrete & continuous

What are the different types of samples that you can take?

simple random sampling, systematic sampling, cluster sampling, stratified sampling, convenience sampling, multistage sampling

What is statistics?

the study of how to gather, organize, analyze, and interpret information

What is a random digit chart?

the table of digits selected at random and placed in a table in Appendix B which can be used to simulate or sample data

What is "replication"?

to verify validity and reliability

What is confidentiality?

when personal information is given, but not shared - only the statistical summaries are made available to other organizations or persons involved in the study

What is a continuous measure?

when the set of possible outcomes are infinite and uncountable - dense (between any 2 values, there exists another value, like mean or rate) - Ex: length, height, age

What is a discrete measure?

when the set of possible outcomes are separate, distinct, and finite, such as a count - can be enumerated - ex: number, count, frequency

What is "control"?

A comparative or control group

What is systematic sampling?

A sample such that every kth individual or item is selected - need some sort of list - evenly samples the whole population - using a system to select the individuals

What is block design?

An experimental design in which individuals or subjects are grouped into categories or blocks and then test blocks are treated as experimental units