Chapter 5 Section 1
Producing Data
Collecting and using sample data wisely
Descriptive Statistics, use of numerical and graphical summaries to characterize a dataset.
Inferential Statistics, use of sample data to make
conclusions about a broader range of individuals than just those who are observed
Statistical Inference, producing answers to a specific question with a
know degree of confidence on correctness.
Fundamental Rule For Using Data For Inference, the
available data can be used to make inferences about a much larger group…if the data can be considered to be
Chapter 5 Section 1
Observational Study & Experiments & Simulations
Observational Study, the researcher simply observes
individuals and measures variables of interest BUT does not attempt to influence the responses
Example…opinions, behaviors or outcomes ( sources of data about
topics)
Observational Studies are a poor way to gauge the effect of an intervention. In order to see a response change, the researchers must impose change. When this goal becomes to understand a cause- and –
effect relationship (the change), a researcher should use an experiment.
Observational studies of the effect of one variable on another often
Chapter 5 Section 1
Observational Study & Experiments & Simulations
Experiment, the researchers manipulate a something and measure the effect of the manipulation on some outcome of interest
Randomized experiments…are were participatians are randomly assigned to engage in one treatment or another.
Designed to pin down the effects of a specific variable of interest Preferred method of science, medicine and industry
Simulations, alternative method of producing data
Used only when…
I. It is not possible to observe individuals directly or to perform an experiement
Who do we study?
Who Do Researchers Study
Population, Entire group of individuals that the researcher wants information about
All NC High School students enrolled in AP Statistics
A census occurs when we attempt to contact the entire population
Example: if we want to know the average GPA of AP Stats
students we would have to collect that data for every HS student currently enrolled in the class
Who Do Researchers Study
Sample,
PART
of the population that the researcher actually measured or surveyed Select specific NC High School students enrolled in AP Stats
In most cases it is impossible or improbable to study the population which is why we generally look at only a small sample size
When study only a part of a population this is called
Sampling.
The GOAL of choosing a sample is a picture of the
population, disturbed as little as possible by the act of gathering information
Population
Sample
Collect data from a
representative Sample...
Sampling
Note: When making conclusions from sampling a sound design for selecting the sample is required
Design, method used to choose the sample from the population
Poor designs lead to misleading data
Use a sample size LARGE enough so that we can see the true
nature of any effects.
Sampling
Voluntary Response Sample, People who choose themselves by responding to a general appeal (People make the choose to respond to the survey)
Problems: • Biased
• Strong (usually negative) opinions are most likely to respond
Selected (convenience) Sampling: People are selected for the sample based upon the discretion of the person
creating the sample
Bias, when the method used to obtain results in systemic favor a particular outcome (either consistently too high or consistently to low)
The goal of every sampling technique/survey Unbiased,
without systematic favor or weight
Common types of Biased…
1. Under-coverage/Selection bias, when the method for selecting the participants in a sample do not represent the population of interest. Somebody was LEFT OUT of the
sample
Mall bias – if shoppers at a mall are surveyed to determine
attitudes about raising the sales tax, the results are not likely to represent all area residents
2. Nonresponse/Nonparticipation bias, when a representative sample is chosen for a survey but a subset cannot be contacted or does not respond
Phone bias – phoning people at home in the evening would omit
people who work during that time. Their opinions would not be represented
3. Response Bias, when the responses of those surveyed are swayed in way direction or the other
Wording of the question Order of choices
Demeanor or appearance of the interviewer Honesty
•
LARGER RANDOM SAMPLES GIVE MORE
ACCURATE RESUSLTS THAN SMALLER
SAMPLES
Sampling: How we select people for a survey
1. Simple Random Sample (SRS)
People are selected at random (draw names, computer
program, random digits table)
Gives each individual an equal chance but also gives every
possible sample an equal chance to be chosen
These other three random sampling designs give each individual an equal chance but not each sample
2. Stratified Sample
People are put into groups of similar characteristics (Age,
Gender, Ethnicity, Income…) known as Strata
THEN choose a separate SRS in each stratum and combine
the SRS’s to form a full sample
Sampling: How we select people for a survey
3. Systematic Sample
Every nth person is selected
4. Clustered Sample
People are grouped by convenience (known as clusters) and then randomly select some of the clusters for your sample.
Typically people are grouped by geographic regions
Multistage Sampling Design
Choosing smaller samples in stages resulting in a sample
consisting of clusters of individuals. Each stage may include a SRS, stratified sample, or another type of sample
Example: when testing a shipment of goods I might randomly
select one box of items and then from that box randomly select which item will be inspected
Or:
1. The US is divided in 2000+ Geographic Areas
2. Each area is divided into small “neighborhood” areas stratified using ethnic or other information
3. Each neighborhood is clustered into groups of
households that can be easily reached for an opinion poll or survey