Statistics is...

1

• Statistics is...
• a collection of techniques for planning
  experiments, obtaining data, and then
  organizing, summarizing, presenting, analyzing,
  interpreting, and drawing conclusions based on
  the data... statistics uses data to gain
  understanding, to answer questions
• the science of learning from data

2
The four-step process Define the problem,
usually by asking questions that need
answers. Collect the Data determined to be needed
to answer the question posed Analyze the data
including numerical and graphical
summaries Interpret the results of the analysis,
including communication of the results
Define the problem
Collect the
data
Process of
Statistical
Investigation
Interpret communicate results
Analyze the
data, summarize
3

• Read Chapter 1 and review the examples given
  there and on my website for insight into the
  statistical process.
• HW In your own work, begin to think of your
  research in terms of this four-step processyou
  probably already do!
• Chapter 2 in the textbook addresses the second
  step, Collecting Data and includes
• specifying the objectives of the study
• choosing the variables to be measured
• picking an appropriate design for the survey or
  experiment
• collecting and organizing the data
• Observational study vs. Experiment
• an observational study in contrast to an
  experiment does not involve the imposition of
  treatments by the investigator but only
  observations example of lead levels in children
  of parents who work in a battery factory example
  of political polling

4

• The latter example has data collected through a
  survey and there are several types of survey
  sampling discussed in 2.2
• simple random sampling where each sample of size
  n has the same chance of being sampled as any
  other sample of the same size.
• stratified random sampling where we first divide
  the population into strata and then randomly
  sample within each stratum.
• cluster sampling involves randomly sampling a
  block, apt. building, etc. from a city and then
  sampling every person in the block, building,
  etc. Consider the example of estimating the
  number of violent deaths in Iraq
• systematic sampling - sample every nth element of
  the population
• There are always potential problems with any type
  of survey
• nonresponse
• measurement problems including bad wording of
  questions, bad delivery of questions, bad memory
  of the respondents, etc.
• How do surveys get their data?
• personal interviews (time-consuming, expensive)
• telephone interviews (the most popular way with
  big polling organizations)
• self-administered questionnaires
• direct observation

5

• HW for Tuesday
• Get the textbook, read 1.1-1.5 2.1-2.6. Pay
  special attention to the examples. Well finish
  our discussion of Ch. 2 next time - note there is
  lots of terminology here!
• Get R and SAS and start exploring them a little
• http//www.r-project.org/
• http//www.uncw.edu/itsd/sassetup.htm