3. Data Collection
īIn research, statisticians use data in many different
ways.
īData can be used to describe situations.
īData can be collected in a variety of ways, BUT if the
sample data is not collected in an appropriate way,
the data may be so completely useless that no amount
of statistical torturing can salvage them.
4.
5. Basic Methods of Sampling
īRandom Sampling
īSelected by using
chance or random
numbers
īEach individual subject
(human or otherwise)
has an equal chance of
being selected
īExamples:
ī Drawing names from a hat
ī Random Numbers
6. Basic Methods of Sampling
īSystematic Sampling
īSelect a random starting point and then select every kth
subject in the population
īSimple to use so it is used often
7. Basic Methods of Sampling
īConvenience Sampling
īUse subjects that are easily accessible
īExamples:
ī Using family members or students in a classroom
ī Mall shoppers
8. Basic Methods of Sampling
īStratified Sampling
īDivide the population into at least two different groups
with common characteristic(s), then draw SOME subjects
from each group (group is called strata or stratum)
īResults in a more representative sample
9. Basic Methods of Sampling
īCluster Sampling
īDivide the population into
groups (called clusters),
randomly select some of
the groups, and then
collect data from ALL
members of the selected
groups
īUsed extensively by
government and private
research organizations
īExamples:
ī Exit Polls
12. Types of Experiments
īObservational Studies
īThe researcher merely observes what is happening or
what has happened in the past and tries to draw
conclusions based on these observations
īNo interaction with subjects, usually
īNo modifications on subjects
īOccur in natural settings, usually
īCan be expensive and time consuming
īExample:
ī Surveys---telephone, mailed questionnaire, personal
interview
13. More on SurveysTelephone Mailed Questionnaire Personal Interviews
Less costly than personal
interviews
Cover a wider geographic
area than telephone or pi
Provides in-depth
responses
Subjects are more candid
than if face to face
Less expensive than
telephone or pi
Interviewers must be
trained
Challenge---some
subjects do not have
phone, will not answer
when called, or hang up
(refusal to participate)
Subjects remain
anonymous
Most costly of three
Tone of voice of
interviewer may
influence subjectsâ
responses
Challenge âlow number
of subjectsâ respond,
inappropriate answers to
questions, subjects have
difficulty
reading/understanding
the questions
Interviewer may be
biased in his/her
selection of subjects
14. Types of Experiments
īExperimental Studies
īThe researcher manipulates one of the variables and
tries to determine how the manipulation influences
other variables
īInteraction with subject occurs, usually
īModifications on subject occurs
īMay occur in unnatural settings (labs or classrooms)
īExample:
ī Clinical trials of new medications ,treatments, etc.
17. Uses of Statistics
īDescribe data
īCompare two or more data sets
īDetermine if a relationship exists between variables
īTest hypothesis (educated guess)
īMake estimates about population characteristics
īPredict past or future behavior of data
Use of statistics can be impressive to employers.
18. īAlmost all fields of human endeavor benefit from the
application of statistical method; however, the
misuses of statistics are just as abundant, if not more
so!
īâThere are three types of lies---lies, damn lies, and
statisticsâ Benjamin Disraeli
īâFigures donât lie, but liars figureâ
īâStatistics can be used to support anything
---especially statisticiansâ Franklin P. Jones
19. Sources of Misuse
īThere are two main sources of misuse of statistics:
īEvil intent on part of a dishonest researcher
īUnintentional errors (stupidity) on part of a researcher
who does not know any better
20. Misuses of Statistics
īSamples
īVoluntary-response sample (or self-selected sample)
ī One in which the subjects themselves decide whether to be
included---creates built-in bias
ī Telephone call-in polls (radio)
ī Mail-in polls
ī Internet polls
īSmall Samples
ī Too few subjects used
īConvenience
ī Not representative since subjects can be easily accessed
21. Misuses of Statistics
īGraphs
īCan be drawn
inappropriately leading
to false conclusions
ī Watch the âscalesâ
ī Omission of labels or
units on the axes
ī Exaggeration of one-
dimensional increase by
using a two-dimensional
graph
22.
23. Misuses of Statistics
īSurvey Questions
īLoaded Questions---unintentional wording to elicit a
desired response
īOrder of Questions
īNonresponse (Refusal)âsubject refuses to answer
questions
īSelf-Interest ---Sponsor of the survey could enjoy
monetary gains from the results
24. Misuses of Statistics
īMissing Data (Partial Pictures)
īDetached Statistics ---no comparison is made
īPercentages --
īPrecise Numbers
īPeople believe this implies accuracy
īImplied Connections
īCorrelation and Causality âwhen we find a statistical
association between two variables, we cannot conclude
that one of the variables is the cause of (or directly
affects) the other variable