austin cole february 16, 2010. outline i. sampling a. bad sampling methods b. random sampling ii....

Data for DecisionsChapter 7

Austin ColeFebruary 16, 2010

Outline

I. Samplinga. Bad Sampling Methodsb. Random Sampling

II. ExperimentsIII. Applying Sample to a PopulationIV. SimulationsV. Confidence IntervalsVI. Discussion

Sampling

Population- entire group of individuals about which we want informationSample- part of population from which information is collected

Unemployment

Monthly unemployment rate based on survey of 60,000 householdsDefine populationDefine unemployedFinal percentage

"Labor Force"

UnemployedNot looking for workEmployed

Bad Sampling Methods

Convenience sample-sample of easiest to reach members of populationBias-systematically favoring a certain outcomeVoluntary Response Sample-people choose to respond to a general appeal

Simple Random Sampling

Every individual in population has equal chance to be sampledTable of random

digits

Cautions about Sample Surveys

Undercoverage-group of the population is left out when choosing sampleNonresponse-individual chosen doesn’t participateWording of questions

Experiments

Observational StudyExperiment-imposes some treatment on individuals to observe their responsesConfounding variables-variable whose effects cannot be distinguishedControl group

Randomized Comparative Experiment

Online vs. classroom courses

Random Sampling Exercise

1.Starting on line x, read 2-digit groups until you have chosen 6 restaurants.2.Ignore groups not in the range and ignore any repeated labels.

Starting at line 105: 07, 19, 14, 17, 13, 15

Thinking about Experiments

Placebo effectDouble-blind experimentProspective studies

From Sample to Population

Statistical inference-using fact of a sample to estimate about whole populationParameter-fixed number that describes populationStatistic-number that describes a sampleSampling Distribution-distribution of values taken by the statistic in all possible samples of the same size from the same population

Simulation

Assessing simulations

ShapeCenter-mean of sampling distribution (g)Spread-standard deviation of sampling distribution

g(1- g)n

Confidence Intervals

Percent of all samples will produce an interval containing the true population parameter68-95-99.7 RuleMargin of error for 95% confidence interval:

ĝ(1- ĝ)n2

95% Confidence Interval

Exercise

A Gallup poll asked a random sample of 1785 adults if they attended church or synagogue in the last 7 days. Of the respondents, 750 said yes. Find the 95% confidence interval.

ĝ(1- ĝ)n

ĝ=.42 =.023

95% Confidence Interval: .376 to .466

Discussion

In real world examples, what are some uses of knowing the spread/standard deviation?Other uses/applications for this information?

9,38,44a (7th edition)

Homework Problems: