Section 2-1 Samples, Good and Bad
Jan 01, 2016
Section 2-1Samples, Good and Bad
Remember: We select a sample in order to get information about some population (entire group of individuals about which we want information)
Read Pg. 63 Town Talk: Why isn’t this a good way to get a sample?
Bad SamplesBias in sampling is bad
The design of a statistical study is biased if it systematically favors certain outcomes
Biased sampling generally results in samples that are not representative of their corresponding populations
Voluntary Response Sample People choose whether to respond.
Attracts people who feel strongly about an issue in question.
It may not represent the opinions of the entire population.
Bias refers to the tendency of a sample statistic to systematically over- or under-estimate a population
Voluntary Response - ExampleWrite in or call in opinion polls—only about
15% of the public has ever responded to a call-in poll which is not a representation of the population as a whole
Talk Town –allowed people to call in rather than actively selecting its own sample therefore the result was biased - the sample was over weighted with people favoring the ambulance monopoly.
Convenience SampleThe selection of whichever individuals are
easiest to reachInterviewer chooses samplePeople tend to pick neat, safe-looking
individuals to selectSometimes a convenience sample may be
drawn from telephone directories and car registration lists. In 1936, people who owned cars and telephones tended to be more affluent. Undercoverage is often a problem with convenience samples.
Convenience Sample - ExampleInterviewing your friends to find out what
school lunch entre is preferred, mall interviews may only target the rich/teenagers/retired people, people tend to pick neat, safe-looking individuals from the stream of customers
Nonresponse BiasSometimes individuals chosen for the sample
are unwilling or unable to participate in the survey.
The bias that results when respondents differ in meaningful ways from nonrespondents.
Since only 25% of the sampled voters actually completed the mail-in survey, survey results overestimated voter support for….
Good SamplesRandom sampling is a procedure for sampling
from a population in which:The selection of a sample unit is based on chance.Every element of the population has a known,
non-zero probability of being selected. Random sampling helps produce representative
samples by eliminating voluntary response bias and guarding against undercoverage bias.
All probability sampling methods rely on random sampling.
Simple Random Sample (SRS)Consists of a group individuals from the
population chosen in such a way that every individual and every mixture group has an equal chance to be in the sample actually chosen.
It is a sample chosen by chance which avoiding bias.
Example – SRSIt is like putting names in a hat (population)
and drawing out a handful (sample)Write 100 names and put in a hat and pick
10…this is an SRS because any ten slips have the same chance of being chosen.
However, using a hat is often impractical so we use computer generated random digits to choose samples.
Random Digits TableA table of random digits is a long string of
the digits 0,1…9
Each entry is equally likely to be any of the 10 digits.
Entries are independent of one another.
Knowledge of 1 part of the table gives no information about any other part
How to Choose an SRSLABEL - assign a numerical label to every
individualYour goal is to use the shortest possible labels1 digit for population up to 10 members (1-9)2 digits for population for 11-100 members (01-99)3 digits for population for 101-1000 members
etc…
(001-999) Recommend starting with 1, 01, or 001 (but
can start with 0, 00 or 000)
Select 5 Students from Statistics Class to be part of a sample surveyStep 1: Label 18 students in alphabetical order.
Then use 2 digits 01, 02,…19 (can use 00,…18)Will depend upon the size of the population
Step 2: Use Table - can enter the table anywhereUse line 104
52, 71, 13, 88, 89, 93, 07, 46, 02, 27, 40, 01, 18
Answer: Sample is composed of students numbered 13, 07, 02, 01 and 18.
Is this an SRS?Assume there are 10 boys and 10 girls
Flip a coin: heads gives all girls, tails gives all boys
Is this an SRS?
No, because we can never get a mix of both girls and boys, so there is a 0% chance of ever getting a mixture of boys and girls.
Classwork/Homework1. Pg. 67-68 #1-4
3. Pg. 74-75 # 7-11