STAT-S 300 Lecture Notes - Lecture 7: Non-Sampling Error, Graphing Calculator, Sampling Frame
Section 4.1 Notes- Sampling and Surveys 8-19-13
• Population- group of individuals about which we want information
• Sample- part of population (pop) from which we collect information- we use info to draw
conclusions about entire pop
o If everyone in pop cannot be sampled you must rethink who/what pop really is
• Sample study- study that uses organized plan to choose sample that represents some
specific pop
• 6 Types of Samples
o How to Sample Well
▪ Simple Random Sample (SRS)- every individual has = chance of being
selected
• Label a list
• Exclude repeats and numbers outside range
• Use randInt. (graphing calculator- math→PRB→5) or Table D
(back of textbook)
o Table of random digits- long string of digits 0-9 w/
following properties:
▪ Each entry in table is equally likely to be any digit
0-9
▪ Entries are independent of each other- no one part
gives info about any other part
▪ Table D
• Give each member of pop a numerical label
of same length
• Read consecutive groups of digits of
appropriate length from table
• Sample contains individuals whose labels
are found, excluding repeats and out-of-
range values
• Chosen by chance rules to prevent favoritism by sampler and self-
selection by respondents
▪ Stratified Sample- pop is divided into strata (groups) based on a
similarity, then SRS taken from each stratum
• Gives more precise info than SRS- reduces variability
• Base strata on information known before sample is taken
▪ Cluster Sample- pop is divided into groups not based on similarity and
SRS taken from each cluster
• Often more convenient than stratified or SRS
• Clusters should mirror traits of pop
▪ Systematic Sample- 1st individual is selected at random, rest of sample is
selected using a fixed period interval
o How to Sample Badly
▪ Voluntary Response- people choose themselves whether or not to
respond
find more resources at oneclass.com
find more resources at oneclass.com