using sas/stat: a gentle introduction to some frequently used
TRANSCRIPT
![Page 1: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/1.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
USING SAS/STAT®:
A GENTLE INTRODUCTION TO
SOME FREQUENTLY USED TOOLS
![Page 2: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/2.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
SCENARIO
• You work for a supermarket and the supermarket is offering a new line of
organic products. Management wants to determine which customers are
likely to purchase these products.
• So you decided to send coupons to customers that are in your loyalty program
so you can see which ones buy items from the new organic line.
• You have collected the data and now you need to determine information about
your customers that have bought the organic line items.
• You have Base SAS and SAS/STAT (SAS/Graph is useful too!)
![Page 3: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/3.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
DATA
Variable Description
ID Unique Customer ID
DEMAFFL Affluence grade on a scale from 1 to 30
DEMAGE Age, in years
DEMCLUSTER Type of Residential Neighborhood – 55 levels
DEMCLSUTERGROUP Neighborhood group - 7 levels
GENDER M=Male, F=Female
DEMGREGION Demographic Region
LOYALTYSTATUS Loyalty status: tin, silver, gold, or platinum
PROMSPEND Total amount spent
PROMTIME Time as loyalty card member
TARGETBUY Organics purchased? 1=Yes, 0=No
TARGETAMT Number of organic products purchased during promotion
PREPROMAMT Number of organic products purchased before promotion
DIFF_AMT TARGETAMT - PREPROMAMT
![Page 4: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/4.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
DATA SAS CODE
PROC PRINT DATA=mydata.organics
(OBS=10)
OBS="Row number"
LABEL;
VAR ID DemAffl DemAge DemCluster
DemClusterGroup DemGender DemReg
PromClass PromSpend PromTime
TargetBuy TargetAmt;
RUN;
![Page 5: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/5.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
DATA SAS ENTERPRISE GUIDE
![Page 6: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/6.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
DATA
Observations 22,223
Variables 14
![Page 7: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/7.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
FIRST THINGS
FIRST…CATEGORICAL VARIABLES - SAS CODE
PROC FREQ DATA=mydata.organics;
TABLES TargetBuy TargetAmt PromClass DemGender;
RUN;
![Page 8: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/8.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
FIRST THINGS
FIRST…CATEGORICAL VARIABLES - SAS ENTERPRISE GUIDE
![Page 9: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/9.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
FIRST THINGS
FIRST…CATEGORICAL VARIABLES
![Page 10: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/10.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
FIRST THINGS
FIRST…CATEGORICAL VARIABLES
![Page 11: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/11.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
FIRST THINGS
FIRST…CONTINUOUS VARIABLES – SAS CODE
PROC MEANS DATA=mydata.organics
VARDEF=DF MEAN STD MIN MAX N NMISS;
VAR PromSpend PromTime DemAge;
RUN;
* Use PROC UNIVARIATE to generate the histograms;
TITLE1 "Summary Statistics";
TITLE2 "Histograms";
PROC UNIVARIATE DATA=mydata.organics NOPRINT;
VAR PromSpend PromTime DemAge;
HISTOGRAM ;
RUN;
![Page 12: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/12.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
FIRST THINGS
FIRST…
CONTINUOUS VARIABLES - SAS ENTERPRISE GUIDE
![Page 13: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/13.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
FIRST THINGS
FIRST…CONTINUOUS VARIABLES
![Page 14: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/14.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
FIRST THINGS
FIRST…CONTINUOUS VARIABLES - SAS CODE
PROC UNIVARIATE DATA = mydata.organics
CIBASIC(TYPE=TWOSIDED ALPHA=0.05)MU0=0;
VAR PromSpend DemAge PromTime;
HISTOGRAM PromSpend DemAge PromTime / NORMAL
( W=1 L=1 COLOR=YELLOW MU=EST SIGMA=EST)
CFRAME=GRAY CAXES=BLACK WAXIS=1
CBARLINE=BLACK CFILL=BLUE PFILL=SOLID ;
RUN;
![Page 15: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/15.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
FIRST THINGS
FIRST…
CONTINUOUS VARIABLES - SAS ENTERPRISE GUIDE
![Page 16: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/16.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
A PICTURE IS
WORTH…
![Page 17: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/17.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
A PICTURE IS
WORTH…
![Page 18: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/18.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
IS IT NORMAL?
![Page 19: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/19.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
JUST TO CLEAR
THINGS UP…
• FREQ
• MEANS
• UNIVARIATE
• CORR
• TTEST
• NPAR1WAY
• ANOVA
• REG
• LOGISTIC
![Page 20: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/20.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
ASSOCIATIONS
![Page 21: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/21.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
ASSOCIATION
• An association exists between two variables if the distribution of
one variable changes when the level (or value) of the other
variable changes
• If there is no association, the distribution of the first variable is the
same regardless of the level of the other variable
![Page 22: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/22.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
TESTS OF
ASSOCIATION
• There is no association between
GENDER and TARGETBUY
• The probability of purchasing organic
items is the same whether you are male
or female
• There is an association between
GENDER and TARGETBUY
• The probability of purchasing organic
items is different between males and
females
ALTERNATIVE HYPOTHESISNULL HYPOTHESIS
![Page 23: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/23.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
INITIAL ANALYSIS SAS CODE
PROC FREQ DATA = mydata.organics2
ORDER=INTERNAL;
TABLES Gender * TargetBuy / FORMAT=COMMA8.
NOROW
NOCOL
NOPERCENT
EXPECTED
NOCUM
ALPHA=0.05;
RUN;
![Page 24: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/24.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
INITIAL ANALYSIS SAS ENTERPRISE GUIDE
![Page 25: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/25.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
INITIAL ANALYSIS SAS ENTERPRISE GUIDE
![Page 26: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/26.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CATEGORICAL VARIABLE TEST
![Page 27: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/27.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CHI-SQUARE TEST
• Association
• Observed frequencies ≠ expected
frequencies
• No Association
• Observed frequencies = expected
frequencies
![Page 28: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/28.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CHI-SQUARE TEST
Chi-square tests and the corresponding p-values
• Determine whether an association exists
• Do not measure the strength of an association
• Depend on and reflect the sample size
![Page 29: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/29.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
P-VALUE FOR CHI-
SQUARE TEST
• Probability of observing a chi-square statistic at least as
large as the one actually observed, given that there is not
association between the variables
• Probability of the association you observe in the data
occurring by chance
![Page 30: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/30.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CHI-SQUARE TEST SAS CODE
PROC FREQ DATA = mydata.organics2
ORDER=INTERNAL;
TABLES Gender * TargetBuy / FORMAT=COMMA8.
NOCOL
NOPERCENT
CELLCHI2
EXPECTED
NOCUM
CHISQ
ALPHA=0.05;
RUN;
![Page 31: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/31.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CHI-SQUARE TEST SAS ENTERPRISE GUIDE
![Page 32: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/32.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
ADDING CHI-SQUARE
TO OUR FREQ
OUTPUT
![Page 33: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/33.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
ADDING CHI-SQUARE
TO OUR FREQ
OUTPUT
(CONTINUED)
![Page 34: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/34.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
STRENGTH OF
ASSOCIATION
Cramer’s V Statistic
• -1 to 1 for 2 by 2 tables
• 0 to 1 for larger tables
• Values further away from 0 indicate the presence of a relatively
strong association
![Page 35: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/35.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
ADDING CHI-SQUARE
TO OUR FREQ
OUTPUT
STRENGTH OF ASSOCIATION – CRAMER’S V
![Page 36: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/36.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
WHEN NOT TO USE
CHI-SQUARE
• When more than 20% of cells have expected counts less than five
• In this case use Fisher’s Exact Test
![Page 37: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/37.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
EXAMPLE FOR
FISHER’S EXACT
TEST
![Page 38: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/38.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
FISHER’S EXACT
TEST
• Useful when sample sizes are small (less than 20-25 total)
• 2x2 tables
• Calculates probabilities by considering every possible table where
the marginal (row and column) totals remain fixed)
• Large datasets may require a prohibitive amount of time and
memory for computing exact p-value.
![Page 39: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/39.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
FISHER’S EXACT
TEST HYPOTHESIS
Null Hypothesis: No Association
Alternative Hypothesis:
Two-Tailed
Left-tailed
Right-tailed
![Page 40: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/40.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
FISCHER’S EXACT
TESTSAS CODE
PROC FREQ DATA =mydata.products
ORDER=INTERNAL;
TABLES Product * Purchased /
NOROW
NOPERCENT
CELLCHI2
EXPECTED
NOCUM
FISHER
SCORES=TABLE
ALPHA=0.05;
RUN;
![Page 41: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/41.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
FISHER’S EXACT
TESTSAS ENTERPRISE GUIDE
![Page 42: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/42.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
FISHER’S EXACT
TEST
![Page 43: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/43.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
FISHER’S EXACT
TEST
![Page 44: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/44.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
ORDINAL VARIABLES TEST
![Page 45: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/45.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
MANTEL HAENSZEL
CHI-SQUARE TEST
• Good with Ordinal association
• Means as one variable increases the other variable tends to increase or
decrease
• Need to have one variable with more than 2 levels
• Does not measure strength of association
Null Hypothesis:
There is no ordinal association between the row and column variables
Alternative Hypothesis:
There is an ordinal association between the row and column variables.
![Page 46: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/46.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
MANTEL HAENSZEL
CHI-SQUARE TESTSAS CODE
PROC FREQ DATA = mydata.organics2
ORDER=INTERNAL;
TABLES LoyaltyStatus * TargetBuy /
NOCOL
NOPERCENT
CELLCHI2
EXPECTED
NOCUM
CHISQ
SCORES=TABLE
ALPHA=0.05;
RUN;
![Page 47: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/47.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
MANTEL HAENSZEL
CHI-SQUARE TESTSAS ENTERPRISE GUIDE
![Page 48: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/48.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
MANTEL HAENSZEL
CHI-SQUARE TESTSAS ENTERPRISE GUIDE
![Page 49: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/49.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
MANTEL HAENSZEL
CHI-SQUARE TEST
![Page 50: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/50.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
MANTEL HAENSZEL
CHI-SQUARE TEST
![Page 51: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/51.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
STRENGTH OF
ASSOCIATION
Spearman Correlation Statistic
• Range -1 to 1
• Values close to 1, relatively high degree of positive correlation
• Values close to -1, relatively high degree of negative correlation
• Only valid if both variables are ordinally scaled and in logical order
![Page 52: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/52.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
MANTEL HAENSZEL
CHI-SQUARE TESTWITH SPEARMAN CORRELATION - SAS ENTERPRISE GUIDE
![Page 53: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/53.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
MANTEL HAENSZEL
CHI-SQUARE TEST WITH SPEARMAN CORRELATION
![Page 54: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/54.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CONTINUOUS VARIABLE TEST
![Page 55: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/55.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS
• Describes the relationship between 2 continuous variables
• Important to view data in scatter plot before you start analysis
![Page 56: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/56.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SCATTER PLOT
• 2 dimensional graphs produced by plotting one variable
against another
• Useful to
• Explore the relationship between 2 variables
• Locate outlying or unusual values
• Identify possible trends
• Identify a basic relationship of Y and X values
• Communicate data analysis results
![Page 57: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/57.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SCATTER PLOT
Typical Scatter Plot
![Page 58: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/58.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SCATTER PLOT
PROMSpend & DEMAge
![Page 59: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/59.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SCATTER PLOT CODE
PROMSpend & DEMAge
PROC GPLOT DATA=mydata.organics2;
PLOT PromSpend * DemAge;
RUN;
*Proc GPLOT is in SAS/Graph, you can substitute PROC PLOT
![Page 60: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/60.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SAS ENTERPRISE GUIDE
*Proc GPLOT is in SAS/Graph
Graph
![Page 61: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/61.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SAS ENTERPRISE GUIDE
*Proc GPLOT is in SAS/Graph
Graph
![Page 62: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/62.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SCATTER PLOT
Image Source: eMathZone.com
![Page 63: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/63.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS PEARSON CORRELATION COEFFICIENT
• Between -1 and 1
• Closer to either extreme, high degree of linear
association between the two variables
• Close to 0, no linear association
• Greater than 0, positive linear association
• Less than 0, negative linear association
![Page 64: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/64.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS HYPOTHESIS TEST
• The parameter representing correlation is ρ
• ρ is estimated by the sample statistic r
Null Hypothesis:
There is no association between the 2 variables, ρ = 0
Alternative Hypothesis:
There is an association between the 2 variables, ρ ≠ 0
• Rejecting H0 indicates only great confidence that ρ is not exactly 0
• A p-value does not measure the magnitude of the association
• Sample size affects the p-value
![Page 65: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/65.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SAS CODE
PROC CORR DATA=mydata.organics
PEARSON;
VAR DemAge PromSpend;
RUN;
![Page 66: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/66.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SAS ENTERPRISE GUIDE
![Page 67: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/67.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SAS ENTERPRISE GUIDE
![Page 68: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/68.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SAS ENTERPRISE GUIDE
![Page 69: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/69.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SAS ENTERPRISE GUIDE
![Page 70: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/70.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS HYPOTHESIS TEST
![Page 71: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/71.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SAS CODE WITH SCATTER PLOT
PROC CORR DATA=mydata.organics
PLOTS=(SCATTER MATRIX)
PEARSON
/* Start of custom user code. */
PLOTS(MAXPOINTS=30000 )
/* End of custom user code. */
;
VAR DemAge PromSpend;
RUN;
![Page 72: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/72.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SAS ENTERPRISE GUIDE
![Page 73: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/73.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SAS ENTERPRISE GUIDE
![Page 74: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/74.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SAS ENTERPRISE GUIDE
![Page 75: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/75.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SAS CODE WITH SCATTER PLOT
![Page 76: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/76.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SCATTER PLOT MATRIXGraph
![Page 77: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/77.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SCATTER PLOT MATRIX CODE
PROC SGSCATTER DATA=mydata.ORGANICS;
TITLE "Scatter Plot Matrix";
MATRIX DemAffl DemAge PromSpend PromTime
TargetBuy TargetAmt/
START=TOPLEFT
ELLIPSE=(ALPHA=0.05 TYPE=PREDICTED)
NOLEGEND;
RUN;
*Proc SGSCATTER is in SAS/Graph.
Graph
![Page 78: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/78.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SAS ENTERPRISE GUIDE
*Proc SGSCATTER is in SAS/Graph
Graph
![Page 79: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/79.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SAS ENTERPRISE GUIDE
*Proc SGSCATTER is in SAS/Graph
Graph
![Page 80: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/80.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SAS CODE WITH SCATTER PLOT MATRIX
PROC CORR DATA=mydata.organics
PLOTS=(SCATTER MATRIX)
PEARSON
/* Start of custom user code. */
PLOTS(MAXPOINTS=1000000 )
/* End of custom user code. */
;
VAR PromTime DemAge TargetAmt PromSpend
DemAffl;
RUN;
![Page 81: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/81.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATIONS SAS CODE WITH SCATTER PLOT MATRIX
![Page 82: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/82.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATION VS CAUSATION
![Page 83: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/83.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CORRELATION VS CAUSATION
![Page 84: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/84.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
CATEGORICAL AND CONTINUOUS VARIABLES
![Page 85: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/85.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST
Three types
1. One Sample
2. Two Sample
3. Paired
![Page 86: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/86.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST ONE SAMPLE
Parametric test to compare sample mean with known
value
• Null Hypothesis: H0: µ = hypothesized value
• Alternative Hypothesis: Ha: µ ≠ hypothesized value
For our Example µ = 47Assumptions
• The data consist of independently chosen random samples
• The sample size is large
![Page 87: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/87.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST ONE SAMPLE
PROC TTEST
DATA = mydata.organics
PLOTS(ONLY)=SUMMARY
ALPHA=0.05
H0 =47
CI = EQUAL;
VAR DemAge;
BY TargetBuy;
RUN;
Can also calculated in PROC UNIVARIATE
![Page 88: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/88.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST ONE SAMPLE - SAS ENTERPRISE GUIDE
![Page 89: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/89.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST ONE SAMPLE - SAS ENTERPRISE GUIDE
![Page 90: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/90.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST ONE SAMPLE
![Page 91: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/91.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST ONE SAMPLE
![Page 92: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/92.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST TWO SAMPLE
Parametric test to compare two independent samples
• Null Hypothesis: H0: µ1 = µ2
• Alternative Hypothesis: Ha: µ1 ≠ µ2
For our Example
µ1 is amount Females spent during the promotion
µ2 is amount Males spent during the promotion
Assumptions
• Independent Observations
• Normally distributed responses for each group
• Equal variances for each group
![Page 93: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/93.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST TWO SAMPLE
PROC TTEST
DATA = mydata.organics
PLOTS(ONLY)=SUMMARY
ALPHA=0.05
H0 =0
CI = EQUAL;
CLASS Gender;
VAR PromSpend;
RUN;
![Page 94: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/94.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST TWO SAMPLE - SAS ENTERPRISE GUIDE
![Page 95: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/95.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST TWO SAMPLE - SAS ENTERPRISE GUIDE
![Page 96: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/96.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST TWO SAMPLE
Descriptive
Statistics
Equality of
Variance
Go with Unequal
Variance Test
Go with Unequal
Variance Test
![Page 97: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/97.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST TWO SAMPLE
![Page 98: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/98.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST PAIRED
Parametric test to compare repeat measures on the same subject
• Null Hypothesis: H0: µpost = µpre
• Alternative Hypothesis: Ha: µpost ≠ µpre
For our Example
µpost is amount of organic items bought during promotion
µpre is amount of organic items bought before promotion
Assumptions
• The subjects are selected randomly
• The distribution of the sample mean differences is normal
![Page 99: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/99.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST PAIRED
PROC TTEST
DATA = mydata.organics
PLOTS(ONLY)=SUMMARY
ALPHA=0.05
H0 =0
CI = EQUAL;
PAIRED TargetAmt * PrePromAmt;
RUN;
![Page 100: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/100.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST PAIRED - SAS ENTERPRISE GUIDE
![Page 101: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/101.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST PAIRED - SAS ENTERPRISE GUIDE
![Page 102: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/102.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST PAIRED
![Page 103: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/103.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
T-TEST PAIRED
![Page 104: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/104.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
NONPARAMETRIC TEST
![Page 105: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/105.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
NONPARAMETRIC
ANALYSIS
Nonparametric analysis are those that rely only on the
assumption that the observations are independent
A nonparametric test is appropriate when
• The data contains valid outliers
• The data is skewed
• The response variable is ordinal and not contiguous
![Page 106: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/106.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
NONPARAMETRIC
ANALYSIS PROC NPAR1WAY
• The rank of each data point is used instead of the raw data
• Rank from smallest to largest
• In the event of a tie the ranks are averaged
• For 2 level variables – Wilcoxon rank-sum test is used
• For more than 2 levels – Kruskal-Wallis test is used
![Page 107: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/107.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
PROC NPAR1WAY
Null Hypothesis: H0: all populations are identical with
respect to scale, shape, and location
Alternative Hypothesis: Ha: all populations are not
identical with respect to scale, shape, and location
• Only assumption is that you have independent
observations
• Used with ordinal, interval and ratio measurement
variables
![Page 108: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/108.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
PROC NPAR1WAY 2 LEVELS
PROC NPAR1WAY DATA=organics2 WILCOXON MEDIAN;
VAR Diff_Amt;
CLASS Gender;
RUN;
![Page 109: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/109.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
PROC NPAR1WAY 2 LEVELS - SAS ENTERPRISE GUIDE
![Page 110: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/110.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
PROC NPAR1WAY 2 LEVEL - SAS ENTERPRISE GUIDE
![Page 111: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/111.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
PROC NPAR1WAY 2 LEVELS
![Page 112: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/112.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
PROC NPAR1WAY 2 LEVELS
![Page 113: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/113.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
PROC NPAR1WAY > 2 LEVELS
proc sort data=mydata.organics2 out=organics2; by
loyaltyStatus;
PROC NPAR1WAY DATA=organics2 WILCOXON MEDIAN;
VAR Diff_Amt;
CLASS LoyaltyStatus;
RUN;
![Page 114: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/114.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
PROC NPAR1WAY > 2 LEVELS - SAS ENTERPRISE GUIDE
![Page 115: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/115.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
PROC NPAR1WAY > 2 LEVEL - SAS ENTERPRISE GUIDE
![Page 116: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/116.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
PROC NPAR1WAY > 2 LEVELS
![Page 117: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/117.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
PROC NPAR1WAY > 2 LEVELS
![Page 118: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/118.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
RESOURCES
![Page 119: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/119.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
RESOURCES
Public SAS Courses
• Statistics 1: Introduction to ANOVA, Regression, and Logistic Regression FREE
• SAS Enterprise Guide: ANOVA, Regression, and Logistic Regression
Online Tutorials
• SAS Online Resources for Statistics Education
• t-tests
• Tests of Association
• Pearson Chi-Square
• Mantel-Haenszel Chi-Square
• Nonparametric Analysis
![Page 120: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/120.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
RESOURCES
• What’s New Documentation• http://support.sas.com/documentation/whatsnew/
• SAS/STAT Newsletter• http://support.sas.com/community/newsletters/index.html
• STAT, IML, OR, ETS Papers
• Statistical Procedures SAS Community
• Frequently Asked-for Statistics
Videos• Youtube.com
http://www.youtube.com/playlist?list=PL0B05D53A5E101AA6
• Video portal to the STAT and OR focus area.
http://support.sas.com/rnd/app/video/index.html
![Page 121: Using SAS/STAT: A Gentle Introduction to Some Frequently Used](https://reader036.vdocuments.net/reader036/viewer/2022062600/585969ff1a28ab6e328f852f/html5/thumbnails/121.jpg)
C op yr i g h t © 2015 , SAS Ins t i t u te Inc . A l l r i g h ts r eser v ed .
QUESTIONS?
Thank you for your time and attention!
Connect with me:
LinkedIn: https://www.linkedin.com/in/melodierush
Twitter: @Melodie_Rush