SAS Functions Flashcards
PROC FREQ
o Frequencies of the variable – how many observations in that group (percentage)
o If select 2 variables to look at can see the cross tabulation – frequency table
o CHI SQUARE
• Has option to include that generates chi-square values
• Ex:
o AGREMENT
• Has option to include /agree for agreement (Kappa)
PROC UNIVARIATE
o Use to explore dataset
o Look at each variable separeately and see the distribuition, summary stats, tests for location , quantiles, extreme observations, plots, can see skewness of data and select to see histogram
PROC TTEST
o Summary statistics (N, mean, standard deviation, standard error, min, max)
o Confidence estimates for the mean and the standard deviations
o T-test results for Pooled (equal variances) and Satterthwaite (when unequal variances)
o Folded F to see Equality of Variances
o One sample t-test/matched pairs has same except equality of variances
o Ex: see if pain rating is different in those who got epidural vs those who didn’t get epidural
PROC ANOVA
o Shows number of observations, anova table, R square, F-value (variation between groups/variation within groups… want 1)
o Doesn’t show which groups are different, just if one of the groups are differnet
o Then have to do multiple comparisons and can do Tukey to check which groups are differnet
o Ex: check if pain rating index of childbirth depends on which age group you are in.
o Ex: check if pain rating index depends on the type of pain killer method used
PROC CONTENTS
o Shows what is in dataset – variable, type, length
PROC CORR
o Shows simple statistics and then Pearson correlation coefficients. Can specify to see Spearman if data not linear
o Can also look at cross correlation matrix
PROC PLOT
Scatterplot
PROC REG
o Use parameter estimates to explain variation in data
o Example: alcohol consumption on life expectancy
o Shows analysis of variance, parameter estiamtes, and fit statistics
