Let

1

• Lets review some of the statistics youve
  learned in your first class
• Univariate analyses (single variable) are done
  both graphically and numerically. For
  categorical variables, use bar graphs and
  percentages. SAS does this with PROC GCHART and
  PROC FREQ
• PROC GCHART DATApadgett VBAR marsh
• title Vertical Bar Graph
• PROC GCHART DATApadgett HBAR marsh
• title Horizontal Bar Graph
• PROC GCHART DATApadgett HBAR plantht
• title Histogram (sort of)
• choose your own midpoints with the option
  /midpoints a to b by
• PROC FREQ DATApadgett TABLES marsh type flower
  run quit

2

• For quantitative variables, use histograms, stem
  and leaf plots, boxplots (especially for
  comparisons - more later), quantile plots
  (probability plots). SAS does this with PROCs
  UNIVARIATE, MEANS, SUMMARY, TABULATE, REPORT,
  etc. Right now, well concentrate on UNIVARIATE
• PROC UNIVARIATE DATApadgett VAR plantht
  totmass title Lots of Statistics run quit
• PROC UNIVARIATE DATApadgett NORMAL PLOT VAR
  plantht totmass title With the NORMAL and PLOT
  options
• run quit
• Better histogram is done in UNIVARIATE with the
  HISTOGRAM statement
• PROC UNIVARIATE DATApadgett VAR plantht
  totmass
• HISTOGRAM plantht /MIDPOINTS 0 to 150 by 25
  NORMAL
• INSET MEANMean (7.2)
• STDStandard Deviation (7.2)/FontCourier
  
• POSNW HEIGHT3 run quit

3

• UNIVARIATE also gives very nice normal quantile
  and normal probability plots
• PROC UNIVARIATE DATApadgett VAR plantht
  totmass
• QQPLOT totmass PROBPLOT plantht run quit