Scripts - Parameters for statistical procedures and graphs

Age-related reference interval

PROC	REFINTERVALAGE
$Variable	Continuous variable
$AgeVariable	Age variable
$Filter	Filter
$Centiles	{1,2.5,5,10,90,95,97.5,99}
$PowersMean	{-2,-1,-0.5,0,0.5,1,2,3}
$PowersSD	{-2,-1,-0.5,0,0.5,1,2,3}
$Log	TRUE \| FALSE
$BoxCoxTransform	TRUE \| FALSE
$Lambda	Real
$Shift	Real
$RefIntervalOutlierTest	NONE \| REED \| TUKEY
$NormalityTest	KOLSMIR \| DAGPEAR \| SHAPWILK \| SHAPFRANC \| CHISQ \| NONE

For details, see Age-related reference interval

Analysis of Covariance (ANCOVA)

PROC	ANCOVA
$Variable	Variable
$Factor1	Variable
$Factor2	Variable
$Filter	Filter
$Covariates[1..24]	Variable(s)
$NormalityTest	KOLSMIR \| DAGPEAR \| SHAPWILK \| SHAPFRANC \| CHISQ \| NONE

For details, see Analysis of covariance

Bland-Altman plot

PROC	BLANDALTMAN
$Variable1	Variable
$Variable2	Variable
$Filter	Filter
$Subgroups	Variable
$BAxAxis	AVERAGE \| METHOD1 \| METHOD2 \| GEOMETRICMEAN \| AVGRANK \| RANK1 \| RANK2
$BAMethod	DIFFERENCES \| DIFFERENCES% \| RATIOS
$MaxDifference	Real
$BALineOfEquality	TRUE \| FALSE
$CIMean	TRUE \| FALSE
$CILimitsOfAgreement	TRUE \| FALSE
$CIErrorBars	TRUE \| FALSE
$BARegression	TRUE \| FALSE
$BARegressionCI	TRUE \| FALSE

For details, see Bland-Altman plot

Bland-Altman plot with multiple measurements per subject

$Variable1	Variable
$Variable2	Variable
$Factor	Variable (subject identification)
$Filter	Filter
$TrueValueConstant	TRUE \| FALSE
$BAxAxis	AVERAGE \| METHOD1 \| METHOD2 \| GEOMETRICMEAN \| AVGRANK \| RANK1 \| RANK2
$BAMethod	DIFFERENCES \| RATIOS
$MaxDifference	Real
$BALineOfEquality	TRUE \| FALSE
$CILimitsOfAgreement	TRUE \| FALSE
$CIErrorBars	TRUE \| FALSE
PROC	BLANDALTMANMULTI

For details, see Bland-Altman plot with multiple measurements per subject

Box-and-Whisker plot

PROC	BOXWHISKER
$Variable	Variable
$Filter	Filter
$Log	TRUE \| FALSE
$PlotAllPoints	TRUE \| FALSE
$Vertical	TRUE \| FALSE

For details, see Box-and-whisker plot

Chi-Squared test

PROC	CHI2TEST
$Factor1	Variable
$Factor2	Variable
$Filter	Filter

For details, see Chi-squared test

Clustered multiple comparison graphs

PROC	GRAPHMULTCOMPCLUST
$Variable	Variable
$Factor1	Variable
$Factor2	Variable
$Filter	Filter
$Bars	TRUE \| FALSE
$MidLines	TRUE \| FALSE
$MidMarkers	TRUE \| FALSE
$ConnectingLines	TRUE \| FALSE
$MeanOrMedian	MEAN \| MEDIAN
$ErrorBarsMean	NONE \| MEANCI \| 1SD \| 2SD \| 3SD \| SEM \| RANGE
$ErrorBarsMedian	$ErrorBarsMedian	NONE \| MEDIANCI \| PR50 \| PR80 \| PR90 \| PR95 \| PR98 \| RANGE (*)
$BoxWhisker	TRUE \| FALSE
$NotchedBW	TRUE \| FALSE
$ViolinPlot	TRUE \| FALSE
$PlotAllPoints	TRUE \| FALSE
$Log	TRUE \| FALSE

For details, see Clustered multiple comparison graphs

(*) Note: PR50 is 50% percentile range (25-75 percentiles; Interquartile range); PR80 is 80% percentile range (10-90 percentile), etc.

Clustered multiple variables graphs

PROC	GRAPHMULTIVARCLUST
$Covariates[1..24]	Variable(s)
$Factor	Variable
$Filter	Filter
$Bars	TRUE \| FALSE
$MidLines	TRUE \| FALSE
$MidMarkers	TRUE \| FALSE
$ConnectingLines	TRUE \| FALSE
$MeanOrMedian	MEAN \| MEDIAN
$ErrorBarsMean	NONE \| MEANCI \| 1SD \| 2SD \| 3SD \| SEM \| RANGE
$ErrorBarsMedian	$ErrorBarsMedian	NONE \| MEDIANCI \| PR50 \| PR80 \| PR90 \| PR95 \| PR98 \| RANGE (*)
$BoxWhisker	TRUE \| FALSE
$NotchedBW	TRUE \| FALSE
$ViolinPlot	TRUE \| FALSE
$PlotAllPoints	TRUE \| FALSE
$DotAndLines	TRUE \| FALSE
$Log	TRUE \| FALSE
$CompleteCasesOnly	TRUE \| FALSE
$ClusteredByVariables	TRUE \| FALSE

For details, see Clustered multiple variables graphs

(*) Note: PR50 is 50% percentile range (25-75 percentiles; Interquartile range); PR80 is 80% percentile range (10-90 percentile), etc.

Cochran–Mantel–Haenszel test

PROC	CMHTEST
$Classifier	Dichotomous outcome variable
$Groups	Group variable
$Factor	Factor variable
$Filter	Filter

For details, see Cochran–Mantel–Haenszel test

Cochran's Q test

PROC	COCHRANQ
$Covariates[1..24]	Variable(s)
$Filter	Filter
$Alpha	Real

For details, see Cochran's Q test

Coefficient of variation from duplicate measurements

PROC	CVFROMDUPLICATES
$VariableX	Variable
$VariableY	Variable
$Filter	Filter
$CVMethod	WSSD \| LOG \| RMS

For details, see Coefficient of variation from duplicate measurements

Comparison of independent ROC Curves

PROC	COMPAREINDEPENDENTROC
$Variable	Variable
$Groups	Group variable
$Classifier	Dichotomous variable
$Filter	Filter
$ROCMethod	DELONG \| HANLEY
$BinomialCI	TRUE \| FALSE
$ShowGraph	TRUE \| FALSE
$MarkPoints	TRUE \| FALSE

For details, see Comparison of independent ROC curves

Comparison of ROC Curves

PROC	COMPAREROC
$Covariates[1..6]	Variables
$Classifier	Dichotomous variable
$Filter	Filter
$ROCMethod	DELONG \| HANLEY
$BinomialCI	TRUE \| FALSE
$ShowGraph	TRUE \| FALSE
$MarkPoints	TRUE \| FALSE

For details, see Comparison of ROC curves

Comparison of partial areas under ROC curves

PROC	PARTIALROCCOMPARE
$Variable1	Variable
$Variable2	Variable
$Classifier	Dichotomous variable
$Filter	Filter
$Paired	TRUE \| FALSE
$PartialInterval	SENSITIVITY \| SPECIFICITY
$ValLow	Real (0..100)
$ValHigh	Real (0..100)
$Bootstrap	TRUE \| FALSE
$BootstrapLoops	Integer
$RandomNrSeed	Integer
$ShowGraph	TRUE \| FALSE

For details, see Comparison of partial areas under ROC curves

Comparison of precision-recall curves

PROC	PRECISIONRECALLCOMPARE
$Variable1	Variable
$Variable2	Variable
$Classifier	Dichotomous variable
$Filter	Filter
$Paired	TRUE \| FALSE
$Bootstrap	TRUE \| FALSE
$BootstrapLoops	Integer
$RandomNrSeed	Integer
$MarkPoints	TRUE \| FALSE

For details, see Comparison of precision-recall curves

Concordance correlation coefficient

PROC	CONCORDANCE
$VariableX	Variable
$VariableY	Variable
$Filter	Filter
$LogX	TRUE \| FALSE
$LogY	TRUE \| FALSE

For details, see Concordance correlation coefficient

Control Chart

PROC	CONTROLCHART
$Variable	Variable
$Filter	Filter
$Labels	Variable
$CtrlType	DATA \| STANDARD \| USER
$CtrlN	Integer
$CtrlMean	Real
$CtrlSD	Real
$CtrlUClim	Real
$CtrlUWlim	Real
$CtrlRefval	Real
$CtrlLWlim	Real
$CtrlLClim	Real
$CtrlApplyrules	TRUE \| FALSE
$CtrlRule1x2S	TRUE \| FALSE
$CtrlRule1x3S	TRUE \| FALSE
$CtrlRule2x2S	TRUE \| FALSE
$CtrlRule4x1S	TRUE \| FALSE
$CtrlRule10X	TRUE \| FALSE
$CtrlRule10N	Integer

For details, see Control chart

Correlation

PROC	CORRELATION
$VariableX	Variable
$VariableY	Variable
$Filter	Filter
$LogX	TRUE \| FALSE
$LogY	TRUE \| FALSE

For details, see Correlation coefficient

Cox proportional-hazards regression

PROC	COXREGRESSION
$Time	Variable
$EndPoint	Variable
$Covariates[1..24]	Variable(s)
$Filter	Filter
$SelectMethod	ENTER \| FORWARD \| BACKWARD \| STEPWISE
$PEnter	Real
$PRemove	Real
$SurvivalGraph	NONE \| SURVIVAL \| 100_SURVIVAL
$Factor	Variable

For details, see Cox proportional-hazards regression

Cronbach's alpha

PROC	CRONBACHALPHA
$Covariates[1..24]	Variable(s)
$Filter	Filter
$ScaleReversal	TRUE \| FALSE

For details, see Cronbach's alpha

Cumulative Distribution Plot

PROC	CUMULATIVE
$Variable	Variable
$Filter	Filter
$Polygon	TRUE \| FALSE
$Complementary	TRUE \| FALSE
$PlotNormalCurve	TRUE \| FALSE
$PlotAllPoints	TRUE \| FALSE

For details, see Cumulative frequency distribution

Deming regression

PROC	DEMING
$VariableX	Variable
$VariableY	Variable
$ReplicateX	Variable
$ReplicateY	Variable
$Filter	Filter
$Subgroups	Variable
$CVx	Real
$CVy	Real
$Scatterdiagram	TRUE \| FALSE
$ResidualsPlot	TRUE \| FALSE

For details, see Passing-Bablok regression

Dot plot

PROC	DOTPLOT
$Variable	Variable
$Filter	Filter
$Bars	TRUE \| FALSE
$MidLines	TRUE \| FALSE
$MidMarkers	TRUE \| FALSE
$ConnectingLines	TRUE \| FALSE
$MeanOrMedian	MEAN \| MEDIAN
$ErrorBarsMean	NONE \| MEANCI \| 1SD \| 2SD \| 3SD \| SEM \| RANGE
$ErrorBarsMedian	NONE \| MEDIANCI \| PR50 \| PR80 \| PR90 \| PR95 \| PR98 \| RANGE (*)
$BoxWhisker	TRUE \| FALSE
$Log	TRUE \| FALSE

For details, see Dot plot

(*) Note: PR50 is 50% percentile range (25-75 percentiles; Interquartile range); PR80 is 80% percentile range (10-90 percentile), etc.

F-test - Variance ratio test

PROC	FTEST
$Variable1	Variable
$Variable2	Variable
$Filter1	Filter
$Filter2	Filter
$Log	TRUE \| FALSE

For details, see Variance ratio test (F-test)

Fisher's exact test

PROC	FISHER
$Factor1	Variable
$Factor2	Variable
$Filter	Filter

For details, see Fisher's exact test

Frequency charts

PROC	FREQUENCYCHART
$Factor1	Variable
$Factor2	Variable
$Filter	Filter
$FrequencyChartType	NONE \| SIMPLE \| CLUSTERED \| STACKED \| 100%STACKED
$RelativeFreq	TRUE \| FALSE
$ReverseX	TRUE \| FALSE
$ReverseY	TRUE \| FALSE
$Horizontal	TRUE \| FALSE
$SortByMagnitude	TRUE \| FALSE

For details, see Frequencies bar charts

Friedman test

PROC	FRIEDMAN
$Covariates[1..24]	Variable(s)
$Filter	Filter
$Alpha	Real

For details, see Friedman test

Function plot

PROC	FUNCTIONPLOT
$Functions[1..6]	Functions
$Xmin	Real
$Xmax	Real
$AutoY	TRUE \| FALSE
$Ymin	Real
$Ymax	Real

For details, see Function plot

Histogram

PROC	HISTOGRAM
$Variable	Variable
$Filter	Filter
$PlotNormalCurve	TRUE \| FALSE
$Relative	TRUE \| FALSE
$SecondaryY	TRUE \| FALSE

For details, see Histogram

Independent samples t-test

PROC	TTEST
$Variable1	Variable
$Variable2	Variable
$Filter1	Filter
$Filter2	Filter
$Log	TRUE \| FALSE
$CI	Integer (%)
$EqualVariance	YES \| NO \| AUTO \| BOTH
$NormalityTest	KOLSMIR \| DAGPEAR \| SHAPWILK \| SHAPFRANC \| CHISQ \| NONE

For details, see Independent samples t-test

Inter-rater agreement (Kappa)

PROC	KAPPA
$Factor1	Variable
$Factor2	Variable
$Filter	Filter
$KappaWeights	NONE \| LINEAR \| QUADRATIC

For details, see Inter-rater agreement (kappa)

Intraclass correlation coefficient

PROC	INTRACLASSCC
$Covariates[1..24]	Variable(s)
$Filter	Filter
$ICCModel	RANDOM \| FIXED
$ICCType	CONSISTENCY \| AGREEMENT

For details, see Intraclass correlation coefficient

Kaplan-Meier survival analysis

PROC	KAPLANMEIER
$Time	Variable
$EndPoint	Variable
$Factor	Variable
$Filter	Filter
$LinearTrend	TRUE \| FALSE
$RMST	TRUE \| FALSE
$RMSTtime	Real
$SurvivalGraph	NONE \| SURVIVAL \| 100MINSURVIVAL
$Include95CI	TRUE \| FALSE
$MarkCensoredData	TRUE \| FALSE
$NumberAtRiskTable	TRUE \| FALSE

For details, see Kaplan-Meier survival curves

Kruskal-Wallis test

PROC	KRUSKALWALLIS
$Variable	Variable
$Factor	Variable
$Filter	Filter
$KWPostHocTest	CONOVER \| DUNN
$Alpha	Real
$Jonckheere	TRUE \| FALSE

For details, see Kruskal-Wallis test

Logistic Regression

PROC	LOGISTICREGRESSION
$Classifier	Dichotomous variable
$Covariates[1..24]	Variable(s)
$Filter	Filter
$SelectMethod	ENTER \| FORWARD \| BACKWARD \| STEPWISE
$PEnter	Real
$PRemove	Real
$CuttOff	Real
$ProbabilityPlot	TRUE \| FALSE
$PlotAllPoints	TRUE \| FALSE

For details, see Logistic regression

Mann-Whitney test (independent samples)

PROC	MANNWHITNEY
$Variable1	Variable
$Variable2	Variable
$Filter1	Filter
$Filter2	Filter

For details, see Mann-Whitney test (independent samples)

McNemar test

PROC	MCNEMAR
$Factor1	Variable
$Factor2	Variable
$Filter	Filter

For details, see McNemar test

Meta-analysis: area under ROC curve

PROC	METAROCAREA
$MetaStudies	Variable
$MetaEstimate	Variable
$MetaSE	Variable
$Filter	Filter
$ForestPlot	TRUE \| FALSE
$RelativeMarkerSz	TRUE \| FALSE
$WeightsModel	FIXED \| RANDOM
$PlotPooledEffects	NONE \| FIXED \| RANDOM \| BOTH
$Diamonds	TRUE \| FALSE
$FunnelPlot	TRUE \| FALSE

For details, see Meta-analysis: area under ROC curve

Meta-analysis: continuous measure

PROC	METACONTINUOUS
$MetaStudies	Variable
$MetaN1	Variable
$MetaMean1	Variable
$MetaSD1	Variable
$MetaN2	Variable
$MetaMean2	Variable
$MetaSD2	Variable
$Filter	Filter
$ForestPlot	TRUE \| FALSE
$RelativeMarkerSz	TRUE \| FALSE
$WeightsModel	FIXED \| RANDOM
$PlotPooledEffects	NONE \| FIXED \| RANDOM \| BOTH
$Diamonds	TRUE \| FALSE
$FunnelPlot	TRUE \| FALSE

For details, see Meta-analysis: continuous measure

Meta-analysis: correlation

PROC	METACORRELATION
$MetaStudies	Variable
$MetaN	Variable
$MetaEstimate	Variable
$Filter	Filter
$ForestPlot	TRUE \| FALSE
$RelativeMarkerSz	TRUE \| FALSE
$WeightsModel	FIXED \| RANDOM
$PlotPooledEffects	NONE \| FIXED \| RANDOM \| BOTH
$Diamonds	TRUE \| FALSE
$FunnelPlot	TRUE \| FALSE

For details, see Meta-analysis: correlation

Meta-analysis: generic inverse variance method

PROC	METAGENERIC
$MetaStudies	Variable
$MetaEstimate	Variable
$MetaSE	Variable
$Filter	Filter
$Log	TRUE \| FALSE
$ForestPlot	TRUE \| FALSE
$RelativeMarkerSz	TRUE \| FALSE
$WeightsModel	FIXED \| RANDOM
$PlotPooledEffects	NONE \| FIXED \| RANDOM \| BOTH
$Diamonds	TRUE \| FALSE
$FunnelPlot	TRUE \| FALSE

For details, see Meta-analysis: generic inverse variance method

Meta-analysis: odds ratio

PROC	METAODDSRATIO
$MetaStudies	Variable
$MetaN1	Variable
$MetaPos1	Variable
$MetaN2	Variable
$MetaPos2	Variable
$Filter	Filter
$ForestPlot	TRUE \| FALSE
$RelativeMarkerSz	TRUE \| FALSE
$WeightsModel	FIXED \| RANDOM
$PlotPooledEffects	NONE \| FIXED \| RANDOM \| BOTH
$Diamonds	TRUE \| FALSE
$FunnelPlot	TRUE \| FALSE

For details, see Meta-analysis: odds ratio

Meta-analysis: proportion

PROC	METAPROPORTION
$MetaStudies	Variable
$MetaN	Variable
$MetaPos	Variable
$Filter	Filter
$ForestPlot	TRUE \| FALSE
$RelativeMarkerSz	TRUE \| FALSE
$WeightsModel	FIXED \| RANDOM
$PlotPooledEffects	NONE \| FIXED \| RANDOM \| BOTH
$Diamonds	TRUE \| FALSE
$FunnelPlot	TRUE \| FALSE

For details, see Meta-analysis: proportion

Meta-analysis: relative risk

PROC	METARELATIVERISK
$MetaStudies	Variable
$MetaN1	Variable
$MetaPos1	Variable
$MetaN2	Variable
$MetaPos2	Variable
$Filter	Filter
$ForestPlot	TRUE \| FALSE
$RelativeMarkerSz	TRUE \| FALSE
$WeightsModel	FIXED \| RANDOM
$PlotPooledEffects	NONE \| FIXED \| RANDOM \| BOTH
$Diamonds	TRUE \| FALSE
$FunnelPlot	TRUE \| FALSE

For details, see Meta-analysis: relative risk

Meta-analysis: risk difference

PROC	METARISKDIFFERENCE
$MetaStudies	Variable
$MetaN1	Variable
$MetaPos1	Variable
$MetaN2	Variable
$MetaPos2	Variable
$Filter	Filter
$ForestPlot	TRUE \| FALSE
$RelativeMarkerSz	TRUE \| FALSE
$WeightsModel	FIXED \| RANDOM
$PlotPooledEffects	NONE \| FIXED \| RANDOM \| BOTH
$Diamonds	TRUE \| FALSE
$FunnelPlot	TRUE \| FALSE

For details, see Meta-analysis: risk difference

Mountain Plot

PROC	MOUNTAINPLOT
$Variable1	Variable
$Variable2	Variable
$Variable3	Variable
$Filter	Filter
$PlotAllPoints	TRUE \| FALSE

For details, see Mountain plot

Multiple comparison graphs

PROC	GRAPHMULTCOMP
$Variable	Variable
$Factor	Variable
$Filter	Filter
$Bars	TRUE \| FALSE
$MidLines	TRUE \| FALSE
$MidMarkers	TRUE \| FALSE
$ConnectingLines	TRUE \| FALSE
$MeanOrMedian	MEAN \| MEDIAN
$ErrorBarsMean	NONE \| MEANCI \| 1SD \| 2SD \| 3SD \| SEM \| RANGE
$ErrorBarsMedian	$ErrorBarsMedian	NONE \| MEDIANCI \| PR50 \| PR80 \| PR90 \| PR95 \| PR98 \| RANGE (*)
$BoxWhisker	TRUE \| FALSE
$NotchedBW	TRUE \| FALSE
$ViolinPlot	TRUE \| FALSE
$PlotAllPoints	TRUE \| FALSE
$Log	TRUE \| FALSE
$ShowEmptyClasses	TRUE \| FALSE

For details, see Multiple comparison graphs

(*) Note: PR50 is 50% percentile range (25-75 percentiles; Interquartile range); PR80 is 80% percentile range (10-90 percentile), etc.

Multiple line graph

PROC	MULTILINEGRAPH
$Covariates[1..6]	Variables
$Labels	Variable
$Filter	Filter
$Log	TRUE \| FALSE
$CompleteCasesOnly	TRUE \| FALSE
$PlotAllPoints	TRUE \| FALSE
$TrendLine	NONE \| MOVINGAVG \| LOESS \| ISOTONIC
$WindowWidth	Integer
$Span	Integer (%)

For details, see Multiple line graph

Multiple variables graphs

PROC	GRAPHMULTIVAR
$Covariates[1..24]	Variable(s)
$Filter	Filter
$Bars	TRUE \| FALSE
$MidLines	TRUE \| FALSE
$MidMarkers	TRUE \| FALSE
$ConnectingLines	TRUE \| FALSE
$MeanOrMedian	MEAN \| MEDIAN
$ErrorBarsMean	NONE \| MEANCI \| 1SD \| 2SD \| 3SD \| SEM \| RANGE
$ErrorBarsMedian	$ErrorBarsMedian	NONE \| MEDIANCI \| PR50 \| PR80 \| PR90 \| PR95 \| PR98 \| RANGE (*)
$BoxWhisker	TRUE \| FALSE
$NotchedBW	TRUE \| FALSE
$ViolinPlot	TRUE \| FALSE
$PlotAllPoints	TRUE \| FALSE
$DotAndLines	TRUE \| FALSE
$Cumulative	TRUE \| FALSE
$Log	TRUE \| FALSE
$CompleteCasesOnly	TRUE \| FALSE

For details, see Multiple variables graphs

(*) Note: PR50 is 50% percentile range (25-75 percentiles; Interquartile range); PR80 is 80% percentile range (10-90 percentile), etc.

Multiple Regression

PROC	MULTIREGRESSION
$VariableY	Variable
$Covariates[1..24]	Variable(s)
$Weights	Variable
$Filter	Filter
$SelectMethod	ENTER \| FORWARD \| BACKWARD \| STEPWISE
$PEnter	Real
$PRemove	Real
$ReportVIF	TRUE \| FALSE
$ZeroOrderCC	TRUE \| FALSE
$NormalityTest	KOLSMIR \| DAGPEAR \| SHAPWILK \| SHAPFRANC \| CHISQ \| NONE

For details, see Multiple regression

Normal plot

PROC	NORMALPLOT
$Variable	Variable
$Filter	Filter
$QQPlot	TRUE \| FALSE
$NormalityTest	KOLSMIR \| DAGPEAR \| SHAPWILK \| SHAPFRANC \| CHISQ \| NONE

For details, see Normal plot

One-sample t-test

PROC	ONESAMPLETTEST
$Variable	Variable
$Filter	Filter
$Log	TRUE \| FALSE
$TestValue	Real

For details, see One sample t-test

One-way ANOVA

PROC	ANOVA
$Variable	Variable
$Factor	Variable
$Filter	Filter
$Log	TRUE \| FALSE
$PostHocTest	SCHEFFE \| TUKEYKRAMER \| SNK
$Alpha	Real
$NormalityTest	KOLSMIR \| DAGPEAR \| SHAPWILK \| SHAPFRANC \| CHISQ \| NONE

For details, see One-way analysis of variance

Outlier Detection

PROC	OUTLIERS
$Variable	Variable
$Filter	Filter
$Log	TRUE \| FALSE
$NormalityTest	KOLSMIR \| DAGPEAR \| SHAPWILK \| SHAPFRANC \| CHISQ \| NONE
$Grubbs	LEFT \| RIGHT \| DOUBLE \| NO
$OutliersAlpha	0.10 \| 0.05 \| 0.025 \| 0.01 \| 0.005 \| 0.001 \| 0.0005 \| 0.0001
$GeneralizedESDtest	TRUE \| FALSE
$MaxOutliers	Integer
$OutliersTukey	TRUE \| FALSE

For details, see Outlier detection

Paired samples t-test

PROC	TTESTPAIRED
$Variable1	Variable
$Variable2	Variable
$Filter	Filter
$Log	TRUE \| FALSE
$CI	Integer (%)
$NormalityTest	KOLSMIR \| DAGPEAR \| SHAPWILK \| SHAPFRANC \| CHISQ \| NONE

For details, see Paired samples t-test

Partial area under ROC curve

PROC	PARTIALROC
$Variable	Variable
$Classifier	Dichotomous variable
$Filter	Filter
$PartialInterval	SENSITIVITY \| SPECIFICITY
$ValLow	Real (0..100)
$ValHigh	Real (0..100)
$Bootstrap	TRUE \| FALSE
$BootstrapLoops	Integer
$RandomNrSeed	Integer
$ShowGraph	TRUE \| FALSE

For details, see Partial area under ROC curve

Partial Correlation

PROC	PARTIALCORRELATION
$VariableX	Variable
$VariableY	Variable
$Covariates[1..24]	Variable(s)
$Filter	Filter
$LogX	TRUE \| FALSE
$LogY	TRUE \| FALSE

For details, see Partial correlation

Passing-Bablok regression

PROC	PASSINGBABLOK
$VariableX	Variable
$VariableY	Variable
$Filter	Filter
$Subgroups	Variable
$PerpendicularResiduals	TRUE \| FALSE
$Spearman	TRUE \| FALSE
$Scatterdiagram	TRUE \| FALSE
$ResidualsPlot	TRUE \| FALSE
$ResidualsByRank	TRUE \| FALSE
$BootstrapCI	TRUE \| FALSE
$BootstrapLoops	Integer
$RandomNrSeed	Integer
$BiasAtLevels	TRUE \| FALSE
$Level1..$Level4	Real

For details, see Passing-Bablok regression

Polar plot

PROC	POLARPLOT
$Radius	Variable
$Angle	Variable
$Filter	Filter
$AngleUnit	RADIANS \| DEGREES \| PERCENTAGE
$ClockWise	TRUE \| FALSE
$ZeroAngle	NORTH \| EAST \| SOUTH \| WEST
$ConnectPoints	TRUE \| FALSE

For details, see Polar plot

Precision-recall curve

PROC	PRECISIONRECALL
$Variable	Variable
$Classifier	Dichotomous variable
$Filter	Filter
$MarkPoints	TRUE \| FALSE

For details, see Precision-recall curve

Probit regression

PROC	PROBITREGRESSION
$ProbitBinaryData	TRUE \| FALSE
$VariableX	Dose variable
For binary data
$VariableY	Response variable
For grouped data
$Variable2	Variable with total number of cases
$Variable3	Variable with number of responses
$Filter	Filter
$Log	TRUE \| FALSE
$DoseResponsePlot	TRUE \| FALSE
$PlotAllPoints	TRUE \| FALSE

For details, see Probit regression (Dose-Response analysis)

Rank Correlation

PROC	RANKCORRELATION
$VariableX	Variable
$VariableY	Variable
$Filter	Filter
$RankCoefficients	SPEARMAN \| KENDALL \| BOTH
$BootstrapLoops	Integer
$RandomNrSeed	Integer

For details, see Rank correlation

Reference interval

PROC	REFINTERVAL
$Variable	Variable
$Filter	Filter
$Interval	90 \| 95 \| 99 \| 99.9 \| 99.99
$IntervalSides	DOUBLE \| LEFT \| RIGHT
$RefIntervalOutlierTest	NONE \| REED \| TUKEY
$FollowCLSI	TRUE \| FALSE
$RobustMethod	TRUE \| FALSE
$Log	TRUE \| FALSE
$BoxCoxTransform	TRUE \| FALSE
$Lambda	Real
$Shift	Real
$NormalityTest	KOLSMIR \| DAGPEAR \| SHAPWILK \| SHAPFRANC \| CHISQ \| NONE
$BootstrapLoops	Integer
$RandomNrSeed	Integer
$ShowGraph	TRUE \| FALSE

For details, see Reference interval

Regression

PROC	REGRESSION
$VariableX	Variable
$VariableY	Variable
$Weights	Variable
$Filter	Filter
$Subgroups	Variable
$ThroughOrigin	TRUE \| FALSE
$RegressionModel	LINEAR \| LOGX \| LOGY \| LOGLOG \| QUADRATIC
$NormalityTest	KOLSMIR \| DAGPEAR \| SHAPWILK \| SHAPFRANC \| CHISQ \| NONE

For details, see Regression

Relative risk and Odds ratio

PROC	RELATIVERISK
$Factor1	Variable
$Factor2	Variable
$Filter	Filter
$RelRisk	TRUE \| FALSE
$NNT	TRUE \| FALSE
$OddsRatio	TRUE \| FALSE

For details, see Relative risk & Odds ratio

Repeated Measures ANOVA

PROC	REPEATED
$Covariates[1..24]	Variable(s)
$Factor	Variable
$Filter	Filter
$Log	TRUE \| FALSE

For details, see Repeated measures analysis of variance

Responsiveness

PROC	RESPONSIVENESS
$Variable1	Variable
$Variable2	Variable
$Filter	Filter
$Paired	TRUE \| FALSE
$FirstMinusSecond	TRUE \| FALSE
$BootstrapLoops	Integer
$RandomNrSeed	Integer

For details, see Responsiveness

ROC Curve analysis

PROC	ROC
$Variable	Variable
$Classifier	Dichotomous variable
$Filter	Filter
$ROCMethod	DELONG \| HANLEY
$BinomialCI	TRUE \| FALSE
$PrevalenceRatio	TRUE \| FALSE
$Prevalence	Real (range 0..1)
$CriterionList	TRUE \| FALSE
$ListSensSpecCI	TRUE \| FALSE
$ListLRatiosCI	TRUE \| FALSE
$ListPredictCI	TRUE \| FALSE
$BootstrapYouden	TRUE \| FALSE
$ShowSummarytable	TRUE \| FALSE
$BootstrapLoops	Integer
$RandomNrSeed	Integer
$CalcOptCutoff	TRUE \| FALSE
$TPC	Real
$TNC	Real
$FPC	Real
$FNC	Real
$ShowGraph	TRUE \| FALSE
$MarkPoints	TRUE \| FALSE
$GraphCI	TRUE \| FALSE

For details, see ROC analysis in MedCalc

Scatter Diagram

PROC	SCATTER
$VariableX	Variable
$VariableY	Variable
$Filter	Filter
$Subgroups	Variable
$LogX	TRUE \| FALSE
$LogY	TRUE \| FALSE
$LineOfEquality	TRUE \| FALSE
$HeatMap	TRUE \| FALSE
$TrendLine	NONE \| MOVINGAVG \| LOESS \| REDUCEDMA \| ISOTONIC
$WindowWidth	Integer
$Span	Integer (%)

For details, see Scatter diagram

Scatter diagram with regression line

PROC	REGRESSIONPLOT
$VariableX	Variable
$VariableY	Variable
$Weights	Variable
$Filter	Filter
$ThroughOrigin	TRUE \| FALSE
$RegressionModel	LINEAR \| LOGX \| LOGY \| LOGLOG \| QUADRATIC
$LineOfEquality	TRUE \| FALSE
$HeatMap	TRUE \| FALSE
$CICurves	TRUE \| FALSE
$PICurves	TRUE \| FALSE
$Subgroups	Variable
$SubgroupsPlot	ALLCASES \| SUBONLY \| BOTH
$ResidualsPlot	TRUE \| FALSE

For details, see Scatter diagram & regression line

Signed rank sum test

PROC	SIGNEDRANKSUMTEST
$Variable	Variable
$Filter	Filter
$TestValue	Real

For details, see Signed rank sum test (one sample)

Summary Statistics

PROC	MEANS
$Variable	Variable
$Filter	Filter
$Subgroups	Variable
$Log	TRUE \| FALSE
$Percentiles	{percentiles list}
$Harmonicmean	TRUE \| FALSE
$Geometricmean	TRUE \| FALSE
$TrimmedMean	TRUE \| FALSE
$TrimPercentage	Real (%)
$NormalityTest	KOLSMIR \| DAGPEAR \| SHAPWILK \| SHAPFRANC \| CHISQ \| NONE

For details, see Summary statistics

Trimmed means: Comparison of independent samples (Yuen-Welch test)

PROC	TRIMMEDMEANTEST
$Variable1	Variable
$Variable2	Variable
$Filter1	Filter
$Filter2	Filter
$TrimPercentage	Real (%)
$CI	Integer (%)

For details, see Yuen-Welch test

Trimmed means: Paired samples test

PROC	PAIREDTRIMMEDMEANTEST
$Variable1	Variable
$Variable2	Variable
$Filter	Filter
$TrimPercentage	Real (%)
$CI	Integer (%)

For details, see Trimmed means: Paired samples test

Two-way ANOVA

PROC	2WAYANOVA
$Variable	Variable
$Factor1	Variable
$Factor2	Variable
$Filter	Filter
$NormalityTest	KOLSMIR \| DAGPEAR \| SHAPWILK \| SHAPFRANC \| CHISQ \| NONE

For details, see Two-way analysis of variance

Waterfall chart

PROC	WATERFALLCHART
$Variable	Variable
$Groups	Group variable
$Filter	Filter
$ShowCaseNumbers	TRUE \| FALSE
$WaterfallHorizontal	TRUE \| FALSE
$WaterfallSort	DESCENDING \| ASCENDING \| NOSORT

For details, see Waterfall chart

Wilcoxon test (paired samples)

PROC	WILCOXON
$Variable1	Variable
$Variable2	Variable
$Filter	Filter

For details, see Wilcoxon test (paired samples)

Scripts - Parameters for statistical procedures and graphs

See also