The SURVEYFREQ Procedure

TABLES Statement

Subsections:

  • TABLES requests < / options > ;

The TABLES statement requests one-way to n-way frequency and crosstabulation tables and statistics for these tables.

If you omit the TABLES statement, PROC SURVEYFREQ generates one-way frequency tables for all DATA= data set variables that are not listed in the other statements.

The following argument is required in the TABLES statement:

requests

specify the frequency and crosstabulation tables to produce. A request is composed of one variable name or several variable names separated by asterisks. To request a one-way frequency table, use a single variable. To request a two-way crosstabulation table, use an asterisk between two variables. To request a multiway table (an n-way table, where n > 2), separate the desired variables with asterisks. The unique values of these variables form the rows, columns, and layers of the table.

For two-way tables to multiway tables, the values of the last variable form the crosstabulation table columns, while the values of the next-to-last variable form the rows. Each level (or combination of levels) of the other variables forms one layer. PROC SURVEYFREQ produces a separate crosstabulation table for each layer. For example, a specification of A*B*C*D in a TABLES statement produces k tables, where k is the number of different combinations of levels for A and B. Each table lists the levels for D (columns) within each level of C (rows).

You can use multiple TABLES statements in a single PROC SURVEYFREQ step. You can also specify any number of table requests in a single TABLES statement. To specify multiple table requests quickly, use a grouping syntax by placing parentheses around several variables and joining other variables or variable combinations. Table 97.3 shows some examples of grouping syntax.

Table 97.3: Grouping Syntax

TABLES Request

Equivalent to

A*(B C)

A*B   A*C

(A B)*(C D)

A*C   B*C   A*D   B*D

(A B C)*D

A*D   B*D   C*D

A – – C

A   B   C

(A – – C)*D

A*D   B*D   C*D


The TABLES statement variables are one or more variables from the DATA= input data set. These variables can be either character or numeric, but the procedure treats them as categorical variables. PROC SURVEYFREQ uses the formatted values of the TABLES variable to determine the categorical variable levels. If you assign a format to a variable by using a FORMAT statement, PROC SURVEYFREQ formats the values before dividing observations into the levels of a frequency or crosstabulation table. See the discussion of the FORMAT procedure in the Base SAS Procedures Guide and the discussions of the FORMAT statement and SAS formats in SAS Formats and Informats: Reference.

By default, the frequency or crosstabulation table lists the values of both character and numeric variables in ascending order based on internal (unformatted) variable values. You can change the order of the values in the table by specifying the ORDER= option in the PROC SURVEYFREQ statement. To list the values in ascending order by formatted value, use ORDER=FORMATTED.

Without Options

If you request a frequency or crosstabulation table without specifying options, PROC SURVEYFREQ produces the following for each table level or cell:

  • frequency (sample size)

  • weighted frequency, which estimates the population total

  • standard deviation of the weighted frequency

  • percentage, which estimates the population proportion

  • standard error of the percentage

The table displays weighted frequencies if your analysis includes a WEIGHT statement, or if you specify the WTFREQ option in the TABLES statement. The table also displays the number of observations that have missing values. For more information, see the sections One-Way Frequency Tables and Crosstabulation Tables.

Options

Table 97.4 summarizes the options available in the TABLES statement. Descriptions of the options follow the table in alphabetical order.

Table 97.4: TABLES Statement Options

Option

Description

Control Statistical Analysis

AGREE

Requests kappa coefficients

ALPHA=

Sets level for confidence limits

CHISQ

Requests Rao-Scott chi-square test

CL

Requests confidence limits for percentages and

 

specifies confidence limit type for percentages

CLWT

Requests confidence limits for weighted frequencies

COV

Requests covariances of frequency estimates

COVP

Requests covariances of proportion estimates

DF=

Specifies degrees of freedom

KAPPA

Requests simple kappa coefficient

LRCHISQ

Requests Rao-Scott likelihood ratio test

OR

Requests odds ratio and relative risks

RISK

Requests risks and risk difference

TESTP=

Specifies null proportions for one-way chi-square test

WCHISQ

Requests Wald chi-square test

WLLCHISQ

Requests Wald log-linear chi-square test

WTKAPPA

Requests weighted kappa coefficient

Request Additional Table Information

CELLCHI2

Displays cell contributions to the Pearson chi-square

CLWT

Displays confidence limits for weighted frequencies

COLUMN

Displays column percentages and standard errors

CV

Displays coefficients of variation for percentages

CVWT

Displays coefficients of variation for weighted frequencies

DEFF

Displays design effects for percentages

DEVIATION

Displays deviations of weighted frequencies

EXPECTED

Displays expected weighted frequencies

PEARSONRES

Displays Pearson residuals

ROW

Displays row percentages and standard errors

VAR

Displays variances of percentages

VARWT

Displays variances of weighted frequencies

WTFREQ

Displays totals and standard errors

 

when there is no WEIGHT statement

Control Displayed Output

NOCELLPERCENT

Suppresses display of overall percentages

NOFREQ

Suppresses display of frequency counts

NOPERCENT

Suppresses display of all percentages

NOPRINT

Suppresses display of tables but displays statistical tests

NOSPARSE

Suppresses display of zero rows and columns

NOSTD

Suppresses display of standard errors for all estimates

NOTOTAL

Suppresses display of row and column totals

NOWT

Suppresses display of weighted frequencies

Produce Statistical Graphics

PLOTS=

Requests plots from ODS Graphics


You can specify the following options in a TABLES statement:

AGREE < (options)>

requests the simple and weighted kappa coefficients with their standard errors and confidence limits. Kappa coefficients can be computed for square two-way tables, where the number of rows equals the number of columns. For $2 \times 2$ tables, the weighted kappa coefficient equals the simple kappa coefficient, and PROC SURVEYFREQ displays only the simple kappa coefficient. For more information, see the section Kappa Coefficients.

Kappa coefficients are available when you specify variance estimation by the jackknife method (VARMETHOD=JACKKNIFE ) or by balanced repeated replication (VARMETHOD=BRR ); kappa coefficients are not available with the Taylor series method (VARMETHOD=TAYLOR ).

The weighted kappa coefficient is computed by using agreement weights that reflect the relative agreement between pairs of variable levels. Agreement weights are not the same as sampling weights, which you provide by specifying the WEIGHT statement. PROC SURVEYFREQ uses sampling weights to compute both the simple and weighted kappa coefficients. For more information, see the section Weighted Kappa Coefficient.

You can specify the level for the kappa confidence limits in the ALPHA= option. By default, ALPHA=0.05, which produces 95% confidence limits.

You can request the simple kappa coefficient or the weighted kappa coefficient separately by specifying the KAPPA or WTKAPPA option, respectively.

You can specify the following options:

PRINTKWTS

displays the agreement weights that PROC SURVEYFREQ uses to compute the weighted kappa coefficient. Agreement weights reflect the relative agreement between pairs of variable levels. By default, PROC SURVEYFREQ uses the Cicchetti-Allison form of agreement weights. If you specify the WT=FC option, the procedure uses the Fleiss-Cohen form of agreement weights. For more information, see the section Weighted Kappa Coefficient.

WT=FC

requests Fleiss-Cohen agreement weights for the weighted kappa computation. By default, PROC SURVEYFREQ uses Cicchetti-Allison agreement weights to compute the weighted kappa coefficient. Agreement weights reflect the relative agreement between pairs of variable levels. For more information, see the section Weighted Kappa Coefficient.

ALPHA=$\alpha $

specifies the level for confidence limits. The value of $\alpha $ must be between 0 and 1; a confidence level of $\alpha $ produces $100(1 - \alpha )$% confidence limits. By default, ALPHA=0.05, which produces 95% confidence limits.

You can request confidence limits for percentages by specifying the CL option, and you can request confidence limits for weighted frequencies by specifying the CLWT option. For more information, see the sections Confidence Limits for Proportions and Confidence Limits for Totals.

The ALPHA= option also applies to confidence limits for the risks and risk difference (which you can request by specifying the RISK option) and to confidence limits for the odds ratio and relative risks (which you can request by specifying the OR option). For more information, see the sections Risks and Risk Difference and Odds Ratio and Relative Risks.

CELLCHI2

displays each table cell’s contribution to the Pearson chi-square statistic in the crosstabulation table. The cell chi-square is computed as $(\emph{weighted frequency} - \emph{expected} )^2 ~  / ~  \emph{expected}$, where weighted frequency is the weighted frequency of the table cell and expected is the expected weighted frequency, which is computed under the null hypothesis that the row and column variables are independent. You can display the expected weighted frequencies by specifying the EXPECTED option, and you can display the deviations (weighted frequencyexpected) by specifying the DEVIATION option. For more information, see the sections Expected Weighted Frequency and Rao-Scott Chi-Square Test. This option has no effect for one-way tables.

CHISQ < (options)>

requests the Rao-Scott chi-square test. This is a design-adjusted test that is computed by applying a design correction to the weighted Pearson chi-square statistic. By default, PROC SURVEYFREQ provides a first-order Rao-Scott chi-square test. If you specify CHISQ(SECONDORDER) , the procedure provides a second-order (Satterthwaite) Rao-Scott chi-square test. For more information, see the section Rao-Scott Chi-Square Test.

For one-way tables, the CHISQ option produces a design-based goodness-of-fit test. By default, this is a goodness-of-fit test for equal proportions. If you specify the null hypothesis proportions in the TESTP= option, the CHISQ option produces a chi-square goodness-of-fit test for the specified proportions.

By default for one-way tables, and for first-order tests for two-way tables, the design correction is computed from proportion estimates. If you specify CHISQ(MODIFIED) , the design correction is computed from null hypothesis proportions. For second-order tests for two-way tables, the design correction is always computed from null hypothesis proportions.

You can specify the following options:

FIRSTORDER

requests a first-order Rao-Scott chi-square test. This is the default for the CHISQ option; if you do not specify CHISQ(SECONDORDER) , the procedure provides a first-order Rao-Scott test.

MODIFIED

uses the null hypothesis proportions to compute the Rao-Scott design correction. By default (if you do not specify CHISQ(MODIFIED)), the procedure uses proportion estimates to compute the design correction for all first-order tests and for second-order tests for one-way tables. For second-order tests for two-way tables, the procedure always uses null hypothesis proportions to compute the design correction.

SECONDORDER

requests a second-order (Satterthwaite) Rao-Scott chi-square test. For more information, see the section Rao-Scott Chi-Square Test.

CL < (options)>

requests confidence limits for the percentages (proportions) in the crosstabulation table. By default, PROC SURVEYFREQ computes standard Wald ("linear") confidence limits for proportions by using the variance estimates that are based on the sample design. For more information, see the section Confidence Limits for Proportions. You can specify the confidence level in the ALPHA= option. By default, ALPHA=0.05, which produces 95% confidence limits.

You can specify options in parentheses after the CL option to control the confidence limit computations. You can use the TYPE= option to request an alternative confidence limit type. In addition to Wald confidence limits, the following types of design-based confidence limits are available for proportions: modified Clopper-Pearson (exact), modified Wilson (score), and logit confidence limits.

If you specify the PSMALL option, PROC SURVEYFREQ uses the alternative confidence limit type for extreme (small or large) proportion estimates and uses Wald confidence limits for all other proportion estimates. If you do not specify the PSMALL option, PROC SURVEYFREQ computes the specified confidence limit type for all proportion values.

You can specify the following options:

ADJUST=NO | YES

controls the degrees-of-freedom adjustment to the effective sample size for the modified Clopper-Pearson and Wilson confidence limits. By default, ADJUST=YES. If you specify ADJUST=NO, the confidence limit computations do not apply the degrees-of-freedom adjustment to the effective sample size. For more information, see the section Modified Confidence Limits.

The ADJUST= option is available for TYPE=CLOPPERPEARSON and TYPE=WILSON confidence limits.

PSMALL < =p >

uses the alternative confidence limit type that you specify in the TYPE= option for extreme (small or large) proportion values.

The PSMALL value p defines the range of extreme proportion values, where those proportions less than or equal to p or greater than or equal to (1 – p) are considered to be extreme, and those proportions between p and (1 – p) are not extreme. If you do not specify a PSMALL value p, PROC SURVEYFREQ uses p = 0.25 by default. For $\Argument{p} = 0.25$, the procedure computes Wald confidence limits for proportions between 0.25 and 0.75 and computes the alternative confidence limit type for proportions less than or equal to 0.25 or greater than or equal to 0.75.

The PSMALL value p must be a nonnegative number. You can specify p as a proportion between 0 and 0.5. Or you can specify p in percentage form as a number between 1 and 50, and PROC SURVEYFREQ converts that number to a proportion. The procedure treats the value 1 as the percentage form 1%.

The PSMALL option is available for TYPE=CLOPPERPEARSON , TYPE=LOGIT , and TYPE=WILSON confidence limits. For more information, see the section Confidence Limits for Proportions.

TRUNCATE=NO | YES

controls the truncation of the effective sample size for the modified Clopper-Pearson and Wilson confidence limits. By default, TRUNCATE=YES truncates the effective sample size if it is larger than the original sample size. If you specify TRUNCATE=NO, the effective sample size is not truncated. For more information, see the section Modified Confidence Limits.

The TRUNCATE= option is available for TYPE=CLOPPERPEARSON and TYPE=WILSON confidence limits.

TYPE=type

specifies the type of confidence limits to compute for proportions. If you do not specify the TYPE= option, PROC SURVEYFREQ computes Wald confidence limits (TYPE=WALD) by default.

If you specify the CL(PSMALL) option, the procedure uses the specified confidence limit type for extreme proportions (outside the PSMALL range) and uses Wald confidence limits for proportions that are not outside the range. If you do not specify the CL(PSMALL) option, the procedure uses the specified confidence limit type for all proportions.

You can specify one of the following confidence limit types:

CLOPPERPEARSON
CP

requests modified Clopper-Pearson (exact) confidence limits for proportions. For more information, see the section Modified Clopper-Pearson Confidence Limits.

LOGIT

requests logit confidence limits for proportions. For more information, see the section Logit Confidence Limits.

WALD

requests standard Wald ("linear") confidence limits for proportions. This is the default confidence limit type if you do not specify the TYPE= option. For more information, see the section Wald Confidence Limits.

WILSON
SCORE

requests modified Wilson (score) confidence limits for proportions. For more information, see the section Modified Wilson Confidence Limits.

CLWT

requests confidence limits for the weighted frequencies (totals) in the crosstabulation table. You can specify the confidence level in the ALPHA= option. By default, ALPHA=0.05, which produces 95% confidence limits. For more information, see the section Confidence Limits for Totals.

COLUMN < (option)>

displays the column percentage (estimated proportion of the column total) for each cell in a two-way table. The COLUMN option also provides the standard errors of the column percentages. For more information, see the section Row and Column Proportions. This option has no effect for one-way tables.

You can specify the following option:

DEFF

displays the design effect for each column percentage in the crosstabulation table. For more information, see the section Design Effect.

COV

requests the covariance matrix of the table cell frequency estimates. For more information, see the section Covariances of Frequency Estimates.

COVP

requests the covariance matrix of the proportion estimates.

CV

displays the coefficient of variation for each percentage (proportion) estimate in the crosstabulation table. For more information, see the section Coefficient of Variation.

CVWT

displays the coefficient of variation for each weighted frequency (estimated total), in the crosstabulation table. For more information, see the section Coefficient of Variation.

DEFF

displays the design effect for each overall percentage (proportion) estimate in the crosstabulation table. For more information, see the section Design Effect.

To request design effects for row or column percentages, specify the DEFF option in parentheses after the ROW or COLUMN option.

DEVIATION

displays the deviations of the weighted frequencies from the expected weighted frequencies (weighted frequencyexpected) in the crosstabulation table. The expected weighted frequencies are computed under the null hypothesis that the row and column variables are independent. You can display the expected values by specifying the EXPECTED option. For more information, see the section Expected Weighted Frequency. This option has no effect for one-way tables.

DF=df

specifies the degrees of freedom for the analysis. The value of df must be a nonnegative number. By default, PROC SURVEYFREQ computes the degrees of freedom as described in the section Degrees of Freedom.

PROC SURVEYFREQ uses the value df to obtain the t-percentile for confidence limits for proportions, totals, and other statistics. For more information, see the section Confidence Limits for Proportions. PROC SURVEYFREQ also uses df to compute the denominator degrees of freedom for the F statistics in the Rao-Scott and Wald chi-square tests. For more information, see the sections Rao-Scott Chi-Square Test, Rao-Scott Likelihood Ratio Chi-Square Test, Wald Chi-Square Test, and Wald Log-Linear Chi-Square Test.

EXPECTED

displays the expected weighted frequencies for the cells in the crosstabulation table. The expected weighted frequencies are computed under the null hypothesis that the row and column variables are independent. For more information, see the section Expected Weighted Frequency. This option has no effect for one-way tables.

KAPPA

requests the simple kappa coefficient with its standard error and confidence limits. The kappa coefficient can be computed for square two-way tables, where the number of rows equals the number of columns. For more information, see the section Simple Kappa Coefficient.

The kappa coefficient is available when you specify variance estimation by the jackknife method (VARMETHOD=JACKKNIFE ) or by balanced repeated replication (VARMETHOD=BRR ); the kappa coefficient is not available with the Taylor series method (VARMETHOD=TAYLOR ).

You can specify the level for the kappa confidence limits in the ALPHA= option. By default, ALPHA=0.05, which produces 95% confidence limits.

LRCHISQ < (options)>

requests the Rao-Scott likelihood ratio chi-square test. This is a design-adjusted test that is computed by applying a design correction to the weighted likelihood ratio chi-square statistic. By default, PROC SURVEYFREQ provides a first-order Rao-Scott likelihood ratio test. If you specify LRCHISQ(SECONDORDER) , the procedure provides a second-order (Satterthwaite) Rao-Scott likelihood ratio test. For more information, see the section Rao-Scott Likelihood Ratio Chi-Square Test.

For one-way tables, the LRCHISQ option produces a design-based likelihood ratio goodness-of-fit test. By default, the null hypothesis is equal proportions. If you specify null hypothesis proportions in the TESTP= option, the LRCHISQ option produces a design-based likelihood ratio test for the specified proportions.

By default for one-way tables, and for first-order tests for two-way tables, the design correction is computed from proportion estimates. If you specify LRCHISQ(MODIFIED) , the design correction is computed from null hypothesis proportions. For second-order tests for two-way tables, the design correction is always computed from null hypothesis proportions.

You can specify the following options:

FIRSTORDER

requests a first-order Rao-Scott likelihood ratio test. This is the default for the LRCHISQ option; if you do not specify LRCHISQ(SECONDORDER) , the procedure provides a first-order Rao-Scott test.

MODIFIED

uses the null hypothesis proportions to compute the Rao-Scott design correction. By default (if you do not specify LRCHISQ(MODIFIED)), the procedure uses proportion estimates to compute the design correction for all first-order tests and for second-order tests for one-way tables. For second-order tests for two-way tables, the procedure always uses null hypothesis proportions to compute the design correction.

SECONDORDER

requests a second-order (Satterthwaite) Rao-Scott likelihood ratio test. For more information, see the section Rao-Scott Likelihood Ratio Chi-Square Test.

NOCELLPERCENT

suppresses the display of overall cell percentages in the crosstabulation table, as well as the standard errors of the percentages. The NOCELLPERCENT option does not suppress the display of row or column percentages, which you can request by specifying the ROW or COLUMN option.

NOFREQ

suppresses the display of cell frequencies in the crosstabulation table. The NOFREQ option also suppresses the display of row, column, and overall table frequencies.

NOPERCENT

suppresses the display of all percentages in the crosstabulation table. The NOPERCENT option also suppresses the display of standard errors of the percentages. Use the NOCELLPERCENT option to suppress display of overall cell percentages but allow display of row or column percentages.

NOPRINT

suppresses the display of frequency and crosstabulation tables but displays all requested statistical tests. This option disables the Output Delivery System (ODS) for the suppressed tables. For more information, see Chapter 20: Using the Output Delivery System.

NOSPARSE

suppresses the display of variable levels with zero frequency in two-way tables. By default, the procedure displays all levels of the column variable within each level of the row variable, including any column variable levels with zero frequency for that row. For multiway tables, the procedure displays all levels of the row variable for each layer of the table by default, including any row variable levels with zero frequency for the layer.

NOSTD

suppresses the display of all standard errors in the crosstabulation table.

NOTOTAL

suppresses the display of row totals, column totals, and overall totals in the crosstabulation table.

NOWT

suppresses the display of weighted frequencies in the crosstabulation table. The NOWT option also suppresses the display of standard errors of the weighted frequencies.

OR
RELRISK

requests estimates of the odds ratio, the column 1 relative risk, and the column 2 relative risk for $2 \times 2$ tables. The OR option also provides confidence limits for these statistics. For more information, see the section Odds Ratio and Relative Risks.

You can specify the confidence level in the ALPHA= option. By default, ALPHA=0.05, which produces 95% confidence limits.

PEARSONRES

displays each crosstabulation table cell’s Pearson residual, which is the square root of the table cell’s contribution to the Pearson chi-square statistic. The Pearson residual is computed as $ (\emph{weighted frequency} - \emph{expected} ) ~  / ~  \sqrt {\emph{expected}} $, where weighted frequency is the weighted frequency of the table cell and expected is the expected weighted frequency, which is computed under the null hypothesis that the row and column variables are independent. You can display the expected values, the deviations, and the cell chi-squares by specifying the EXPECTED , DEVIATION , and CELLCHI2 options, respectively. For more information, see the sections Expected Weighted Frequency and Rao-Scott Chi-Square Test. This option has no effect for one-way tables.

PLOTS < (global-plot-options)> < =plot-request < (plot-options)> >
PLOTS < (global-plot-options)>
< =(plot-request < (plot-options)> < …plot-request < (plot-options)> > )>

controls the plots that are produced through ODS Graphics. Plot-requests identify the plots, and plot-options control the appearance and content of the plots. You can specify plot-options in parentheses after a plot-request. A global-plot-option applies to all plots for which it is available unless it is altered by a specific plot-option. You can specify global-plot-options in parentheses after the PLOTS option.

When you specify only one plot-request, you can omit the parentheses around the plot-request. For example:

 plots=all
 plots=wtfreqplot
 plots=(wtfreqplot oddsratioplot)
 plots(only)=(riskdiffplot relriskplot)

ODS Graphics must be enabled before plots can be requested. For example:

ods graphics on;
proc surveyfreq;
   tables treatment*response / chisq plots=wtfreqplot;
   weight wt;
run;
ods graphics off;

For more information about enabling and disabling ODS Graphics, see the section Enabling and Disabling ODS Graphics in Chapter 21: Statistical Graphics Using ODS.

If ODS Graphics is enabled but you do not specify the PLOTS= option, PROC SURVEYFREQ produces all plots that are associated with the analyses that you request, with the exception of weighted frequency plots and mosaic plots. To produce a weighted frequency plot or mosaic plot when ODS Graphics is enabled, you must specify the WTFREQPLOT or MOSAICPLOT plot-request, or you must specify the PLOTS=ALL option. PROC SURVEYFREQ produces the remaining plots (listed in Table 97.5) by default when you request the corresponding TABLES statement options.

You can suppress default plots and request specific plots by using the PLOTS(ONLY)= option; PLOTS(ONLY)=(plot-requests) produces only the plots that are specified as plot-requests. You can suppress all plots by specifying the PLOTS=NONE option.

See Figure 97.4 and Figure 97.7 for examples of plots that PROC SURVEYFREQ produces. For information about ODS Graphics, see Chapter 21: Statistical Graphics Using ODS.

Plot Requests

Table 97.5 lists the available plot-requests together with their required TABLES statement options. Descriptions of the plot-requests follow the table in alphabetical order.

Table 97.5: Plot Requests

Plot Request

Description

Required TABLES Statement Option

ALL

All plots

None

KAPPAPLOT

Kappa plot

AGREE or KAPPA ($h \times r \times r$ table)

MOSAICPLOT

Mosaic plot

Crosstabulation table request

NONE

No plots

None

ODDSRATIOPLOT

Odds ratio plot

OR ($h \times 2 \times 2$ table)

RELRISKPLOT

Relative risk plot

OR ($h \times 2 \times 2$ table)

RISKDIFFPLOT

Risk difference plot

RISK , RISK1 , or RISK2 ($h \times 2 \times 2$ table)

WTFREQPLOT

Weighted frequency plot

Frequency or crosstabulation table request

WTKAPPAPLOT

Weighted kappa plot

AGREE or WTKAPPA ($h \times r \times r$ table, r > 2)


The following plot-requests are available:

ALL

requests all plots that are associated with the specified analyses. If you specify the PLOTS=ALL option, PROC SURVEYFREQ also produces the weighted frequency and mosaic plots that are associated with the tables that you request. (PROC SURVEYFREQ does not produce weighted frequency and mosaic plots by default when ODS Graphics is enabled.)

KAPPAPLOT < (plot-options)>

requests a plot of kappa coefficients with confidence limits. Kappa plots are available for multiway square tables and display the simple kappa coefficient (with confidence limits) for each two-way table layer. To produce a kappa plot, you must specify the KAPPA or AGREE option in the TABLES statement to compute kappa coefficients.

Table 97.6 lists the plot-options that are available for kappa plots. For descriptions of the plot-options, see the subsection "Plot Options."

Table 97.6: Plot Options for KAPPAPLOT and WTKAPPAPLOT

Plot Option

Description

Values

CLDISPLAY=

Error bar type

BAR, LINE, LINEARROW,

   

SERIF$^*$, or SERIFARROW

NPANELPOS=

Statistics per graphic

Number (All$^*$)

ORDER=

Order of two-way levels

ASCENDING or DESCENDING

RANGE=

Range to display

Values or CLIP

STATS

Statistic values

None

$^*$Default


MOSAICPLOT < (plot-options)>

requests a mosaic plot. Mosaic plots are available for crosstabulation tables. For multiway tables, PROC SURVEYFREQ provides a mosaic plot for each two-way table layer.

To produce a mosaic plot, you must specify the MOSAICPLOT plot-request in the PLOTS= option, or you must specify the PLOTS=ALL option. PROC SURVEYFREQ does not produce mosaic plots by default when ODS Graphics is enabled.

Mosaic plots display tiles that correspond to the crosstabulation table cells. The areas of the tiles are proportional to the weighted frequencies of the table cells. The column variable is displayed on the X axis, and the tile widths are proportional to the relative weighted frequencies of the column variable levels. The row variable is displayed on the Y axis, and the tile heights are proportional to the relative weighted frequencies of the row levels within column levels. For more information, see Friendly (2000).

By default, the colors of the tiles correspond to the row variable levels. If you specify the COLORSTAT plot-option, the tiles are colored according to the values of the Pearson residuals.

You can specify the following plot-options:

COLORSTAT < =PEARSONRES >

colors the mosaic plot tiles according to the values of the Pearson residuals. A table cell’s Pearson residual is the square root of its contribution to the Pearson chi-square statistic. The Pearson residual is computed as $ (\emph{weighted frequency} - \emph{expected} ) ~  / ~  \sqrt {\emph{expected}} $, where weighted frequency is the weighted frequency of the table cell and expected is the expected weighted frequency. You can specify the PEARSONRES option to display the Pearson residuals in the crosstabulation table.

SQUARE

produces a square mosaic plot, where the height of the Y axis equals the width of the X axis. In a square mosaic plot, the scale of the relative weighted frequencies is the same on both axes. By default, PROC SURVEYFREQ produces a rectangular mosaic plot.

NONE

suppresses all plots.

ODDSRATIOPLOT < (plot-options)>

requests a plot of odds ratios with confidence limits. Odds ratio plots are available for multiway $2 \times 2$ tables and display the odds ratio (with confidence limits) for each $2 \times 2$ table layer. To produce an odds ratio plot, you must specify the OR option in the TABLES statement for a multiway $2 \times 2$ table.

Table 97.7 lists the plot-options that are available for odds ratio plots. For descriptions of the plot-options, see the subsection "Plot Options."

Table 97.7: Plot Options for ODDSRATIOPLOT, RELRISKPLOT, and RISKDIFFPLOT

Plot Option

Description

Values

CLDISPLAY=

Error bar type

BAR, LINE, LINEARROW,

   

SERIF$^*$, or SERIFARROW

COLUMN= $^{**}$

Risk column

1$^*$ or 2

LOGBASE= $^{***}$

Axis scale

2, E, or 10

NPANELPOS=

Statistics per graphic

Number (All$^*$)

ORDER=

Order of two-way levels

ASCENDING or DESCENDING

RANGE=

Range to display

Values or CLIP

STATS

Statistic values

None

$^*$Default

$^{**}$Available for RELRISKPLOT and RISKDIFFPLOT

$^{***}$Available for ODDSRATIOPLOT and RELRISKPLOT


RELRISKPLOT < (plot-options)>

requests a plot of relative risks with confidence limits. Relative risk plots are available for multiway $2 \times 2$ tables and display the relative risk (with confidence limits) for each $2 \times 2$ table layer. To produce a relative risk plot, you must specify the OR option in the TABLES statement for a multiway $2 \times 2$ table.

Table 97.7 lists the plot-options that are available for relative risk plots. For descriptions of the plot-options, see the subsection "Plot Options."

RISKDIFFPLOT < (plot-options)>

requests a plot of risk differences with confidence limits. Risk difference plots are available for multiway $2 \times 2$ tables and display the risk difference (with confidence limits) for each $2 \times 2$ table layer. To produce a risk difference plot, you must specify the RISK , RISK1 , or RISK2 option in the TABLES statement for a multiway $2 \times 2$ table.

Table 97.7 lists the plot-options that are available for risk difference plots. For descriptions of the plot-options, see the subsection "Plot Options."

WTFREQPLOT < (plot-options)>

requests a weighted frequency plot. Weighted frequency plots are available for frequency and crosstabulation tables. For multiway tables, PROC SURVEYFREQ provides a two-way weighted frequency plot for each two-way table layer.

To produce a weighted frequency plot, you must specify the WTFREQPLOT plot-request in the PLOTS= option, or you must specify the PLOTS=ALL option. PROC SURVEYFREQ does not produce weighted frequency plots by default when ODS Graphics is enabled.

By default, PROC SURVEYFREQ displays weighted frequency plots as bar charts. You can specify the TYPE=DOTPLOT plot-option to display frequency plots as dot plots. You can plot weighted percentages instead of frequencies by specifying the SCALE=PERCENT plot-option. There are four frequency plot layouts available, which you can request by specifying the TWOWAY= plot-option. For more information, see the subsection "Plot Options."

By default, the primary grouping of graph cells in a two-way layout is by column variable. Row variable levels are then displayed within column variable levels. You can specify the GROUPBY=ROW plot-option to group first by row variable.

Weighted frequency plots for one-way tables display confidence limits by default. For two-way tables, weighted frequency plots display confidence limits by default in the TWOWAY=GROUPVERTICAL and TWOWAY=GROUPHORIZONTAL layouts. You can suppress confidence limits by specifying the CLBAR=NO plot-option. Confidence limits are not available for two-way plots in the TWOWAY=CLUSTER and TWOWAY=STACKED layouts.

Table 97.8 lists the plot-options that are available for weighted frequency plots. For descriptions of the plot-options, see the subsection "Plot Options."

Table 97.8: Plot Options for WTFREQPLOT

Plot Option

Description

Values

CLBAR=

Confidence limit bars

NO or YES$^*$

GROUPBY= $^{**}$

Primary group

COLUMN$^*$ or ROW

NPANELPOS= $^{**}$

Sections per panel

Number (4$^*$)

ORIENT=

Orientation

HORIZONTAL or VERTICAL$^*$

SCALE=

Scale

PERCENT or WTFREQ$^*$

TWOWAY= $^{**}$

Two-way layout

CLUSTER, GROUPHORIZONTAL,

   

GROUPVERTICAL$^*$, or STACKED

TYPE=

Type

BARCHART$^*$ or DOTPLOT

$^*$Default

   

$^{**}$For two-way tables

   


WTKAPPAPLOT < (plot-options)>

requests a plot of weighted kappa coefficients with confidence limits. Weighted kappa plots are available for multiway square tables and display the weighted kappa coefficient (with confidence limits) for each two-way table layer. To produce a weighted kappa plot, you must specify the WTKAPPA or AGREE option in the TABLES statement to compute weighted kappa coefficients, and the table dimension must be greater than 2.

Table 97.6 lists the plot-options that are available for weighted kappa plots. For descriptions of the plot-options, see the subsection "Plot Options."

Global Plot Options

A global-plot-option applies to all plots for which the option is available unless it is altered by an individual plot-option. All plot-options that are listed in Table 97.8 and Table 97.7 are available as global-plot-options. The ONLY option is also available as a global-plot-option.

You can specify global-plot-options in parentheses after the PLOTS option. For example:

 plots(order=ascending stats)=(riskdiffplot oddsratioplot)
 plots(only)=wtfreqplot

In addition to the plot-options that are listed in Table 97.8 and Table 97.7, you can specify the following global-plot-option in parentheses after the PLOTS option:

ONLY

suppresses the default plots and requests only the plots that are specified as plot-requests.

Plot Options

You can specify the following plot-options in parentheses after a plot-request.

CLBAR=NO | YES

controls the confidence limit error bars in weighted frequency plots (WTFREQPLOT ). By default, CLBAR=YES, which displays confidence limits error bars; CLBAR=NO suppresses confidence limit error bars.

This plot-option applies to all weighted frequency plots except those two-way plots that are displayed in the TWOWAY=CLUSTER or TWOWAY=STACKED layout. Confidence limit error bars are not available in the TWOWAY=CLUSTER and TWOWAY=STACKED layouts.

CLDISPLAY=BAR < width > | LINE | LINEARROW | SERIF | SERIFARROW

controls the appearance of the confidence limit error bars. This plot-option is available for the following plots: KAPPAPLOT , ODDSRATIOPLOT , RELRISKPLOT , RISKDIFFPLOT , and WTKAPPAPLOT .

The default is CLDISPLAY=SERIF, which displays the confidence limits as lines with serifs. CLDISPLAY=LINE displays the confidence limits as plain lines without serifs. The CLDISPLAY=SERIFARROW and CLDISPLAY=LINEARROW plot-options display arrowheads on any error bars that are clipped by the RANGE= plot-option; if an entire error bar is cut from the plot, the plot displays an arrowhead that points toward the statistic.

CLDISPLAY=BAR displays the confidence limits as bars. By default, the width of the bars equals the size of the marker for the estimate. You can control the width of the bars and the size of the marker by specifying the value of width as a percentage of the distance between bars, $0 < \Argument{width} \le 1$. The bar might disappear when the value of width is very small.

COLUMN=1 | 2

specifies the $2 \times 2$ table column to use to compute the risk (proportion). This plot-option is available for the relative risk plot (RELRISKPLOT ) and the risk difference plot (RISKDIFFPLOT ). If you specify COLUMN=1, the plot displays the column 1 relative risks or the column 1 risk differences. Similarly, if you specify COLUMN=2, the plot displays the column 2 relative risks or risk differences.

For relative risk plots, the default is COLUMN=1. For risk difference plots, the default if COLUMN=1 if you request computation of both column 1 and column 2 risk differences with the RISK option. If you request computation of only column 1 (or only column 2) risk differences by specifying the RISK1 (or RISK2 ) option, by default the risk difference plot displays these risk differences.

GROUPBY=COLUMN | ROW

specifies the primary grouping for two-way weighted frequency plots, which you can request by specifying the WTFREQPLOT plot-request.

The default is GROUPBY=COLUMN, which groups graph cells first by column variable and displays row variable levels within column variable levels. You can specify GROUPBY=ROW to group first by row variable. In two-way and multiway table requests, the column variable is the last variable specified and forms the columns of the crosstabulation table. The row variable is the next-to-last variable specified and forms the rows of the table.

By default for a bar chart that is displayed in the TWOWAY=STACKED layout, bars correspond to the column variable levels and row levels are displayed (stacked) within each column bar. By default for a bar chart that is displayed in the TWOWAY=CLUSTER layout, bars are first grouped by column variable levels, and row levels are displayed as adjacent bars within each column-level group. You can reverse the default row and column variable groupings by specifying GROUPBY=ROW.

LOGBASE=2 | E | 10

applies to the odds ratio plot (ODDSRATIOPLOT ) and the relative risk plot (RELRISKPLOT ). This plot-option displays the odds ratio or relative risk axis on the log scale that you specify.

NPANELPOS=n

divides the plot into multiple panels that display at most $|\Argument{n}|$ statistics or sections.

If n is positive, the number of statistics or sections per panel is balanced; if n is negative, the number of statistics per panel is not balanced. For example, suppose you want to display 21 odds ratios. NPANELPOS=20 displays two panels, the first with 11 odds ratios and the second with 10 odds ratios; NPANELPOS=–20 displays 20 odds ratios in the first panel but only 1 in the second panel. This plot-option is available for all plots except mosaic plots and one-way weighted frequency plots.

For two-way weighted frequency plots (WTFREQPLOT ), NPANELPOS=n requests that panels display at most $|\Argument{n}|$ sections, where sections correspond to row or column variable levels, depending on the type of plot and the grouping. By default, n=4 and each panel includes at most four sections. This plot-option applies to two-way plots that are displayed in the TWOWAY=GROUPVERTICAL or TWOWAY=GROUPHORIZONTAL layout. The NPANELPOS= plot-option does not apply to the TWOWAY=CLUSTER and TWOWAY=STACKED layouts, which are always displayed in a single panel.

For plots that display statistics with confidence limits, NPANELPOS=n requests that panels display at most $|\Argument{n}|$ statistics. By default, n=0 and all statistics are displayed in a single panel. This plot-option applies to the following plots: KAPPAPLOT , ODDSRATIOPLOT , RELRISKPLOT , RISKDIFFPLOT , and WTKAPPAPLOT .

ORDER=ASCENDING | DESCENDING

displays the two-way table (layer) statistics in order of the statistic value. If you specify ORDER=ASCENDING or ORDER=DESCENDING, the plot displays the statistics in ascending or descending order, respectively. By default, the order of the statistics in the plot matches the order that the two-way table layers appear in the multiway table.

This plot-option is available for the following plots: KAPPAPLOT , ODDSRATIOPLOT , RELRISKPLOT , RISKDIFFPLOT , and WTKAPPAPLOT .

ORIENT=HORIZONTAL | VERTICAL

controls the orientation of weighted frequency plots (WTFREQPLOT ). This plot-option places the variable levels on the Y axis and the weighted frequencies or percentages on the X axis. ORIENT=VERTICAL places the variable levels on the X axis. The default orientation is ORIENT=VERTICAL for bar charts (TYPE=BARCHART ) and ORIENT=HORIZONTAL for dot plots (TYPE=DOTPLOT ).

RANGE=(< min > < , max > )| CLIP

specifies the range of values to display. If you specify RANGE=CLIP, the confidence limits are clipped and the display range is determined by the minimum and maximum values of the estimates. By default, the display range includes all confidence limits.

This plot-option is available for the following plots: KAPPAPLOT , ODDSRATIOPLOT , RELRISKPLOT , RISKDIFFPLOT , and WTKAPPAPLOT .

SCALE=PERCENT | WTFREQ

specifies the scale of the frequencies in weighted frequency plots (WTFREQPLOT ). SCALE=WTFREQ displays weighted frequencies (totals), and SCALE=PERCENT displays percentages. The default scale is SCALE=WTFREQ.

STATS

displays the values of the statistics and their confidence limits on the right side of the plot. If you do not specify this plot-option, the statistic values are not displayed.

This plot-option is available for the following plots: KAPPAPLOT , ODDSRATIOPLOT , RELRISKPLOT , RISKDIFFPLOT , and WTKAPPAPLOT .

TWOWAY=CLUSTER | GROUPHORIZONTAL | GROUPVERTICAL | STACKED

specifies the layout for two-way weighted frequency plots (WTFREQPLOT ).

All TWOWAY= layouts are available for bar charts (TYPE=BARCHART ). All TWOWAY= layouts except TWOWAY=CLUSTER are available for dot plots (TYPE=DOTPLOT ). Confidence limits (CLBAR= ) can be displayed in the GROUPVERTICAL and GROUPHORIZONTAL layouts. Confidence limits are not available in the STACKED and CLUSTER layouts. The ORIENT= and GROUPBY= plot-options are available for all TWOWAY= layouts.

The default two-way layout is TWOWAY=GROUPVERTICAL, which produces a grouped plot that has a vertical common baseline. By default for bar charts (TYPE=BARCHART , ORIENT=VERTICAL ), the X axis displays column variable levels, and the Y axis displays weighted frequencies. The plot includes a vertical (Y-axis) block for each row variable level. The relative positions of the graph cells in this plot layout are the same as the relative positions of the table cells in the crosstabulation table. You can reverse the default row and column grouping by specifying the GROUPBY=ROW plot-option.

The TWOWAY=GROUPHORIZONTAL layout produces a grouped plot that has a horizontal common baseline. By default (GROUPBY=COLUMN ), the plot displays a block on the X axis for each column variable level. Within each column-level block, the plot displays row variable levels.

The TWOWAY=STACKED layout produces stacked displays of weighted frequencies. By default (GROUPBY=COLUMN ) in a stacked bar chart, the bars correspond to column variable levels, and row levels are stacked within each column level. By default in a stacked dot plot, the dotted lines correspond to column levels, and cell weighted frequencies are plotted as data dots on the corresponding column line. The dot color identifies the row level.

The TWOWAY=CLUSTER layout, which is available only for bar charts, displays groups of adjacent bars. By default, the primary grouping is by column variable level, and row levels are displayed within each column level.

You can reverse the default row and column grouping in any layout by specifying the GROUPBY=ROW plot-option. The default is GROUPBY=COLUMN , which groups first by column variable.

TYPE=BARCHART | DOTPLOT

specifies the type (form) of the weighted frequency plots (WTFREQPLOT ). TYPE=BARCHART produces a bar chart and TYPE=DOTPLOT produces a dot plot. The default type is TYPE=BARCHART.

RISK
RISKDIFF

requests risk statistics for $2 \times 2$ tables. The RISK option also provides standard errors and confidence limits for these statistics. Risk statistics include the row 1 risk (proportion), row 2 risk, overall risk, and risk difference. For more information, see the section Risks and Risk Difference.

The RISK option provides both column 1 and column 2 risks. To request only column 1 or column 2 risks, use the RISK1 or RISK2 option.

You can specify the confidence level in the ALPHA= option. By default, ALPHA=0.05, which produces 95% confidence limits.

RISK1
RISKDIFF1

requests column 1 risk statistics for $2 \times 2$ tables, together with their standard errors and confidence limits. Risk statistics include the row 1 risk (proportion), row 2 risk, overall risk, and risk difference. For more information, see the section Risks and Risk Difference.

You can specify the confidence level in the ALPHA= option. By default, ALPHA=0.05, which produces 95% confidence limits.

RISK2
RISKDIFF2

requests column 2 risk statistics for $2 \times 2$ tables, together with their standard errors and confidence limits. Risk statistics include the row 1 risk (proportion), row 2 risk, overall risk, and risk difference. For more information, see the section Risks and Risk Difference.

You can specify the confidence level in the ALPHA= option. By default, ALPHA=0.05, which produces 95% confidence limits.

ROW < (option)>

displays the row percentage (estimated proportion of the row total) for each cell in a two-way table. The ROW option also provides the standard errors of the row percentages. For more information, see the section Row and Column Proportions. This option has no effect for one-way tables.

You can specify the following option:

DEFF

displays the design effect for each row percentage in the crosstabulation table. For more information, see the section Design Effect.

TESTP=(values)

specifies null hypothesis proportions (test percentages) for chi-square tests for one-way tables (goodness-of-fit tests). You can separate values with blanks or commas, and you must enclose the list of values in parentheses. Specify values in probability form as numbers between 0 and 1, where the proportions sum to 1. Or specify values in percentage form as numbers between 0 and 100, where the percentages sum to 100. PROC SURVEYFREQ treats the value 1 as the percentage form 1%. The number of TESTP= values must equal the number of variable levels in the one-way table. List these values in the same order in which the corresponding variable levels appear in the output.

When you specify the TESTP= option, PROC SURVEYFREQ displays the specified test percentages in the one-way frequency table. The TESTP= option has no effect for two-way tables.

PROC SURVEYFREQ uses the TESTP= values for the one-way Rao-Scott chi-square test (CHISQ ) and for the one-way Rao-Scott likelihood ratio chi-square test (LRCHISQ ). See the sections Rao-Scott Chi-Square Test and For more information, see the section Rao-Scott Likelihood Ratio Chi-Square Test.

VAR

displays the variance estimate for each percentage in the crosstabulation table. For more information, see the section Proportions. By default, PROC SURVEYFREQ displays the standard errors of the percentages.

VARWT

displays the variance estimate for each weighted frequency, or estimated total, in the crosstabulation table. For more information, see the section Totals. By default, PROC SURVEYFREQ displays the standard deviations of the weighted frequencies.

WCHISQ

requests the Wald chi-square test for two-way tables. For more information, see the section Wald Chi-Square Test.

WLLCHISQ

requests the Wald log-linear chi-square test for two-way tables. For more information, see the section Wald Log-Linear Chi-Square Test.

WTFREQ

displays totals (weighted frequencies) and their standard errors when you do not specify a WEIGHT or REPWEIGHTS statement. By default, PROC SURVEYFREQ displays the weighted frequencies only when you specify a WEIGHT or REPWEIGHTS statement. When you do not specify a WEIGHT or REPWEIGHTS statement, PROC SURVEYFREQ assigns all observations a weight of one.

WTKAPPA < (options)>

requests the weighted kappa coefficient with its standard error and confidence limits. Weighted kappa coefficients can be computed for square two-way tables, where the number of rows equals the number of columns. For $2 \times 2$ tables, the weighted kappa coefficient equals the simple kappa coefficient, and PROC SURVEYFREQ displays only the simple kappa coefficient. For more information, see the section Weighted Kappa Coefficient.

Weighted kappa coefficients are available when you specify variance estimation by the jackknife method (VARMETHOD=JACKKNIFE ) or by balanced repeated replication (VARMETHOD=BRR ); weighted kappa coefficients are not available with the Taylor series method (VARMETHOD=TAYLOR ).

The weighted kappa coefficient is computed by using agreement weights that reflect the relative agreement between pairs of variable levels. Agreement weights are not the same as sampling weights, which you provide by specifying the WEIGHT statement. PROC SURVEYFREQ uses the sampling weights to compute both the simple kappa and weighted kappa coefficients. For more information, see the section Weighted Kappa Coefficient.

You can specify the confidence level in the ALPHA= option. By default, ALPHA=0.05, which produces 95% confidence limits.

You can specify the following options:

PRINTKWTS

displays the agreement weights that PROC SURVEYFREQ uses to compute the weighted kappa coefficient. Agreement weights reflect the relative agreement between pairs of variable levels. By default, PROC SURVEYFREQ computes these weights by using the Cicchetti-Allison form. If you specify the WT=FC option, the procedure uses the Fleiss-Cohen form of agreement weights. For more information, see the section Weighted Kappa Coefficient.

WT=FC

requests Fleiss-Cohen agreement weights for the weighted kappa computation. By default, PROC SURVEYFREQ uses Cicchetti-Allison agreement weights to compute the weighted kappa coefficient. Agreement weights reflect the relative agreement between pairs of variable levels. For more information, see the section Weighted Kappa Coefficient.