Syntax: LSMEANS Statement 
LSmeans can be computed for any effect in the statistical model that involves only CLASS variables. You can specify multiple effects in one LSMEANS statement or in multiple LSMEANS statements, and all LSMEANS statements must appear after the MODEL statement. If you do not specify modeleffects, the options in the LSMEANS statement are applied to all suitable model effects.
As in the ESTIMATE statement, the matrix is tested for estimability; if this test fails, the procedure displays "Nonest" for the LSmeans entries. Note that linear functions of LSmeans, such as differences, can be estimable, even if the means themselves are not estimable. Estimability checks for differences are thus applied separately from checks for the means.
Assuming the LSmean is estimable, the procedure constructs an approximate t test to test the null hypothesis that the associated population quantity equals zero.
Table 19.20 summarizes important options in the LSMEANS statement. All LSMEANS options are subsequently discussed in alphabetical order.
Option 
Description 

Construction and Computation of LSMeans 

Modifies the covariate value in computing LSmeans 

Computes separate margins 

Requests differences of LSmeans 

Specifies the weighting scheme for LSmeans computation as determined by the input data set 

Tunes estimability checking 

Degrees of Freedom and pvalues 

Determines the method for multiple comparison adjustment of LSmeans differences 

Determines the confidence level () 

Adjusts multiple comparison pvalues further in a stepdown fashion 

Statistical Output 

Constructs confidence limits for means and mean differences 

Displays the correlation matrix of LSmeans 

Displays the covariance matrix of LSmeans 

Prints the matrix 

Produces a "Lines" display for pairwise LSmeans differences 

Prints the LSmeans 

Requests ODS statistical graphics of means and mean comparisons 

Specifies the seed for computations that depend on random numbers 

Generalized Linear Modeling 

Exponentiates and displays estimates of LSmeans or LSmeans differences 

Computes and displays estimates and standard errors of LSmeans (but not differences) on the inverse linked scale 

Reports (simple) differences of least squares means in terms of odds ratios if permitted by the link function 
You can specify the following options in the LSMEANS statement after a slash (/):
specifies how denominator degrees of freedom are determined when values and confidence limits are adjusted for multiple comparisons with the ADJUST= option. When you do not specify the ADJDFE= option or when you specify ADJDFE=SOURCE, the denominator degrees of freedom for multiplicityadjusted results are the denominator degrees of freedom for the LSmean effect in the "Type III Tests of Fixed Effects" table. When you specify ADJDFE=ROW, the denominator degrees of freedom for multiplicityadjusted results correspond to the degrees of freedom that are displayed in the DF column of the "Differences of Least Squares Means" table.
The ADJDFE=ROW setting is particularly useful if you want multiplicity adjustments to take into account that denominator degrees of freedom are not constant across LSmean differences.
In oneway models with heterogeneous variance, combining certain ADJUST= options with the ADJDFE=ROW option corresponds to particular methods of performing multiplicity adjustments in the presence of heteroscedasticity. For example, the following statements fit a heteroscedastic oneway model and perform Dunnett’s T3 method (Dunnett 1980), which is based on the studentized maximum modulus (ADJUST=SMM):
proc glimmix; class A; model y = A / ddfm=satterth; random _residual_ / group=A; lsmeans A / adjust=smm adjdfe=row; run;
If you combine the ADJDFE=ROW option with ADJUST=SIDAK, the multiplicity adjustment corresponds to the T2 method of Tamhane (1979), and ADJUST=TUKEY corresponds to the method of GamesHowell (Games and Howell 1976). Note that ADJUST=TUKEY gives the exact results for the case of fractional degrees of freedom in the oneway model, but it does not take into account that the degrees of freedom are subject to variability. A more conservative method, such as ADJUST=SMM, might protect the overall error rate better.
Unless the ADJUST= option is specified in the LSMEANS statement, the ADJDFE= option has no effect. The option is not supported by the procedures that perform chisquarebased inference (GENMOD, LOGISTIC, PHREG, and SURVEYLOGISTIC).
requests a multiple comparison adjustment for the pvalues and confidence limits for the differences of LSmeans. The adjusted quantities are produced in addition to the unadjusted quantities. By default, the procedure performs all pairwise differences. If you specify ADJUST=DUNNETT, the procedure analyzes all differences with a control level. If you specify ADJUST=NELSON, ANOM differences are taken. The ADJUST= option implies the DIFF option.
The BON (Bonferroni) and SIDAK adjustments involve correction factors described in Chapter 41, The GLM Procedure, and Chapter 60, The MULTTEST Procedure ; also see Westfall and Young (1993) and Westfall et al. (1999). When you specify ADJUST=TUKEY and your data are unbalanced, the procedure uses the approximation described in Kramer (1956) and identifies the adjustment as "TukeyKramer" in the results. Similarly, when you specify ADJUST=DUNNETT or ADJUST=NELSON and the LSmeans are correlated, the procedure uses the factoranalytic covariance approximation described in Hsu (1992) and identifies the adjustment in the results as "DunnettHsu" or "NelsonHsu," respectively. The approximation derives an approximate "effective sample sizes" for which exact critical values are computed. Computing the exact adjusted values and critical values for unbalanced designs can be computationally intensive, in particular for ADJUST=NELSON. A simulationbased approach, as specified by the ADJUST=SIM option, while nondeterministic, can provide inferences that are sufficiently accurate in much less time. The preceding references also describe the SCHEFFE and SMM adjustments.
Nelson’s adjustment applies only to the analysis of means (Ott 1967; Nelson 1982, 1991, 1993), where LSmeans are compared against an average LSmean. It does not apply to all pairwise differences of least squares means. See the DIFF=ANOM option for more details regarding the analysis of means with the procedure.
The SIMULATE adjustment computes adjusted pvalues and confidence limits from the simulated distribution of the maximum or maximum absolute value of a multivariate t random vector. All covariance parameters, except the residual scale parameter, are fixed at their estimated values throughout the simulation, potentially resulting in some underdispersion. The simulation estimates , the true th quantile, where is the confidence coefficient. The default is 0.05, and you can change this value with the ALPHA= option in the LSMEANS statement.
The number of samples is set so that the tail area for the simulated is within of with % confidence. In equation form,
where is the simulated and is the true distribution function of the maximum; see Edwards and Berry (1987) for details. By default, = 0.005 and = 0.01, placing the tail area of within 0.005 of 0.95 with 99% confidence. You can specify the following simoptions in parentheses after the ADJUST=SIMULATE option:
specifies the target accuracy radius of a % confidence interval for the true probability content of the estimated th quantile. The default value is ACC=0.005.
specifies the value for a % confidence interval for the true probability content of the estimated th quantile. The default value for the accuracy confidence is 99%, which corresponds to EPS=0.01.
specifies the sample size for the simulation. By default, is set based on the values of the target accuracy radius and accuracy confidence % for an interval for the true probability content of the estimated th quantile. With the default values for , , and (0.005, 0.01, and 0.05, respectively), NSAMP=12,604 by default.
specifies an integer that is used to start the pseudorandom number generator for the simulation. If you do not specify a seed, or specify a value less than or equal to zero, the seed is by default generated from reading the time of day from the computer’s clock.
specifies that the computational work for the simulation be divided into parallel threads, where the number of threads is the value of the SAS system option CPUCOUNT=. For large simulations (as specified directly using the NSAMP= simoption or indirectly using the ACC= or EPS= simoptions), parallel processing can markedly speed up the computation of adjusted pvalues and confidence intervals. However, because the parallel processing has different pseudorandom number streams, the precise results are different from the default ones, which are computed in sequence rather than in parallel. This option overrides the SAS system option THREADS  NOTHREADS.
specifies that the computational work for the simulation be performed in sequence rather than in parallel. NOTHREADS is the default. This option overrides the SAS system option THREADS  NOTHREADS.
If the STEPDOWN option is in effect, the pvalues are further adjusted in a stepdown fashion. For certain options and data, this adjustment is exact under an iid model for the dependent variable, in particular for the following:
for ADJUST=DUNNETT when the means are uncorrelated
for ADJUST=TUKEY with STEPDOWN(TYPE=LOGICAL) when the means are balanced and uncorrelated.
The first case is a consequence of the nature of the successive stepdown hypotheses for comparisons with a control; the second uses an extension of the maximum studentized range distribution appropriate for partition hypotheses (Royen 1989). Finally, for STEPDOWN(TYPE=FREE), ADJUST=TUKEY employs the Royen (1989) extension in such a way that the resulting pvalues are conservative.
requests that a t type confidence interval be constructed for each of the LSmeans with confidence level . The value of number must be between 0 and 1; the default is 0.05.
modifies the values of the covariates that are used in computing LSmeans. By default, all covariate effects are set equal to their mean values for computation of standard LSmeans. The AT option enables you to assign arbitrary values to the covariates. Additional columns in the output table indicate the values of the covariates.
If there is an effect that contains two or more covariates, the AT option sets the effect equal to the product of the individual means rather than the mean of the product (as with standard LSmeans calculations). The AT MEANS option sets covariates equal to their mean values (as with standard LSmeans) and incorporates this adjustment to crossproducts of covariates.
As an example, consider the following statements:
class A; model Y = A x1 x2 x1*x2; lsmeans A; lsmeans A / at means; lsmeans A / at x1=1.2; lsmeans A / at (x1 x2)=(1.2 0.3);
For the first two LSMEANS statements, the LSmeans coefficient for x1 is (the mean of x1) and for x2 is (the mean of x2). However, for the first LSMEANS statement, the coefficient for x1*x2 is , but for the second LSMEANS statement, the coefficient is . The third LSMEANS statement sets the coefficient for x1 equal to and leaves it at for x2, and the final LSMEANS statement sets these values to and , respectively.
Even if you specify a WEIGHT variable, the unweighted covariate means are used for the covariate coefficients if there is no AT specification. If you specify the AT option, WEIGHT or FREQ variables are taken into account as follows. The weighted covariate means are then used for the covariate coefficients for which no explicit AT values are given, or if you specify AT MEANS. Observations that do not contribute to the analysis because of a missing dependent variable are included in computing the covariate means. Use the E option in conjunction with the AT option to check that the modified LSmeans coefficients are the ones you want.
The AT option is disabled if you specify the BYLEVEL option.
requests that separate margins be computed for each level of the LSMEANS effect.
The standard LSmeans have equal coefficients across classification effects. The BYLEVEL option changes these coefficients to be proportional to the observed margins. This adjustment is reasonable when you want your inferences to apply to a population that is not necessarily balanced but has the margins observed in the input data set. In this case, the resulting LSmeans are actually equal to raw means for fixedeffects models and certain balanced randomeffects models, but their estimated standard errors account for the covariance structure that you have specified. If a WEIGHT statement is specified, the procedure uses weighted margins to construct the LSmeans coefficients.
If the AT option is specified, the BYLEVEL option disables it.
requests that t type confidence limits be constructed for each of the LSmeans. The confidence level is 0.95 by default; this can be changed with the ALPHA= option. If you specify an ADJUST= option, then the confidence limits are adjusted for multiplicity. But if you also specify STEPDOWN, then only pvalues are stepdown adjusted, not the confidence limits.
displays the estimated correlation matrix of the least squares means as part of the "Least Squares Means" table.
displays the estimated covariance matrix of the least squares means as part of the "Least Squares Means" table.
specifies the degrees of freedom for the t test and confidence limits. The default is the denominator degrees of freedom taken from the "Type III Tests" table that corresponds to the LSmeans effect. The option is not supported by the procedures that perform chisquarebased inference (GENMOD, LOGISTIC, PHREG and SURVEYLOGISTIC).
requests that differences of the LSmeans be displayed. You can use one of the following optional difftype values to specify which differences to produce:
requests all pairwise differences; this is the default.
requests differences between each LSmean and the average LSmean, as in the analysis of means (Ott 1967). The average is computed as a weighted mean of the LSmeans, the weights being inversely proportional to the diagonal entries of the matrix. If LSmeans are nonestimable, this designbased weighted mean is replaced with an equally weighted mean. Note that the ANOM procedure in SAS/QC software implements both tables and graphics for the analysis of means with a variety of response types. For oneway designs and normal data with identity link, the DIFF=ANOM computations are equivalent to the results of PROC ANOM. If the LSmeans being compared are uncorrelated, exact adjusted pvalues and critical values for confidence limits can be computed in the analysis of means; see Nelson (1982, 1991, 1993) and Guirguis and Tobias (2004) in addition to the documentation for the ADJUST=NELSON option.
requests differences with a control, which, by default, is the first valid level of each of the specified LSMEANS effects. For example, suppose the effects A and B are classification variables, both of them have two levels 1 and 2, and the A=1, B=1 cell is missing. Unless the procedure supports a MISSING option in the CLASS statement and the option is in effect, the following LSMEANS statement uses the level (1,2) of A*B as the control:
lsmeans A*B / diff=control;
Nevertheless, you can still specify a valid level as the control—for example, (2,1) of A*B. To specify which levels of the effects are the controls, list the quoted formatted values in parentheses after the CONTROL keyword. For example, if the effects A, B, and C are classification variables, each having two levels, 1 and 2, the following LSMEANS statement specifies the (1,2) level of A*B and the (2,1) level of B*C as controls:
lsmeans A*B B*C / diff=control('1' '2' '2' '1');
For multiple effects, the results depend upon the order of the list, and so you should check the output to make sure that the controls are correct.
Twotailed tests and confidence limits are associated with the CONTROL difftype. For onetailed results, use either the CONTROLL or CONTROLU difftype.
tests whether the noncontrol levels are significantly smaller than the control; the upper confidence limits for the control minus the noncontrol levels are considered to be infinity and are displayed as missing.
tests whether the noncontrol levels are significantly larger than the control; the upper confidence limits for the noncontrol levels minus the control are considered to be infinity and are displayed as missing.
If you want to perform multiple comparison adjustments on the differences of LSmeans, you must specify the ADJUST= option.
The differences of the LSmeans are displayed in a table titled "Differences of Least Squares Means."
requests that the matrix coefficients for the LSMEANS effects be displayed.
requests exponentiation of the LSmeans or LSmean differences. When you model data with the logit, cumulative logit, or generalized logit link functions, and the estimate represents a log odds ratio or log cumulative odds ratio, the EXP option produces an odds ratio. In proportional hazards model, the exponentiation of the LSmean differences produces estimates of hazard ratios. If you specify the CL or ALPHA= option, the (adjusted) confidence bounds are also exponentiated.
The EXP option is supported only by PROC PHREG, PROC SURVEYPHREG, the procedures that support generalized linear modeling (GENMOD, LOGISTIC, and SURVEYLOGISTIC), and PROC PLM when it is used to perform statistical analyses on item stores that are created by these procedures.
requests that estimates and their standard errors in the "Least Squares Means" table also be reported on the scale of the mean (the inverse linked scale). This enables you to obtain estimates of predicted probabilities and their standard errors in logistic models, for example. The option is specific to an LSMEANS statement. If you also specify the CL option, the procedure computes confidence intervals for the predicted means by applying the inverse link transform to the confidence limits on the linked (linear) scale. Standard errors on the inverse linked scale are computed by the delta method.
The ILINK option is supported only by the procedures that support generalized linear modeling (GENMOD, LOGISTIC and SURVEYLOGISTIC) and by PROC PLM when it is used to perform statistical analyses on item stores that are created by these procedures.
presents results of comparisons between all pairs of least squares means by listing the means in descending order and indicating nonsignificant subsets by line segments beside the corresponding LSmeans. When all differences have the same variance, these comparison lines are guaranteed to accurately reflect the inferences that are based on the corresponding tests, which are made by comparing the respective pvalues to the value of the ALPHA= option (0.05 by default). However, equal variances might not be the case for differences between LSmeans. If the variances are not all the same, then the comparison lines might be conservative, in the sense that if you base your inferences on the lines alone, you will detect fewer significant differences than the tests indicate. If there are any such differences, the procedure lists the pairs of means that are inferred to be significantly different by the tests but not by the comparison lines. However, even though the variances in many cases are unequal, they are similar enough that the comparison lines accurately reflect the test inferences.
determines whether to print the least squares means themselves. For most procedure, MEANS is the default behavior. For example, the NOMEANS option is the default for the PHREG procedure. You can then use the MEANS option to produce the table of least squares means, if desired.
requests that LSmean differences (DIFF, ADJUST= options) are also reported in terms of odds ratios. The ODDSRATIO option is ignored unless you use either the logit, cumulative logit, or generalized logit link function. If you specify the CL or ALPHA= option, confidence intervals for the odds ratios are also computed. These intervals are adjusted for multiplicity when you specify the ADJUST= option.
The ODDSRATIO option is supported only by the procedures that support generalized linear modeling (GENMOD, LOGISTIC and SURVEYLOGISTIC) and by PROC PLM when it is used to perform statistical analyses on item stores created by these procedures.
specifies a potentially different weighting scheme for the computation of LSmeans coefficients. The standard LSmeans have equal coefficients across classification effects; however, the OM option changes these coefficients to be proportional to those found in the OMdataset. This adjustment is reasonable when you want your inferences to apply to a population that is not necessarily balanced but has the margins that are observed in OMdataset.
By default, OMdataset is the same as the analysis data set. You can optionally specify another data set that describes the population for which you want to make inferences. This data set must contain all model variables except for the dependent variable (which is ignored if it is present). In addition, the levels of all CLASS variables must be the same as those that occur in the analysis data set. If a level of a classification effect in the original data set is not present in the OMdataset, the LSmeans for that level are undefined. The corresponding rows of the LSMeans table are displayed as missing. Specifying an OMdataset enables you to construct arbitrarily weighted LSmeans.
In computing the observed margins, the procedure uses all observations for which there are no missing or invalid independent variables, including those for which there are missing dependent variables. Also, if you use a WEIGHT statement, the procedure computes weighted margins to construct the LSmeans coefficients. If your data are balanced, the LSmeans are unchanged by the OM option.
The BYLEVEL option modifies the observedmargins LSmeans. Instead of computing the margins across all of the OMdataset, the procedure computes separate margins for each level of the LSMEANS effect in question. In this case the resulting LSmeans are actually equal to raw means for fixedeffects models and certain balanced randomeffects models, but their estimated standard errors account for the covariance structure that you have specified.
You can use the E option in conjunction with either the OM or BYLEVEL option to verify that the modified LSmeans coefficients are the ones you want. It is possible that the modified LSmeans are not estimable when the standard ones are estimable, or vice versa.
is the same as the DIFF option.
requests that graphics related to least squares means be produced via ODS Graphics, provided that ODS Graphics is enabled and the plotrequest does not conflict with other options in the LSMEANS statement. For general information about ODS Graphics, see Chapter 21, Statistical Graphics Using ODS.
The available options and suboptions are as follows:
requests that the default plots that correspond to this LSMEANS statement be produced. The default plot depends on the options in the statement.
requests an analysisofmeans display in which least squares means are compared to an average least squares mean. Least squares mean ANOM plots are produced only for those model effects that are listed in LSMEANS statements and have options that do not contradict with the display. For example, the following statements produce analysisofmean plots for effects A and C:
lsmeans A / diff=anom plot=anom; lsmeans B / diff plot=anom; lsmeans C / plot=anom;
The DIFF option in the second LSMEANS statement implies all pairwise differences.
produces box plots of the distribution of the least squares mean or least squares mean differences across a posterior sample. For example, this plot is available in procedures that support a Bayesian analysis through the BAYES statement.
A separate box is generated for each estimable function, and all boxes appear on a single graph by default. You can affect the appearance of the box plot graph with the following options:
specifies the orientation of the boxes. The default is vertical orientation of the box plots.
specifies how to break the series of box plots across multiple panels. If the NPANELPOS option is not specified, or if number equals zero, then all box plots are displayed in a single graph; this is the default. If a negative number is specified, then exactly up to number of box plots are displayed per panel. If number is positive, then the number of boxes per panel is balanced to achieve small variation in the number of box plots per graph.
requests a display in which least squares means are visually compared against a reference level. These plots are produced only for statements with options that are compatible with control differences. For example, the following statements produce control plots for effects A and C:
lsmeans A / diff=control('1') plot=control; lsmeans B / diff plot=control; lsmeans C plot=control;
The DIFF option in the second LSMEANS statement implies all pairwise differences.
requests a display of all pairwise least squares mean differences and their significance. The display is also known as a "meanmean scatter plot" when it is based on arithmetic means (Hsu 1996 and Hsu and Peruggia 1994). For each comparison a line segment, centered at the LSmeans in the pair, is drawn. The length of the segment corresponds to the projected width of a confidence interval for the least squares mean difference. Segments that fail to cross the 45degree reference line correspond to significant least squares mean differences.
LSmean difference plots are produced only for statements with options that are compatible with the display. For example, the following statements request differences against a control level for the A effect, all pairwise differences for the B effect, and the least squares means for the C effect:
lsmeans A / diff=control('1') plot=diff; lsmeans B / diff plot=diff; lsmeans C plot=diff;
The DIFF= type in the first statement is incompatible with a display of all pairwise differences.
You can specify the following diffplotoptions:
determines the positioning of the line segments in the plot. This is the default diffplotoptions. When the ABS option is in effect, all line segments are shown on the same side of the reference line.
determines the positioning of the line segments in the plot. The NOABS option separates comparisons according to the sign of the difference.
marks the center point for each comparison. This point corresponds to the intersection of two least squares means.
suppresses the display of the line segments that represent the confidence bounds for the differences of the least squares means. The NOLINES option implies the CENTER option. The default is to draw line segments in the upper portion of the plot area without marking the center point.
generates panels of histograms with a kernel density overlaid if the analysis has access to a set of posterior parameter estimates. For example, this plot is available in procedures that support a Bayesian analysis through the BAYES statement. A separate plot in each panel contains the results for each least squares mean or least squares mean differences. You can sepcify the following distplotoptions in parentheses:
controls the display of a horizontal box plot of the estimable function’s distribution across the posterior sample below the graph. The BOX option is enabled by default.
controls the display of the histogram of the estimable function’s distribution across the posterior sample. The HIST option is enabled by default.
controls the display of a normal density estimate on the graph. The NONORMAL option is enabled by default.
controls the display of a kernel density estimate on the graph. The KERNEL option is enabled by default.
specifies the highest number of rows in a panel. The default is 3.
specifies the highest number of columns in a panel. The default is 3.
unpacks the panel into separate graphics.
requests displays of the least squares means.
The following meanplotoptions control the display of the least squares means.
displays the least squares means in ascending order. This option has no effect if means are displayed in separate plots.
displays upper and lower confidence limits for the least squares means. By default, 95% limits are drawn. You can change the confidence level with the ALPHA= option. Confidence limits are drawn by default if the CL option is specified in the LSMEANS statement.
displays confidence limits as bands. This option implies the JOIN option.
displays the least squares means in descending order. This option has no effect if means are displayed in separate plots.
requests that means (and confidence limits) be displayed on the inverse linked scale.
connects the least squares means with lines. This option is implied by the CLBAND option. If the effect contains nested variables and a SLICEBY= effect contains classification variables that appear as crossed effects, this option is ignored.
specifies an effect by which to group the means in a single plot. For example, the following statement requests a plot in which the levels of A are placed on the horizontal axis and the means that belong to the same level of B are joined by lines:
lsmeans A*B / plot=meanplot(sliceby=b join);
Unless the LSmean effect contains at least two classification variables, the SLICEBY= option has no effect. The fixedeffect does not have to be an effect in your MODEL statement, but it must consist entirely of classification variables and it must be contained in the LSmean effect.
specifies an effect by which to break interaction plots into separate displays. For example, the following statement requests for each level of C one plot of the A*B cell means that are associated with that level of C:
lsmeans A*B*C / plot=meanplot(sliceby=b plotby=c clband);
In each plot, levels of A are displayed on the horizontal axis, and confidence bands are drawn around the means that share the same level of B.
The PLOTBY= option has no effect unless the LSmean effect contains at least three classification variables. The fixedeffect does not have to be an effect in the MODEL statement, but it must consist entirely of classification variables and it must be contained in the LSmean effect.
requests that no plots be produced.
When LSmean calculations are adjusted for multiplicity by using the ADJUST= option, the plots are adjusted accordingly.
specifies the seed for the samplingbased components of the computations for the LSMEANS statement (for example, chibarsquare statistics and simulated pvalues). number specifies an integer that is used to start the pseudorandomnumber generator for the simulation. If you do not specify a seed, or if you specify a value less than or equal to zero, the seed is generated from reading the time of day from the computer clock. Note that there could be multiple LSMEANS statements with SEED= specifications and there could be other statements that can supply a random number seed. Since the procedure has only one random number stream, the initial seed is shown in the SAS log.
tunes the estimability checking. If is a vector, define ABS() to be the largest absolute value of the elements of . If ABS() is greater than *number for any row of in the contrast, then is declared nonestimable. Here, is the Hermite form matrix , and is ABS(), except when it equals 0, and then is 1. The value for number must be between 0 and 1; the default is 1E4.
requests that multiple comparison adjustments for the pvalues of LSmean differences be further adjusted in a stepdown fashion. Stepdown methods increase the power of multiple comparisons by taking advantage of the fact that a pvalue is never declared significant unless all smaller pvalues are also declared significant. The STEPDOWN adjustment combined with ADJUST=BON corresponds to the methods of Holm (1979) "Method 2" of Schaffer (1986); this is the default. Using stepdownadjusted pvalues combined with ADJUST=SIMULATE corresponds to the method of Westfall (1997).
If the denominator degrees of freedom are computed by the KenwardRoger (Kenward and Roger 1997) or Satterthwaite method in a mixed model, then stepdownadjusted pvalues are produced only if the ADJDFE=ROW option is in effect.
Also, STEPDOWN affects only pvalues, not confidence limits. For ADJUST=SIMULATE, the generalized least squares hybrid approach of Westfall (1997) is used to increase Monte Carlo accuracy.
specifies the time (in seconds) to be spent computing the maximal logically consistent sequential subsets of equality hypotheses for TYPE=LOGICAL. The default is MAXTIME=60. If the MAXTIME value is exceeded, the adjusted tests are not computed. When this occurs, you can try increasing the MAXTIME value. However, note that there are common multiple comparisons problems for which this computation requires a huge amount of time—for example, all pairwise comparisons between more than 10 groups. In such cases, try to use TYPE=FREE (the default) or TYPE=LOGICAL() for small .
specifies that a report on the stepdown adjustment be displayed, including a listing of the sequential subsets (Westfall 1997) and, for ADJUST=SIMULATE, the stepdown simulation results.
specifies how stepdown adjustment are made. If you specify TYPE=LOGICAL, the stepdown adjustments are computed by using maximal logically consistent sequential subsets of equality hypotheses (Shaffer 1986, Westfall 1997). Alternatively, for TYPE=FREE, sequential subsets are computed ignoring logical constraints. The TYPE=FREE results are more conservative than those for TYPE=LOGICAL, but they can be much more efficient to produce for many comparisons. For example, it is not feasible to take logical constraints between all pairwise comparisons of more than 10 groups. For this reason, TYPE=FREE is the default.
However, you can reduce the computational complexity of taking logical constraints into account by limiting the depth of the search tree used to compute them, specifying the optional depth parameter as a number in parentheses after TYPE=LOGICAL. As with TYPE=FREE, results for TYPE=LOGICAL() are conservative relative to the true TYPE=LOGICAL results. But even for TYPE=LOGICAL() they can be appreciably less conservative than TYPE=FREE, and they are computationally feasible for much larger numbers of comparisons. If you do not specify or if , the full search tree is used.