Assigning Data to Roles

To run the Multidimensional Preferences Analysis task, you must select an input data source. In the input data source, each subject (also referred to as a rater) must be a separate column, and each object that is being rated must be a separate row. Therefore, the data is a transpose of the usual multivariate data matrix. In other words, the columns are the people. In a more typical matrix, the rows represent people. To filter the input data source, click Filter Icon

You must assign at least two variables to the Raters role.

Option Name	Description
Roles
Raters	specifies the variables to analyze.
Level of measurement	specifies the level of measurement for the analysis variables.
Transformation	specifies the transformation to use in the analysis. The transformation options depend on the value selected from the Level of measurement drop-down list. If you select Interval from the Level of measurement drop-down list, these options are available: Linear specifies the optimal linear transformation of each variable. For variables with no missing values, the transformed variable is the same as the original variable. For variables with missing values, the transformed nonmissing values have a different scale and origin than the original values. B-spline finds a B-spline transformation (De Boor 1978) of each variable. Monotonic B-spline (mspline). No transformation (identity). If you select Ordinal from the Level of measurement drop-down list, these options are available: Monotonic finds a monotonic transformation of each variable, with the restriction that ties are preserved. The Kruskal (1964) secondary least squares monotonic transformation is used. This transformation weakly preserves order and category membership (ties). Rank transforms variables to ranks. Ranks are averaged within ties. The smallest input value is assigned the smallest rank. If you select Nominal from the Level of measurement drop-down list, the optimal scoring transformation is used. No transformation options are available.
Degree of spline	specifies the degree of the B-spline transformation. The degree must be a nonnegative integer. The defaults are 3 degrees for B-spline variables and 2 degrees for monotonic B-spline variables. Note: This option is available only for B-spline and Monotonic B-spline transformations.
Number of knots	creates n knots, the first at the percentile, the second at the percentile, and so on. Knots are always placed at data values; there is no interpolation. For example, if the number of knots is 3, knots are placed at the 25th percentile, the median, and the 75th percentile. Note: This option is available only for B-spline and Monotonic B-spline transformations.
Larger values for preferences	specifies whether to reflect the transformation after the iterations are completed and before the final standardization and results calculations. By default, the Higher preferences option is not selected and the transformation is not reflected.
Additional Roles
Frequency count	lists a numeric variable whose value represents the frequency of the observation. If you assign a variable to this role, the task assumes that each observation represents n observations, where n is the value of the frequency variable. If n is not an integer, SAS truncates it. If n is less than 1 or is missing, the observation is excluded from the analysis. The sum of the frequency variable represents the total number of observations.
Group analysis by	obtains separate analyses of observations in each unique group.