MTABULATE procedure

Forms tables classified by multiple-response factors (R.W. Payne).

Options

PRINT = string

Controls printed output (counts, totals, nobservations, means, minima, maxima, variances, quantiles, sds, skewness, kurtosis, semeans); default * i.e. none

CLASSIFICATION = factors

Non multiple-response factors classifying the tables

MRESPONSE = pointers

Pointers to factors defining the multiple-responses for the tables

MRFACTOR = identifiers

Identifier of factors to index the sets of multiple responses in the tables

COUNTS = table

Saves a table counting the number of units with each factor combination; default *

MARGINS = string

Whether the tables should be given margins (yes, no); default no

WEIGHTS = variate

Weights to be used in the tabulations; default * indicates that all units have weight 1

PERCENTQUANTILES = scalar or variate

Percentages for which quantiles are required; default 50 i.e. median

Parameters

DATA = variates

Data values to be tabulated

TOTALS = tables

Tables to contain totals

NOBSERVATIONS = tables

Tables containing the numbers of non-missing values in each cell

MEANS = tables

Tables of means

MINIMA = tables

Tables of minimum values in each cell

MAXIMA = tables

Tables of maximum values in each cell

VARIANCES = tables

Tables of cell variances

QUANTILES = tables or pointers

Table to contain quantiles at a single PERCENTQUANTILE, or pointer of pointers to tables for several PERCENTQUANTILES

SDS = tables

Tables of standard deviations

SKEWNESS = tables

Tables of skewness coefficients

KURTOSIS = tables

Tables of kurtosis coefficients

SEMEANS = tables

Tables of standard errors of means

Description

Multiple responses occur in surveys as the result of open-ended questions like "Which cities have you visited?". In GenStat, these can be formed by the FMFACTORS procedure and are represented by a pointer containing a factor for each possible response code. The factors have levels 0 and 1, and corresponding labels 'absent' and 'present'. If the original response codes were textual, the various strings are used as labels of the pointer; while if they were numerical, the numbers are used as the pointer suffixes.

The multiple responses for the tables are specified by the MRESPONSE option, while any ordinary factors are specified by the CLASSIFICATION option. The MARGINS option indicates whether or not the tables are to contain margins. For the multiple responses, these represent summaries not over the responses but over the respondents (who may each have given several responses). MTABULATE needs to generate an ordinary factor to classify the dimension of the tables corresponding to each set of multiple responses. You can supply identifiers for these factors (thus allowing them to be accessed outside the procedure), using the MRFACTOR option.

The other options and parameters are similar to those of the TABULATE directive. The COUNTS option can save a table containing the frequencies of the various responses. The DATA parameter provides information about the respondents who made the multiple responses. (So, for example, you could set DATA to the incomes of the respondents and then tabulate the average incomes of the people who have visited each of the cities.) The other parameters allow you to save the various types of numerical summary: totals, numbers of non-missing values, means, minima, maxima, variances, quantiles, standard deviations, skewness and kurtosis coefficients and (within-cell) standard errors of means.

The PERCENTQUANTILES option specifies which quantiles you want. By default just the median (the 50% quantile) is produced. However, you can set PERCENTQUANTILES to a scalar to request another percentage point, or to a variate to request several. The QUANTILE parameter will then return a pointer with length equal to the required number of quantiles, instead of a single table.

The PRINT option allows you to print the tables (as well as, or instead of, saving them). By default nothing is printed.

Options: PRINT, CLASSIFICATION, MRESPONSE, MRFACTOR, COUNTS, MARGINS, WEIGHTS, PERCENTQUANTILES, SDS, SKEWNESS, KURTOSIS, SEMEANS..

Parameters: DATA, TOTALS, NOBSERVATIONS, MEANS, MINIMA, MAXIMA, VARIANCES, QUANTILES.

Method

MTABULATE uses TABULATE to form tables for each multiple response or combination of multiple responses, and then EQUATE to put them all into a single table.

Action with RESTRICT

MTABULATE takes account of any restrictions on the classification or multiple-response factors or the DATA or WEIGHT variates.