DATA Subcommand (BAYES ONESAMPLE: Binomial command)

DATA specifies recode data and set parameters for the estimation of Bayes factors. At least one /DATA subcommand is required to run the analysis. Multiple DATA subcommands are allowed.

Numerical variables

VARIABLES = sublist. Specify the variables to use for the procedure. The parameter is required. The available variables are a subset of the original variable list. Each variable in the list is analyzed.

SUCCESS. Specify how“success”, the data values tested against the test value, is defined for numerical variables. Valid when DISTRIBUTION = BINOMIAL.
  • LAST is the default setting. The parameter performs the binomial test using the last value found in the category after sorted in an ascending order to define “success”.
  • FIRST uses the first value found in the category after sorted in an ascending order.
  • MIDPOINT uses the values ≥ the midpoint as cases. A midpoint value is the average of the minimum and maximum sample data.
  • CUTPOINT(value) uses the values ≥ a specified cutoff values as cases. The value must be a single numeric.
  • LEVEL(values) treats the user-specified values (there can be more than 1) as cases.

NULLPRIORPARAM = value1 value 2. Specify the shape parameter a0 and the scale parameter b0 under the null hypothesis of Binomial inference. Valid when DISTRIBUTION= BINOMIAL. For Binomial-Beta relationship, the support range is value1, value2 > 0 or value1 = value2 = 0 to accommodate the Beta and the Haldane’s priors.

ALTERPRIORPARAM = value1 value 2. Specify a1 and b1 under the alternative hypothesis of Binomial inference if Bayes factor is to be estimated. The support range is the same as those in NULLPRIORPARAM. The parameter is valid when DISTRIBUTION = BINOMIAL. For Binomial-Beta relationship, the support range is value1, value2 > 0 or value1 = value2 = 0 to accommodate the Beta and the Haldane’s priors.

String variables

VARIABLES = sublist. Specify the variables to use for the procedure. The parameter is required. The available variables are a subset of the original variable list. Each variable in the list is analyzed.

SUCCESS. Specify how “success”, the data values tested against the test value, is defined for string variables. Valid when DISTRIBUTION = BINOMIAL.
  • LAST is the default setting. The parameter performs the binomial test using the last string found in the category after sorted in an ascending order to define “success”.
  • FIRST uses the first string found in the category after sorted in an ascending order.
  • LEVEL(values) treats the user-specified strings (there can be more than 1) as cases.

NULLPRIORPARAM = value1 value 2. Specify the shape parameter a0 and the scale parameter b0 under the null hypothesis of Binomial inference. Valid when DISTRIBUTION= BINOMIAL. For Binomial-Beta relationship, the support range is value1, value2 > 0 or value1 = value2 = 0 to accommodate the Beta and the Haldane’s priors.

ALTERPRIORPARAM = value1 value 2. Specify a1 and b1 under the alternative hypothesis of Binomial inference if Bayes factor is to be estimated. The support range is the same as those in NULLPRIORPARAM. The parameter is valid when DISTRIBUTION = BINOMIAL. For Binomial-Beta relationship, the support range is value1, value2 > 0 or value1 = value2 = 0 to accommodate the Beta and the Haldane’s priors.