`BLDATMAN`

BLDATMAN (Short form for Bland Altman analysis) is a command that is required to invoke the procedure.

The Bland Altman analysis is a graphical technique to evaluate the bias between mean differences. Bland Altman proposed by Altman and Bland, 1983, is a procedure based on the quantification of the agreement between two quantitative measurements by exploring the mean differences and constructing the limits of agreement.

The Bland Altman analysis is available in Statistics Base Edition.

BLDATMAN variable list1 [ WITH variable list2 [(PAIRED)] ]

variable_list1 is required and allows only numeric variables. variable_list1 may be followed by an optional keyword WITH. If WITH is not specified, variable_list1 must contain at least two different numeric variables to run the procedure. If WITH is not specified, each variable is compared with every other variable in variable_list1. If WITH is specified, it must be followed by variable_list2, which allows only numeric variables. If WITH is specified, every variable in variable_list1 is compared with every variable in variable_list2. If WITH is specified, an optional keyword PAIRED may be specified in parentheses and put at the end of the command. When PAIRED is specified, the first variable before WITH is compared with the first variable after WITH, the second variable before WITH is compared with the second variable after WITH, and so on. The same number of variables should be specified before and after WITH. Those unmatched variables are ignored and a warning message is issued. PAIRED generates an error message if keyword WITH is not specified. If WITH is specified, regardless PAIRED, at least one pair of two different variables must be matched to run the procedure.