Takane, Young, and de Leeuw proposed a procedure called FACTALS for th
e analysis of variables of mixed measurement levels (numerical, ordina
l, or nominal). Mooijaart pointed out that their algorithm does not ne
cessarily converge, and Nevels proposed a new algorithm for the case o
f nominal variables. In the present paper it is shown that Nevels' pro
cedure is incorrect, and a new procedure for handling nominal variable
s is proposed. In addition, a procedure for handling ordinal variables
is proposed. Using these results, a monotonically convergent algorith
m is constructed for FACTALS of any mixture of variables.