Methods for determining the statistical part worth value of factors in conjoint analysis

MONANOVA is one type of conjoint analysis used for measuring the part worth value of factors to the total evaluation, exclusively using preference ranking data of a group of commercial products designed by presorted factors. Its criterion, called Stress, is the same as that of monotone regression in MDS. Consequently, part worth values obtained from MONANOVA do not necessarily lead to definite solutions but give an approximate comparison of each factorʹs contribution to the total evaluation of the products. s paper, we would like to discuss two problems with MONANOVA: namely, its reliability and stability. We would then like to propose two possible solutions to these problems: the first combines regression and monotone methods; the second employs quadratic fractional programming. With these, we hope to obtain each factorʹs contribution to the total evaluation as a partial correlation coefficient and to demonstrate that one can compare the factorʹs contribution to the total evaluation with constant stability.