A problem often arising in engineering applications of computer models is t
o determine the importance of each data item in the large pool of required
input factors. This paper explores a statistical approach for investigating
factor sensitivities. The methodology is demonstrated with the HDM-III hig
hway life-cycle cost analysis model. Specifically, the net present value (N
PV) of life-cycle costs predicted by the HDM-III model is analyzed, and sen
sitivities of NPV to the link characterization input factors are investigat
ed. In the statistically designed experiment, combinations of the input fac
tors are chosen using a method called Latin hypercube sampling, which is we
ll suited to the deterministic HDM-III model. Two analyses of the output da
ta are performed, based on a first-order regression approximation and a Gau
ssian stochastic-process model. For NPV, the factor rankings are similar, b
ut the sensitivities obtained from the two techniques show some marked diff
erences. This demonstrates the greater flexibility of the stochastic-proces
s model in dealing with nonlinearities and factor interactions in complex i
nput-output relationships.