| In DEA frontier estimation, stability and outliers became an important issue. Referring to this issue, P. W. Bauer (Journal of Econometrics, vol. 46, no. 1/2, 1990) stated," ... the calculated frontier may be warped if the data are contaminated by statistical noise." In an attempt to encounter this problem, DEA sensitivity analysis studies were conducted in two different directions. Early on, Charnes et al. (1985) developed the envelopment space approach. More recently, Thompson et al. (1990) studied the sensitivity of efficiency measures by application of a multiplier space approach.;An initial application was made of the multiplier space approach by Thompson et al. to Kansas farming and Illinois Coal Mining. This dissertation focuses on extending the application of this approach to four more primary data sets, namely, 11 small business development centers (University of Houston), 30 independent oil and gas extraction firms, 14 major oil and gas extraction firms and the nation's 100 large commercial banks. The results are compared with those found by applying a theorem developed by Thrall, which facilitates comparisons of the two approaches. |
