Forum: help
Monitor Forum | Start New ThreadRE: Monotonicity and curvature check [ Reply ] By: Arne Henningsen on 2013-03-19 20:54 | [forum:39226] |
Dear Tshepelayi Your questions is not related to the topic "Monotonicity and curvature check." Please re-post it in a thread with an appropriate title. Best regards, Arne |
RE: Monotonicity and curvature check [ Reply ] By: Tshepelayi Kabata on 2013-03-19 20:03 | [forum:39225] |
Dear Dr Henningsen, Thank you for shedding a light on my 2 previous issues. I have another concern. For some unidentified reasons, I am getting z-value and Pr(>|z|) throughout the coef(summary(sfastep1Result)) whereas I was expecting to get the t-value for the estimates. Is there any unobserved default setting for z-value and t-value ? |
RE: Monotonicity and curvature check [ Reply ] By: Arne Henningsen on 2013-03-01 23:52 | [forum:38995] |
Dear Tshepelayi The commands in Appendix 3 of my paper still give the same results as in the paper. You get different results, because you have a small typo in your commands: you wrote "LABOR" instead of "lLABOR" (the logarithm of the labor input). Please note that monotonicity does not require that all b_i are positive but that the right-hand side of equation (16) is positive. The right-hand side of equation (16) can be positive for some sets of input quantities even though a b_i (e.g. b_LABOR) is negative. Best regards, Arne |
Monotonicity and curvature check [ Reply ] By: Tshepelayi Kabata on 2013-02-17 20:40 | [forum:38868] |
Dear Dr Henningsen, Using R 2.12.0 to replicate your the result in your JPA 2009 paper , I find equivalent estimates but the result of the monotonicity and quasiconcavity differ ( see step 1 and 2 below ). What might be the issue ? If monotoncity is 100% fulfilled how do you explain the persistent negative sign of lAREA in the second step(page 223) ? Also how can you recover estimates of lAREA, lLABOR lNPK from the final stochastic frontier estimation (table 3, page 223)? First Step > summary(translogCheckMono(xNames=c("lAREA", "LABOR", "lNPK"), data= riceProdPhil, coef=uCoef, dataLogged=TRUE)) This translog function is monotonically increasing in lAREA, LABOR, lNPK at 340 out of 344 observations (98.8%) The monotonicity condition for the exogenous variable - 'lAREA' is fulfilled at 344 out of 344 observations (100%) - 'LABOR' is fulfilled at 342 out of 344 observations (99.4%) - 'lNPK' is fulfilled at 340 out of 344 observations (98.8%) > translogCheckCurvature(xNames=c("lAREA", "LABOR", "lNPK"), data= riceProdPhil, coef=uCoef, dataLogged=TRUE,convexity=FALSE, quasi=TRUE) This translog function is quasiconcave at 343 out of 344 observations (99.7%) Second Step > summary(translogCheckMono(xNames=c("lAREA", "LABOR", "lNPK"), data= riceProdPhil, coef=cCoef, dataLogged=TRUE)) This translog function is monotonically increasing in lAREA, LABOR, lNPK at 323 out of 344 observations (93.9%) The monotonicity condition for the exogenous variable - 'lAREA' is fulfilled at 344 out of 344 observations (100%) - 'LABOR' is fulfilled at 331 out of 344 observations (96.2%) - 'lNPK' is fulfilled at 323 out of 344 observations (93.9%) > translogCheckCurvature(xNames=c("lAREA", "LABOR", "lNPK"), data= riceProdPhil, coef=cCoef, dataLogged=TRUE,convexity=FALSE, quasi=TRUE) This translog function is quasiconcave at 343 out of 344 observations (99.7%) |