Conditional Maximum Likelihood Estimation in Probability-Branched Multistage Designs

This article describes the conditional maximum likelihood-based item parameter estimation in probabilistic multistage designs. In probabilistic multistage designs, the routing is not solely based on a raw score j and a cut score c as well as a rule for routing into a module such as j < c or j ≤ c but is based on a probability p(j) for each raw score j. It can be shown that the use of a conventional conditional maximum likelihood parameter estimate in multistage designs leads to severely biased item parameter estimates. Zwitser and Maris (2013) were able to show that with deterministic routing, the integration of the design into the item parameter estimation leads to unbiased estimates. This article extends this approach to probabilistic routing and, at the same time, represents a generalization. In a simulation study, it is shown that the item parameter estimation in probabilistic designs leads to unbiased item parameter estimates.

Conditional Maximum Likelihood Estimation in Probability-Branched Multistage Designs

This article describes the conditional maximum likelihood-based item parameter estimation in probabilistic multistage designs. In probabilistic multistage designs, the routing is not solely based on a raw score j and a cut score c as well as a rule for routing into a module such as j < c or j ≤ c but is based on a probability p(j) for each raw score j. It can be shown that the use of a conventional conditional maximum likelihood parameter estimate in multistage designs leads to severely biased item parameter estimates. Zwitser and Maris (2013) were able to show that with deterministic routing, the integration of the design into the item parameter estimation leads to unbiased estimates. This article extends this approach to probabilistic routing and, at the same time, represents a generalization. In a simulation study, it is shown that the item parameter estimation in probabilistic designs leads to unbiased item parameter estimates.

Land Rights and Improvement Investments Among Crop Farmers in Southwest, Nigeria

This study examined the effects of land rights on investment decisions of farmers and determinants of land rights as well as the impact of land rights on farm productivity in Southwest Nigeria. Multistage sampling process was used in the selection of 320 respondents for the study. Two-stage conditional maximum likelihood (2SCML) approach of multivariate probit regression model and Endogenous switching regression (ESR) model were used for the analysis of the collected data. The results of Average Treatment Effect on the Treated (ATT) estimates from ESR revealed that land rights increase farm productivity. This study further confirmed the importance of land rights in facilitating investment in soil-improving measures which subsequently leads to higher productivity. One of the policy implications emanating from this study suggests that having use and transfer rights as against use only rights goes a long way in enhancing investment in soil-improving or conservation measures. Keywords: Crop, Investment, Land, Multivariate Probit Regression Model, Rights, Two-stage Conditional Maximum Likelihood.

Evaluation of direct and indirect methods for modelling the joint distribution of tree diameter and height data with the bivariate Johnson’s SBB function to forest stands

Aim of study: In this study, both the direct and indirect methods by conditional maximum likelihood (CML) and moments for fitting Johnson’s SBB were evaluated. To date, Johnson’s SBB has been fitted by either indirect (two-stage) method using well-known procedures for the marginal diameter and heights, or direct methods, where all parameters are estimated at once. Application of bivariate Johnson’s SBB for predicting height and improving volume estimation requires a suitable fitting method.Area of study: E. globulus, P. pinaster and P. radiata stands in northwest Spain.Material and methods: The data set comprised of 308, 184 and 96 permanent sample plots (PSPs) from the aforementioned species. The suitability of the method was evaluated based on height and volume prediction. Indices including coefficient of determination (R2), root mean square Error (RMSE), model efficiency (MEF), Bayesian Information Criterion (BIC) and Hannan-Quinn Criterion (HQC) were used to assess the model predictions. Significant difference between observed and predicted tree height and volumes were tested using paired sample t-test at 5% level for each plot by species.Main results: The indirect method by CML was the most suitable method for height and volume prediction in the three species. The R2 and RMSE for height prediction ranged from 0.994 – 0.820 and 1.454 – 1.676, respectively. The percentage of plot in which the observed and predicted heights were significant was 0.32%. The direct method was the least performed method especially for height prediction in E. globulus.Research highlights: The indirect (two-stage) method, especially by conditional maximum likelihood, was the most suitable method for the bivariate Johnson’s SBB distribution.Keywords: conditional maximum likelihood; moments; two-stage method; direct method; tree volume.