Multiplicative Measurement Error and the Simulation Extrapolation Method

Abstract

Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Abstract Whereas the literature on additive measurement error has known a considerable treatment, less work has been done for multiplicative noise. In this paper we concentrate on multiplicative measurement error in the covariates, which contrary to additive error not only modifies proportionally the original value, but also conserves the structural zeros. Terms of use: Documents inThis paper compares three variants to specify the multiplicative measurement error model in the simulation step of the Simulation-Extrapolation (SIMEX) method originally proposed by Cook and Stefanski (1994): i) as an additive one without using a logarithmic transformation, ii) as the well-known logarithmic transformation of the multiplicative error model, and iii) as an approach using the multiplicative measurement error model as such. The aim of the paper is to analyze how well these three approaches reduce the bias caused by the multiplicative measurement error. We apply three variants to the case of data masking by multiplicative measurement error, in order to obtain parameter estimates of the true data generating process. We produce Monte Carlo evidence on how the reduction of data quality can be minimized.JEL classification: C13, C21

Revisiting the Distance Duality Relation using a non-parametric regression method

“…Initially in the simulation step, some additional error is introduced by hand in the data with some controlling parameter ξ. Next by using regression analysis on this new dataset, we try to trace the effect of the measured error in our original dataset [45,46] Figure 3: Plot of CV vs α for all the datasets. For Dataset A, we choose α to be 0.9 , while for Dataset B and Simulated dataset minimum value of CV is at α = 0.8 and α = 0.6 respectively.…”

“…Initially in the simulation step, some additional error is introduced by hand in the data with some controlling parameter ξ. Next by using regression analysis on this new dataset, we try to trace the effect of the measured error in our original dataset [45,46].…”

“…Without assuming any prior or cosmological model, LOESS recovers the global trend of data by studying the appropriate number of neighbouring data points around the focus point. Further, to take measurement errors of data into account, we use SIMulation and EXtrapolation method (SIMEX) [43][44][45]. In Cosmology, this technique has been used in detail by Montiel et al for reconstruction of the cosmic expansion history [46].…”

“…Next by using regression analysis on this new dataset, we try to trace the effect of the measured error in our original dataset [45,46].…”

Estimation of linear models with anonymised panel data

“…This method was originally designed by Cook and Stefanski (1994) in order to handle additive noise. Biewen et al (2008) show that the SIMEX approach can be used in the multiplicative case as well and present three approaches to deal with multiplicative stochastic noise. 18 We test the SIMEX correction using a simulation study.…”