Bernadette Govaerts

Encompassing in stationary linear dynamic models

A model M1 encompasses a rival model M2 if M1 can explain M2s results. A Wald Encompassing Test (WET) checks if a statistic of interest to M2 coincides with an estimator of its predicted value under M1. We propose techniques for evaluating WETs in stationary, linear, dynamic, single equations with weakly exogenous regressors, extending results for strong exogeneity. Dynamics can constrain M1s predictions of M2s findings, so encompassing tests can differ from existing tests as examples illustrate. Their asymptotic power functions are compared with the outcomes in a small Monte Carlo. The results support the use of parsimonious encompassing tests.

Occupational exposure to noise and the prevalence of hearing loss in a Belgian military population : a cross-sectional study

The armed forces are highly exposed to occupational noise. The aim of this study was to evaluate the prevalence and noise exposures associated with the severity of hearing loss (HL) in a Belgian military population. A cross-sectional study was carried out at the Centre for Medical Expertise (CME) and in four Units of Occupational Medicine (UOM). Hearing thresholds were determined by audiometry. The examination included a questionnaire on hearing-related medical history, and noise exposure in military and leisure time activity. A multinomial logistic regression model was used to assess the association of the severity of HL with tinnitus, with the military occupation, and with noise exposures. Of the 2055 subjects aged 18-55 years, 661 (32.2%) had a slight HL (25-40 dB), 280 (13.6%) had a moderate HL (45-60 dB) and 206 (10.0%) had a severe HL (> 60 dB) of 4 and 6 kHz for both ears. The prevalence of slight, moderate and severe HL increased significantly with age and was higher for subjects from Paracommando and infantry units. Fighting in Built-Up Area (FIBUA) training, shooting with large caliber weapons, and participation in military exercises were the best determinants of HL in this population. These results suggest that subjects from infantry and Paracommando units run the highest risk of HL because they are exposed to very loud noises in their professional life, like large caliber shooting and FIBUA training.

Control charts and efficient sampling methodologies in the field of photovoltaics

Industries must have their process under control in order to become more performant. This is true in particular for the photovoltaic (PV) industry. A Statistical Process Control (SPC) study has been realised for the semi-industrial process at IMEC. In these circumstances control charts have been built and sampling schemes have been investigated. It has been found to be powerful to detect special causes of variation and will still be used in the future for early detection of out of control situations.

A statistical hypothesis testing approach to the determination of supportable aircraft separation standards

This paper deals with the setting of a statistical hypothesis testing approach to the determination of aircraft separation standards that can be supported within a given flight information region. Given a minimal safety distance to be maintained between two aircraft, we derive a criterion that must be obeyed by their radar plots in order to meet the required separation at a known level of confidence. Conversely, given that a minimal safety distance is maintained between two radar plots, one can find the actual separation that can be guaranteed between both aircraft. The results described here can help to implement automated conflict alert and minimum safe altitude warning facilities.<<ETX>>

A note on a method to compute the asymptotic distribution of the sample second order moments of dynamic linear normal variables

This paper describes a method for the computation of the asymptotic distribution of any statistic or estimator which is a continuous function of sample second order moments of random variables having a dynamic linear normal distribution of the form : Zt = IIt-1 + et where et iN(0, σ) The method uses a general theorem due to Hannan [1970] and simplifies it in decomposing II in Jordan canonical form.