From Empirical Bayes to Full Bayes: Methods for Analysing Traffice Safety Data, October 25, 2004

Traffic safety engineers are among the early adopters of Bayesian statistical tools for analyzing crash data. As in many other areas of application, empirical Bayes methods were their first choice, perhaps because they represent an intuitively appealing, yet relatively easy to implement alternative to purely classical approaches. With the enormous progress in numerical methods made in recent years and with the availability of free, easy to use software that permits implementing a fully Bayesian approach, however, there is now ample justification to progress towards fully Bayesian analyses of crash data. The fully Bayesian approach, in particular as implemented via multi-level hierarchical models, has many advantages over the empirical Bayes approach. In a full Bayesian analysis, prior information and all available data are seamlessly integrated into posterior distributions on which practitioners can base their inferences. All uncertainties are thus accounted for in the analyses and there is no need to pre-process data to obtain Safety Performance Functions and other such prior estimates of the effect of covariates on the outcome of interest. In this slight, fully Bayesian methods may well be less costly to implement and may result in safety estimates with more realistic standard errors. In this manuscript, we present the full Bayesian approach to analyzing traffic safety data and focus on highlighting the differences between the empirical Bayes and the full Bayes approaches. We use an illustrative example to discuss a step-by-step Bayesian analysis of the data and to show some of the types of inferences that are possible within the full Bayesian framework.

From Empirical Bayes to Full Bayes: Methods for Analyzubg Traffic Safey Data, October 25, 2004

Traffic safety engineers are among the early adopters of Bayesian statistical tools for analyzing crash data. As in many other areas of application, empirical Bayes methods were their first choice, perhaps because they represent an intuitively appealing, yet relatively easy to implement alternative to purely classical approaches. With the enormous progress in numerical methods made in recent years and with the availability of free, easy to use software that permits implementing a fully Bayesian approach, however, there is now ample justification to progress towards fully Bayesian analyses of crash data. The fully Bayesian approach, in particular as implemented via multi-level hierarchical models, has many advantages over the empirical Bayes approach. In a full Bayesian analysis, prior information and all available data are seamlessly integrated into posterior distributions on which practitioners can base their inferences. All uncertainties are thus accounted for in the analyses and there is no need to pre-process data to obtain Safety Performance Functions and other such prior estimates of the effect of covariates on the outcome of interest. In this light, fully Bayesian methods may well be less costly to implement and may result in safety estimates with more realistic standard errors. In this manuscript, we present the full Bayesian approach to analyzing traffic safety data and focus on highlighting the differences between the empirical Bayes and the full Bayes approaches. We use an illustrative example to discuss a step-by-step Bayesian analysis of the data and to show some of the types of inferences that are possible within the full Bayesian framework.

Iowa Calibration of MEPDG Performance Prediction Models, 2013

This study aims to improve the accuracy of AASHTO Mechanistic-Empirical Pavement Design Guide (MEPDG) pavement performance predictions for Iowa pavement systems through local calibration of MEPDG prediction models. A total of 130 representative pavement sites across Iowa were selected. The selected pavement sites represent flexible, rigid, and composite pavement systems throughout Iowa. The required MEPDG inputs and the historical performance data for the selected sites were extracted from a variety of sources. The accuracy of the nationally-calibrated MEPDG prediction models for Iowa conditions was evaluated. The local calibration factors of MEPDG performance prediction models were identified to improve the accuracy of model predictions. The identified local calibration coefficients are presented with other significant findings and recommendations for use in MEPDG/DARWin-ME for Iowa pavement systems.

Iowa Calibration of MEPDG Performance Prediction Models: Tech Transfer Summary

This study aims to improve the accuracy of AASHTO Mechanistic-Empirical Pavement Design Guide (MEPDG) pavement performance predictions for Iowa pavement systems through local calibration of MEPDG prediction models. A total of 130 representative pavement sites across Iowa were selected. The selected pavement sites represent flexible, rigid, and composite pavement systems throughout Iowa. The required MEPDG inputs and the historical performance data for the selected sites were extracted from a variety of sources. The accuracy of the nationally-calibrated MEPDG prediction models for Iowa conditions was evaluated. The local calibration factors of MEPDG performance prediction models were identified to improve the accuracy of model predictions. The identified local calibration coefficients are presented with other significant findings and recommendations for use in MEPDG/DARWin-ME for Iowa pavement systems.

MEPDG Work Plan Task No. 8: Validation of Pavement Performance Curves for the Mechanistic-Empirical Pavement Design Guide, 2009

The objective of this research is to determine whether the nationally calibrated performance models used in the Mechanistic-Empirical Pavement Design Guide (MEPDG) provide a reasonable prediction of actual field performance, and if the desired accuracy or correspondence exists between predicted and monitored performance for Iowa conditions. A comprehensive literature review was conducted to identify the MEPDG input parameters and the MEPDG verification/calibration process. Sensitivities of MEPDG input parameters to predictions were studied using different versions of the MEPDG software. Based on literature review and sensitivity analysis, a detailed verification procedure was developed. A total of sixteen different types of pavement sections across Iowa, not used for national calibration in NCHRP 1-47A, were selected. A database of MEPDG inputs and the actual pavement performance measures for the selected pavement sites were prepared for verification. The accuracy of the MEPDG performance models for Iowa conditions was statistically evaluated. The verification testing showed promising results in terms of MEPDG’s performance prediction accuracy for Iowa conditions. Recalibrating the MEPDG performance models for Iowa conditions is recommended to improve the accuracy of predictions. ****************** Large File**************************

Investigation of AASHTOWare Pavement ME Design/DARWin-ME Performance Prediction Models for Iowa Pavement Analysis and Design, 2015

The Mechanistic-Empirical Pavement Design Guide (MEPDG) was developed under National Cooperative Highway Research Program (NCHRP) Project 1-37A as a novel mechanistic-empirical procedure for the analysis and design of pavements. The MEPDG was subsequently supported by AASHTO’s DARWin-ME and most recently marketed as AASHTOWare Pavement ME Design software as of February 2013. Although the core design process and computational engine have remained the same over the years, some enhancements to the pavement performance prediction models have been implemented along with other documented changes as the MEPDG transitioned to AASHTOWare Pavement ME Design software. Preliminary studies were carried out to determine possible differences between AASHTOWare Pavement ME Design, MEPDG (version 1.1), and DARWin-ME (version 1.1) performance predictions for new jointed plain concrete pavement (JPCP), new hot mix asphalt (HMA), and HMA over JPCP systems. Differences were indeed observed between the pavement performance predictions produced by these different software versions. Further investigation was needed to verify these differences and to evaluate whether identified local calibration factors from the latest MEPDG (version 1.1) were acceptable for use with the latest version (version 2.1.24) of AASHTOWare Pavement ME Design at the time this research was conducted. Therefore, the primary objective of this research was to examine AASHTOWare Pavement ME Design performance predictions using previously identified MEPDG calibration factors (through InTrans Project 11-401) and, if needed, refine the local calibration coefficients of AASHTOWare Pavement ME Design pavement performance predictions for Iowa pavement systems using linear and nonlinear optimization procedures. A total of 130 representative sections across Iowa consisting of JPCP, new HMA, and HMA over JPCP sections were used. The local calibration results of AASHTOWare Pavement ME Design are presented and compared with national and locally calibrated MEPDG models.

Empirical study of correlated survival times for recurrent events with proportional hazards margins and the effect of correlation and censoring.

Background: In longitudinal studies where subjects experience recurrent incidents over a period of time, such as respiratory infections, fever or diarrhea, statistical methods are required to take into account the within-subject correlation. Methods: For repeated events data with censored failure, the independent increment (AG), marginal (WLW) and conditional (PWP) models are three multiple failure models that generalize Cox"s proportional hazard model. In this paper, we revise the efficiency, accuracy and robustness of all three models under simulated scenarios with varying degrees of within-subject correlation, censoring levels, maximum number of possible recurrences and sample size. We also study the methods performance on a real dataset from a cohort study with bronchial obstruction. Results: We find substantial differences between methods and there is not an optimal method. AG and PWP seem to be preferable to WLW for low correlation levels but the situation reverts for high correlations. Conclusions: All methods are stable in front of censoring, worsen with increasing recurrence levels and share a bias problem which, among other consequences, makes asymptotic normal confidence intervals not fully reliable, although they are well developed theoretically.

An Empirical Study on Faculty Perceptions and Teaching Practices of Wikipedia

Some faculty members from different universities around the world have begun to use Wikipedia as a teaching tool in recent years. These experiences show, in most cases, very satisfactory results and a substantial improvement in various basic skills, as well as a positive influence on the students' motivation. Nevertheless and despite the growing importance of e-learning methodologies based on the use of the Internet for higher education, the use of Wikipedia as a teaching resource remains scarce among university faculty.Our investigation tries to identify which are the main factors that determine acceptance or resistance to that use. We approach the decision to use Wikipedia as a teaching tool by analyzing both the individual attributes of faculty members and the characteristics of the environment where they develop their teaching activity. From a specific survey sent to all faculty of the Universitat Oberta de Catalunya (UOC), pioneer and leader in online education in Spain, we have tried to infer the influence of these internal and external elements. The questionnaire was designed to measure different constructs: perceived quality of Wikipedia, teaching practices involving Wikipedia, use experience, perceived usefulness and use of 2.0 tools. Control items were also included for gathering information on gender, age, teaching experience, academic rank, and area of expertise.Our results reveal that academic rank, teaching experience, age or gender, are not decisive factors in explaining the educational use of Wikipedia. Instead, the decision to use it is closely linked to the perception of Wikipedia's quality, the use of other collaborative learning tools, an active attitude towards web 2.0 applications, and connections with the professional non-academic world. Situational context is also very important, since the use is higher when faculty members have got reference models in their close environment and when they perceive it is positively valued by their colleagues. As far as these attitudes, practices and cultural norms diverge in different scientific disciplines, we have also detected clear differences in the use of Wikipedia among areas of academic expertise. As a consequence, a greater application of Wikipedia both as a teaching resource and as a driver for teaching innovation would require much more active institutional policies and some changes in the dominant academic culture among faculty members.

Timing of Capacity Change: Models for Capital Intensive Industry

The suitable timing of capacity investments is a remarkable issue especially in capital intensive industries. Despite its importance, fairly few studies have been published on the topic. In the present study models for the timing of capacity change in capital intensive industry are developed. The study considers mainly the optimal timing of single capacity changes. The review of earlier research describes connections between cost, capacity and timing literature, and empirical examples are used to describe the starting point of the study and to test the developed models. The study includes four models, which describe the timing question from different perspectives. The first model, which minimizes unit costs, has been built for capacity expansion and replacement situations. It is shown that the optimal timing of an investment can be presented with the capacity and cost advantage ratios. After the unit cost minimization model the view is extended to the direction of profit maximization. The second model states that early investments are preferable if the change of fixed costs is small compared to the change of the contribution margin. The third model is a numerical discounted cash flow model, which emphasizes the roles of start-up time, capacity utilization rate and value of waiting as drivers of the profitable timing of a project. The last model expands the view from project level to company level and connects the flexibility of assets and cost structures to the timing problem. The main results of the research are the solutions of the models and analysis or simulations done with the models. The relevance and applicability of the results are verified by evaluating the logic of the models and by numerical cases.

Monte Carlo Methods in Parameter Estimation of Nonlinear Models

Yksi keskeisimmistä tehtävistä matemaattisten mallien tilastollisessa analyysissä on mallien tuntemattomien parametrien estimointi. Tässä diplomityössä ollaan kiinnostuneita tuntemattomien parametrien jakaumista ja niiden muodostamiseen sopivista numeerisista menetelmistä, etenkin tapauksissa, joissa malli on epälineaarinen parametrien suhteen. Erilaisten numeeristen menetelmien osalta pääpaino on Markovin ketju Monte Carlo -menetelmissä (MCMC). Nämä laskentaintensiiviset menetelmät ovat viime aikoina kasvattaneet suosiotaan lähinnä kasvaneen laskentatehon vuoksi. Sekä Markovin ketjujen että Monte Carlo -simuloinnin teoriaa on esitelty työssä siinä määrin, että menetelmien toimivuus saadaan perusteltua. Viime aikoina kehitetyistä menetelmistä tarkastellaan etenkin adaptiivisia MCMC menetelmiä. Työn lähestymistapa on käytännönläheinen ja erilaisia MCMC -menetelmien toteutukseen liittyviä asioita korostetaan. Työn empiirisessä osuudessa tarkastellaan viiden esimerkkimallin tuntemattomien parametrien jakaumaa käyttäen hyväksi teoriaosassa esitettyjä menetelmiä. Mallit kuvaavat kemiallisia reaktioita ja kuvataan tavallisina differentiaaliyhtälöryhminä. Mallit on kerätty kemisteiltä Lappeenrannan teknillisestä yliopistosta ja Åbo Akademista, Turusta.

Are R&D collaborative agreements persistent at the firm level? Empirical evidence for the Spanish case

We provide evidence on the dynamics in firms’ R&D cooperation behaviour. Our main objective is to analyse if R&D collaborative agreements are persistent at the firm level, and in such a case, to study what are the main drivers of this phenomenon. R&D cooperation activities at the firm level can be persistent due to true state dependence, this implying that cooperating in a given period enhances the probability of doing it in the subsequent period and it can also be a consequence of firms’ individual heterogeneity, so that certain firms have certain characteristics that make them more likely to carry out technological alliances.

Are R&D collaborative agreements persistent at the firm level? Empirical evidence for the Spanish case

We provide evidence on the dynamics in firms’ R&D cooperation behaviour. Our main objective is to analyse if R&D collaborative agreements are persistent at the firm level, and in such a case, to study what are the main drivers of this phenomenon. R&D cooperation activities at the firm level can be persistent due to true state dependence, this implying that cooperating in a given period enhances the probability of doing it in the subsequent period and it can also be a consequence of firms’ individual heterogeneity, so that certain firms have certain characteristics that make them more likely to carry out technological alliances.

Mobile E-commerce business model – A value web based approach to business models in mobile gaming industry

Tämän tutkimuksen tavoitteena oli tutkia langattomien internet palveluiden arvoverkkoa ja liiketoimintamalleja. Tutkimus oli luonteeltaan kvalitatiivinen ja siinä käytettiin strategiana konstruktiivista case-tutkimusta. Esimerkkipalveluna oli Treasure Hunters matkapuhelinpeli. Tutkimus muodostui teoreettisesta ja empiirisestä osasta. Teoriaosassa liitettiin innovaatio, liiketoimintamallit ja arvoverkko käsitteellisesti toisiinsa, sekä luotiin perusta liiketoimintamallien kehittämiselle. Empiirisessä osassa keskityttiin ensin liiketoimintamallien luomiseen kehitettyjen innovaatioiden pohjalta. Lopuksi pyrittiin määrittämään arvoverkko palvelun toteuttamiseksi. Tutkimusmenetelminä käytettiin innovaatiosessiota, haastatteluja ja lomakekyselyä. Tulosten pohjalta muodostettiin useita liiketoimintakonsepteja sekä kuvaus arvoverkon perusmallista langattomille peleille. Loppupäätelmänä todettiin että langattomat palvelut vaativat toteutuakseen useista toimijoista koostuvan arvoverkon.

A territorial approach to R&D subsidies: Empirical evidence for Catalonian firms

Using a database of 2,263 responses to R&D public calls in Catalonia, during the period 2007–2010, this paper proceeds to analyse the potential interaction of the territorial and policy dimensions with the propensity to apply for, and be awarded, a public R&D subsidy. Controlling for characteristics at the firm and project level, we estimate models using a twostep procedure. In the first step, our results suggest that large firms which export and which belong to high-tech manufactures are more likely to participate in a public R&D call. Furthermore, both urban location and past experience of such calls have a positive effect. Our territorial proxy of information spillovers shows a positive sign, but this is only significant at intra-industry level. Membership of one of the sectors prioritized by the Catalan government, perhaps surprisingly, does not have a significant impact. In the second step, our results show that cooperative projects, SMEs or old firms shows a positive effect on the probability of obtaining a public subsidy. Finally, the cluster policy does not show a clear relationship with the public R&D call, suggesting that cluster policies and R&D subsidies follow different goals. Our results are in line with previous results in the literature, but they highlight the unequal territorial distribution of the firms which apply and the fact that policymakers should interlink the decision criteria for their public call with other policies.

Completion and Publication Rates of Randomized Controlled Trials in Surgery: An Empirical Study.

OBJECTIVE: To investigate the prevalence of discontinuation and nonpublication of surgical versus medical randomized controlled trials (RCTs) and to explore risk factors for discontinuation and nonpublication of surgical RCTs. BACKGROUND: Trial discontinuation has significant scientific, ethical, and economic implications. To date, the prevalence of discontinuation of surgical RCTs is unknown. METHODS: All RCT protocols approved between 2000 and 2003 by 6 ethics committees in Canada, Germany, and Switzerland were screened. Baseline characteristics were collected and, if published, full reports retrieved. Risk factors for early discontinuation for slow recruitment and nonpublication were explored using multivariable logistic regression analyses. RESULTS: In total, 863 RCT protocols involving adult patients were identified, 127 in surgery (15%) and 736 in medicine (85%). Surgical trials were discontinued for any reason more often than medical trials [43% vs 27%, risk difference 16% (95% confidence interval [CI]: 5%-26%); P = 0.001] and more often discontinued for slow recruitment [18% vs 11%, risk difference 8% (95% CI: 0.1%-16%); P = 0.020]. The percentage of trials not published as full journal article was similar in surgical and medical trials (44% vs 40%, risk difference 4% (95% CI: -5% to 14%); P = 0.373). Discontinuation of surgical trials was a strong risk factor for nonpublication (odds ratio = 4.18, 95% CI: 1.45-12.06; P = 0.008). CONCLUSIONS: Discontinuation and nonpublication rates were substantial in surgical RCTs and trial discontinuation was strongly associated with nonpublication. These findings need to be taken into account when interpreting surgical literature. Surgical trialists should consider feasibility studies before embarking on full-scale trials.