Surrogate science: The idol of a universal method for scientific inference

The application of statistics to science is not a neutral act. Statistical tools have shaped and were also shaped by its objects. In the social sciences, statistical methods fundamentally changed research practice, making statistical inference its centerpiece. At the same time, textbook writers in the social sciences have transformed rivaling statistical systems into an apparently monolithic method that could be used mechanically. The idol of a universal method for scientific inference has been worshipped since the "inference revolution" of the 1950s. Because no such method has ever been found, surrogates have been created, most notably the quest for significant p values. This form of surrogate science fosters delusions and borderline cheating and has done much harm, creating, for one, a flood of irreproducible results. Proponents of the "Bayesian revolution" should be wary of chasing yet another chimera: an apparently universal inference procedure. A better path would be to promote both an understanding of the various devices in the "statistical toolbox" and informed judgment to select among these.

Isotope ratio mass spectrometry as a tool for source inference in forensic science: a critical review

Isotope ratio mass spectrometry (IRMS) has been used in numerous fields of forensic science in a source inference perspective. This review compiles the studies published on the application of isotope ratio mass spectrometry (IRMS) to the traditional fields of forensic science so far. It completes the review of Benson et al. [1] and synthesises the extent of knowledge already gathered in the following fields: illicit drugs, flammable liquids, human provenancing, microtraces, explosives and other specific materials (packaging tapes, safety matches, plastics, etc.). For each field, a discussion assesses the state of science and highlights the relevance of the information in a forensic context. Through the different discussions which mark out the review, the potential and limitations of IRMS, as well as the needs and challenges of future studies are emphasized. The paper elicits the various dimensions of the source which can be obtained from the isotope information and demonstrates the transversal nature of IRMS as a tool for source inference.

Optimizing the Testing Process in Module Testing – Functional Testing Interface

Today's business environment has become increasingly unexpected and fast changing because of the global competition. This new environment requires the companies to organize their control differently, e.g. by logistic process thinking. Logistic process thinking in software engineering applies the principles of production process to immaterial products. Processes must be optimized, so that every phase adds value to the customer, and the lead times can be cut shorter to meet the new customer requirements. The purpose of this thesis is to examine and optimize the testing processes of software engineering concentrating on module testing, functional testing and their interface. The concept of logistic process thinking is introduced through production process, value added model and process management. Also theory of testing based on literature is presented, concentrating on module testing and functional testing. The testing processes of the Case Company are presented together with the project models in which they are implemented. The real life practices in module testing and functional testing and their interface are examined through interviews. These practices are analyzed against the processes and the testing theory, through which ideas for optimizing the testing process are introduced. The project world of the Case Company is also introduced together with two example testing projects in different life cycle phases. The examples give a view of how much effort of the project is put in different types of testing.

Testing a new multigroup inference approach to reconstructing past environmental conditions

A new, quantitative, inference model for environmental reconstruction (transfer function), based for the first time on the simultaneous analysis of multigroup species, has been developed. Quantitative reconstructions based on palaeoecological transfer functions provide a powerful tool for addressing questions of environmental change in a wide range of environments, from oceans to mountain lakes, and over a range of timescales, from decades to millions of years. Much progress has been made in the development of inferences based on multiple proxies but usually these have been considered separately, and the different numeric reconstructions compared and reconciled post-hoc. This paper presents a new method to combine information from multiple biological groups at the reconstruction stage. The aim of the multigroup work was to test the potential of the new approach to making improved inferences of past environmental change by improving upon current reconstruction methodologies. The taxonomic groups analysed include diatoms, chironomids and chrysophyte cysts. We test the new methodology using two cold-environment training-sets, namely mountain lakes from the Pyrenees and the Alps. The use of multiple groups, as opposed to single groupings, was only found to increase the reconstruction skill slightly, as measured by the root mean square error of prediction (leave-one-out cross-validation), in the case of alkalinity, dissolved inorganic carbon and altitude (a surrogate for air-temperature), but not for pH or dissolved CO2. Reasons why the improvement was less than might have been anticipated are discussed. These can include the different life-forms, environmental responses and reaction times of the groups under study.

Inference of Evolutionary Forces Acting on Human Biological Pathways.

Because natural selection is likely to act on multiple genes underlying a given phenotypic trait, we study here the potential effect of ongoing and past selection on the genetic diversity of human biological pathways. We first show that genes included in gene sets are generally under stronger selective constraints than other genes and that their evolutionary response is correlated. We then introduce a new procedure to detect selection at the pathway level based on a decomposition of the classical McDonald-Kreitman test extended to multiple genes. This new test, called 2DNS, detects outlier gene sets and takes into account past demographic effects and evolutionary constraints specific to gene sets. Selective forces acting on gene sets can be easily identified by a mere visual inspection of the position of the gene sets relative to their two-dimensional null distribution. We thus find several outlier gene sets that show signals of positive, balancing, or purifying selection but also others showing an ancient relaxation of selective constraints. The principle of the 2DNS test can also be applied to other genomic contrasts. For instance, the comparison of patterns of polymorphisms private to African and non-African populations reveals that most pathways show a higher proportion of nonsynonymous mutations in non-Africans than in Africans, potentially due to different demographic histories and selective pressures.

Hire me: Computational inference of hirability in employment interviews based on nonverbal behavior

Understanding the basis on which recruiters form hirability impressions for a job applicant is a key issue in organizational psychology and can be addressed as a social computing problem. We approach the problem from a face-to-face, nonverbal perspective where behavioral feature extraction and inference are automated. This paper presents a computational framework for the automatic prediction of hirability. To this end, we collected an audio-visual dataset of real job interviews where candidates were applying for a marketing job. We automatically extracted audio and visual behavioral cues related to both the applicant and the interviewer. We then evaluated several regression methods for the prediction of hirability scores and showed the feasibility of conducting such a task, with ridge regression explaining 36.2% of the variance. Feature groups were analyzed, and two main groups of behavioral cues were predictive of hirability: applicant audio features and interviewer visual cues, showing the predictive validity of cues related not only to the applicant, but also to the interviewer. As a last step, we analyzed the predictive validity of psychometric questionnaires often used in the personnel selection process, and found that these questionnaires were unable to predict hirability, suggesting that hirability impressions were formed based on the interaction during the interview rather than on questionnaire data.

The caspase-3-p120-RasGAP module generates a NF-κB repressor in response to cellular stress.

The nuclear factor κB (NF-κB) transcription factor is a master regulator of inflammation. Short-term NF-κB activation is generally beneficial. However, sustained NF-κB might be detrimental, directly causing apoptosis of cells or leading to a persistent damaging inflammatory response. NF-κB activity in stressed cells needs therefore to be controlled for homeostasis maintenance. In mildly stressed cells, caspase-3 cleaves p120 RasGAP, also known as RASA1, into an N-terminal fragment, which we call fragment N. We show here that this fragment is a potent NF-κB inhibitor. Fragment N decreases the transcriptional activity of NF-κB by promoting its export from the nucleus. Cells unable to generate fragment N displayed increased NF-κB activation upon stress. Knock-in mice expressing an uncleavable p120 RasGAP mutant showed exaggerated NF-κB activation when their epidermis was treated with anthralin, a drug used for the treatment of psoriasis. Our study provides biochemical and genetic evidence of the importance of the caspase-3-p120-RasGAP stress-sensing module in the control of stress-induced NF-κB activation.

The activity of the anti-apoptotic fragment generated by the caspase-3/p120 RasGAP stress-sensing module displays strict Akt isoform specificity.

The caspase-3/p120 RasGAP module acts as a stress sensor that promotes pro-survival or pro-death signaling depending on the intensity and the duration of the stressful stimuli. Partial cleavage of p120 RasGAP generates a fragment, called fragment N, which protects stressed cells by activating Akt signaling. Akt family members regulate many cellular processes including proliferation, inhibition of apoptosis and metabolism. These cellular processes are regulated by three distinct Akt isoforms: Akt1, Akt2 and Akt3. However, which of these isoforms are required for fragment N mediated protection have not been defined. In this study, we investigated the individual contribution of each isoform in fragment N-mediated cell protection against Fas ligand induced cell death. To this end, DLD1 and HCT116 isogenic cell lines lacking specific Akt isoforms were used. It was found that fragment N could activate Akt1 and Akt2 but that only the former could mediate the protective activity of the RasGAP-derived fragment. Even overexpression of Akt2 or Akt3 could not rescue the inability of fragment N to protect cells lacking Akt1. These results demonstrate a strict Akt isoform requirement for the anti-apoptotic activity of fragment N.

PeerHood as UMSIC middleware module

The UMSIC project will produce an application for social inclusion of children. This application will run on Nokia N810 internet tablet. PeerHood is an implementation of Peer-to-Peer neighborhood and communication concept, it is going to be used as a part of middleware in the project. PeerHood is responsible for providing neighboring information and connections to nearby devices. This thesis will present the requirements that the project sets to PeerHood including the general state of the art middleware requirements and the requirements set by the target device. These requirements will not be analyzed further. One main focus of this thesis is to analyze PeerHood from the UMSIC project point of view. In this thesis the results of PeerHood analysis are presented, including memory consumption testing, performance testing and testing of networking components of PeerHood. As a result of these tests modifications for PeerHood are introduced. The modifications are related to PeerHood usage in real mobile environment. Based on these requirements a framework was implemented that, when used properly, will enhance the context awareness of PeerHood, especially in mobile devices. The techniques used in framework are presented and instructions about how to use the framework are given. The approaches used in the implemented framework are analysed in this thesis. As an outcome of this thesis the context awareness of PeerHood is improved. As an additional outcome the guidelines for future development of PeerHood are introduced that are based on the results of the analysis of both PeerHood and the implemented framework.

Indirect likelihood inference (revised)

Standard indirect Inference (II) estimators take a given finite-dimensional statistic, Z_{n} , and then estimate the parameters by matching the sample statistic with the model-implied population moment. We here propose a novel estimation method that utilizes all available information contained in the distribution of Z_{n} , not just its first moment. This is done by computing the likelihood of Z_{n}, and then estimating the parameters by either maximizing the likelihood or computing the posterior mean for a given prior of the parameters. These are referred to as the maximum indirect likelihood (MIL) and Bayesian Indirect Likelihood (BIL) estimators, respectively. We show that the IL estimators are first-order equivalent to the corresponding moment-based II estimator that employs the optimal weighting matrix. However, due to higher-order features of Z_{n} , the IL estimators are higher order efficient relative to the standard II estimator. The likelihood of Z_{n} will in general be unknown and so simulated versions of IL estimators are developed. Monte Carlo results for a structural auction model and a DSGE model show that the proposed estimators indeed have attractive finite sample properties.

Usability Centered Implementation of a Web Based Value Added Module

Customer specific functionalities are a challenging part of procurement and invoice automation environments. In Basware Enterprise Purchase to Payment product family the customer specific reports are supported only in a basic level without any seamless interface between all EPP products. Also other customer specific functionalities are not supported as there is no customizable interface between the applications and only the most common features are implemented to the products themselves. In this thesis foundations are created for a new web based value added module where it is possible to create seamless customer specific functionalities throughout the whole EPP product family. The work is implemented in a Proof of Concept type of piloting. The system is created in user centered way where the users are able to explain their requests and determine their needs. The result is an excellent foundation for a module that can be developed further.