Evaluation use within a humanitarian non-governmental organization’s health care user-fee exemption program in West Africa

L’évaluation de l’action humanitaire (ÉAH) est un outil valorisé pour soutenir l’imputabilité, la transparence et l’efficience de programmes humanitaires contribuant à diminuer les inéquités et à promouvoir la santé mondiale. L’EAH est incontournable pour les parties prenantes de programme, les bailleurs de fonds, décideurs et intervenants souhaitant intégrer les données probantes aux pratiques et à la prise de décisions. Cependant, l’utilisation de l’évaluation (UÉ) reste incertaine, l’ÉAH étant fréquemment menée, mais inutilisé. Aussi, les conditions influençant l’UÉ varient selon les contextes et leur présence et applicabilité au sein d’organisations non-gouvernementales (ONG) humanitaires restent peu documentées. Les évaluateurs, parties prenantes et décideurs en contexte humanitaire souhaitant assurer l’UÉ pérenne détiennent peu de repères puisque rares sont les études examinant l’UÉ et ses conditions à long terme. La présente thèse tend à clarifier ces enjeux en documentant sur une période de deux ans l’UÉ et les conditions qui la détermine, au sein d’une stratégie d’évaluation intégrée au programme d’exemption de paiement des soins de santé d’une ONG humanitaire. L’objectif de ce programme est de faciliter l’accès à la santé aux mères, aux enfants de moins de cinq ans et aux indigents de districts sanitaires au Niger et au Burkina Faso, régions du Sahel où des crises alimentaires et économiques ont engendré des taux élevés de malnutrition, de morbidité et de mortalité. Une première évaluation du programme d’exemption au Niger a mené au développement de la stratégie d’évaluation intégrée à ce même programme au Burkina Faso. La thèse se compose de trois articles. Le premier présente une étude d’évaluabilité, étape préliminaire à la thèse et permettant de juger de sa faisabilité. Les résultats démontrent une logique cohérente et plausible de la stratégie d’évaluation, l’accessibilité de données et l’utilité d’étudier l’UÉ par l’ONG. Le second article documente l’UÉ des parties prenantes de la stratégie et comment celle-ci servit le programme d’exemption. L’utilisation des résultats fut instrumentale, conceptuelle et persuasive, alors que l’utilisation des processus ne fut qu’instrumentale et conceptuelle. Le troisième article documente les conditions qui, selon les parties prenantes, ont progressivement influencé l’UÉ. L’attitude des utilisateurs, les relations et communications interpersonnelles et l’habileté des évaluateurs à mener et à partager les connaissances adaptées aux besoins des utilisateurs furent les conditions clés liées à l’UÉ. La thèse contribue à l’avancement des connaissances sur l’UÉ en milieu humanitaire et apporte des recommandations aux parties prenantes de l’ONG.

Personalized recommendation on multi-layer context graph

Recommender systems have been successfully dealing with the problem of information overload. A considerable amount of research has been conducted on recommender systems, but most existing approaches only focus on user and item dimensions and neglect any additional contextual information, such as time and location. In this paper, we propose a Multi-Layer Context Graph (MLCG) model which incorporates a variety of contextual information into a recommendation process and models the interactions between users and items for better recommendation. Moreover, we provide a new ranking algorithm based on Personalized PageRank for recommendation in MLCG, which captures users' preferences and current situations. The experiments on two real-world datasets demonstrate the effectiveness of our approach.

A Comparison of the Effects of Classroom and Multi-User Virtual Environments on the Perceived Speaking Anxiety of Adult Post-Secondary English Language Learners

The population of English Language Learners (ELLs) globally has been increasing substantially every year. In the United States alone, adult ELLs are the fastest growing portion of learners in adult education programs (Yang, 2005). There is a significant need to improve the teaching of English to ELLs in the United States and other English-speaking dominant countries. However, for many ELLs, speaking, especially to Native English Speakers (NESs), causes considerable language anxiety, which in turn plays a vital role in hindering their language development and academic progress (Pichette, 2009; Woodrow, 2006). Task-based Language Teaching (TBLT), such as simulation activities, has long been viewed as an effective approach for second-language development. The current advances in technology and rapid emergence of Multi-User Virtual Environments (MUVEs) have provided an opportunity for educators to consider conducting simulations online for ELLs to practice speaking English to NESs. Yet to date, empirical research on the effects of MUVEs on ELLs’ language development and speaking is limited (Garcia-Ruiz, Edwards, & Aquino-Santos, 2007). This study used a true experimental treatment control group repeated measures design to compare the perceived speaking anxiety levels (as measured by an anxiety scale administered per simulation activity) of 11 ELLs (5 in the control group, 6 in the experimental group) when speaking to Native English Speakers (NESs) during 10 simulation activities. Simulations in the control group were done face-to-face, while those in the experimental group were done in the MUVE of Second Life. The results of the repeated measures ANOVA revealed after the Huynh-Feldt epsilon correction, demonstrated for both groups a significant decrease in anxiety levels over time from the first simulation to the tenth and final simulation. When comparing the two groups, the results revealed a statistically significant difference, with the experimental group demonstrating a greater anxiety reduction. These results suggests that language instructors should consider including face-to-face and MUVE simulations with ELLs paired with NESs as part of their language instruction. Future investigations should investigate the use of other multi-user virtual environments and/or measure other dimensions of the ELL/NES interactions.

Off-line and On-line Affective Recognition of a Computer User through A Biosignal Processing Approach

Physiological signals, which are controlled by the autonomic nervous system (ANS), could be used to detect the affective state of computer users and therefore find applications in medicine and engineering. The Pupil Diameter (PD) seems to provide a strong indication of the affective state, as found by previous research, but it has not been investigated fully yet. In this study, new approaches based on monitoring and processing the PD signal for off-line and on-line affective assessment (“relaxation” vs. “stress”) are proposed. Wavelet denoising and Kalman filtering methods are first used to remove abrupt changes in the raw Pupil Diameter (PD) signal. Then three features (PDmean, PDmax and PDWalsh) are extracted from the preprocessed PD signal for the affective state classification. In order to select more relevant and reliable physiological data for further analysis, two types of data selection methods are applied, which are based on the paired t-test and subject self-evaluation, respectively. In addition, five different kinds of the classifiers are implemented on the selected data, which achieve average accuracies up to 86.43% and 87.20%, respectively. Finally, the receiver operating characteristic (ROC) curve is utilized to investigate the discriminating potential of each individual feature by evaluation of the area under the ROC curve, which reaches values above 0.90. For the on-line affective assessment, a hard threshold is implemented first in order to remove the eye blinks from the PD signal and then a moving average window is utilized to obtain the representative value PDr for every one-second time interval of PD. There are three main steps for the on-line affective assessment algorithm, which are preparation, feature-based decision voting and affective determination. The final results show that the accuracies are 72.30% and 73.55% for the data subsets, which were respectively chosen using two types of data selection methods (paired t-test and subject self-evaluation). In order to further analyze the efficiency of affective recognition through the PD signal, the Galvanic Skin Response (GSR) was also monitored and processed. The highest affective assessment classification rate obtained from GSR processing is only 63.57% (based on the off-line processing algorithm). The overall results confirm that the PD signal should be considered as one of the most powerful physiological signals to involve in future automated real-time affective recognition systems, especially for detecting the “relaxation” vs. “stress” states.

User interaction monitoring and analysis framework

An overview is given of a user interaction monitoring and analysis framework called BaranC. Monitoring and analysing human-digital interaction is an essential part of developing a user model as the basis for investigating user experience. The primary human-digital interaction, such as on a laptop or smartphone, is best understood and modelled in the wider context of the user and their environment. The BaranC framework provides monitoring and analysis capabilities that not only records all user interaction with a digital device (e.g. smartphone), but also collects all available context data (such as from sensors in the digital device itself, a fitness band or a smart appliances). The data collected by BaranC is recorded as a User Digital Imprint (UDI) which is, in effect, the user model and provides the basis for data analysis. BaranC provides functionality that is useful for user experience studies, user interface design evaluation, and providing user assistance services. An important concern for personal data is privacy, and the framework gives the user full control over the monitoring, storing and sharing of their data.

A service-oriented user interaction analysis framework supporting adaptive applications

A comprehensive user model, built by monitoring a user's current use of applications, can be an excellent starting point for building adaptive user-centred applications. The BaranC framework monitors all user interaction with a digital device (e.g. smartphone), and also collects all available context data (such as from sensors in the digital device itself, in a smart watch, or in smart appliances) in order to build a full model of user application behaviour. The model built from the collected data, called the UDI (User Digital Imprint), is further augmented by analysis services, for example, a service to produce activity profiles from smartphone sensor data. The enhanced UDI model can then be the basis for building an appropriate adaptive application that is user-centred as it is based on an individual user model. As BaranC supports continuous user monitoring, an application can be dynamically adaptive in real-time to the current context (e.g. time, location or activity). Furthermore, since BaranC is continuously augmenting the user model with more monitored data, over time the user model changes, and the adaptive application can adapt gradually over time to changing user behaviour patterns. BaranC has been implemented as a service-oriented framework where the collection of data for the UDI and all sharing of the UDI data are kept strictly under the user's control. In addition, being service-oriented allows (with the user's permission) its monitoring and analysis services to be easily used by 3rd parties in order to provide 3rd party adaptive assistant services. An example 3rd party service demonstrator, built on top of BaranC, proactively assists a user by dynamic predication, based on the current context, what apps and contacts the user is likely to need. BaranC introduces an innovative user-controlled unified service model of monitoring and use of personal digital activity data in order to provide adaptive user-centred applications. This aims to improve on the current situation where the diversity of adaptive applications results in a proliferation of applications monitoring and using personal data, resulting in a lack of clarity, a dispersal of data, and a diminution of user control.

A simultaneous extraction of context and community from pervasive signals using nested dirichlet process

Understanding user contexts and group structures plays a central role in pervasive computing. These contexts and community structures are complex to mine from data collected in the wild due to the unprecedented growth of data, noise, uncertainties and complexities. Typical existing approaches would first extract the latent patterns to explain human dynamics or behaviors and then use them as a way to consistently formulate numerical representations for community detection, often via a clustering method. While being able to capture high-order and complex representations, these two steps are performed separately. More importantly, they face a fundamental difficulty in determining the correct number of latent patterns and communities. This paper presents an approach that seamlessly addresses these challenges to simultaneously discover latent patterns and communities in a unified Bayesian nonparametric framework. Our Simultaneous Extraction of Context and Community (SECC) model roots in the nested Dirichlet process theory which allows a nested structure to be built to summarize data at multiple levels. We demonstrate our framework on five datasets where the advantages of the proposed approach are validated.

The effect of gender context on children's social behaviour in a limited resources situation : an observational study

Knowing when to compete and when to cooperate to maximize opportunities for equal access to activities and materials in groups is critical to children's social and cognitive development. The present study examined the individual (gender, social competence) and contextual factors (gender context) that may determine why some children are more successful than others. One hundred and fifty-six children (M age=6.5 years) were divided into 39 groups of four and videotaped while engaged in a task that required them to cooperate in order to view cartoons. Children within all groups were unfamiliar to one another. Groups varied in gender composition (all girls, all boys, or mixed-sex) and social competence (high vs. low). Group composition by gender interaction effects were found. Girls were most successful at gaining viewing time in same-sex groups, and least successful in mixed-sex groups. Conversely, boys were least successful in same-sex groups and most successful in mixed-sex groups. Similar results were also found at the group level of analysis; however, the way in which the resources were distributed differed as a function of group type. Same-sex girl groups were inequitable but efficient whereas same-sex boy groups were more equitable than mixed groups but inefficient compared to same-sex girl groups. Social competence did not influence children's behavior. The findings from the present study highlight the effect of gender context on cooperation and competition and the relevance of adopting an unfamiliar peer paradigm when investigating children's social behavior.