The use of active compliance enables robots to carry out tasks in the presence of significant sensing and control errors. Compliant motions are quite difficult for humans to specify, however. Furthermore, robot programs are quite sensitive to details of geometry and to error characteristics and must, therefore, be constructed anew for each task. These factors motivate the need for automatic synthesis tools for robot programming, especially for compliant motion. This paper describes a formal approach to the synthesis of compliant motion strategies from geometric descriptions of assembly operations and explicit estimates of errors in sensing and control. A key aspect of the approach is that it provides correctness criteria for compliant motion strategies.

Thesis (Ph.D. (Pastoral))--North-West University, Potchefstroom Campus, 2006.

Emotions play an important role in affecting human¡¦s behaviors. It is obvious especially in the negotiation process. However, these emotional interactions disappear in the electronic communication media. Emoticons provide a new way for negotiators to show their emotions in e-negotiation. E-negotiation support system is composed of electronic communication and decision support system. Using emoticons could help negotiators display their emotions to counterparts. This study explored the effect of emoticon on negotiation process from the aspect of communication. In general, e-communication can be separated into social-emotional communication and task-oriented communication. Emoticons retrieve multiple cues and enable negotiators transmit emotions so that the interactions increase. As a result, more social-emotional communication happens between negotiators. Previous studies indicate that nonverbal cues could add the meaning of messages. In the negotiation process, negotiators may lower down the opportunity of misunderstanding by using emoticons to express their thought in text and increase the quality of communication. The negotiation process may be more effective and increase the ability to reach agreement. Negotiators will be more satisfied of the process and their counterparts when the outcome is equal or better than what they expected. The purpose of this study is to explore whether supporting emoticons influence negotiators communication behavior and change the negotiation process. For this purpose, we conducted an experiment and adopted content analysis to define the social-emotional communication and task-oriented communication behaviors. We also measured the perceived communication effectiveness and satisfaction of negotiation process by questionnaire. The results showed that emoticons increase the frequency of social-emotional communication. Social-emotional communication also increases the negotiators¡¦ task-oriented communication. As a result the communication becomes more effective. Negotiators have better satisfaction of negotiation process.

Stroke rehabilitation is to train the motor function of a patients limb. In this process, functional assessment is of importance, and it is primarily based on a patients task performance. The context of the rehabilitation discussed in this thesis is such that functional assessment is conducted through a computer system and the Internet. In particular, a patient performs the task at home in a haptic virtual environment, and the task performance is transmitted to the therapist over the Internet. One problem with this approach to functional assessment is that a patients mind state is little known to the therapist. This immediately leads to one question, that is, whether an elevated mind state will have some significant effect on the patients task performance? If so, this approach can result in a considerable error. The overall objective of this thesis study was to generate an answer to the aforementioned question. The study focused on a patients elevated fatigue state. The specific objectives of the study include: (i) developing a haptic virtual environment prototype system for functional assessment, (ii) developing a physiological-based inference system for fatigue state, and (iii) performing an experiment to generate knowledge regarding the fatigue effect on task performance. With a limited resource in recruiting patients in the experiment, the study conducted few experiments on patients but mostly on healthy subjects. The study has concluded: (1) the proposed haptic virtual environment system is effective for the wrist coordination task and is likely promising to other tasks, (2) the accuracy of proposed fatigue inference system achieves 89.54%, for two levels of fatigue state, which is promising, (3) the elevated fatigue state significantly affects task performance in the context of wrist coordination task, and (4) the accuracy of the individual-based inference approach is significantly higher than that of the group-based inference approach. The main contributions of the thesis are (1) generation of the new knowledge regarding the fatigue effect on task performance in the context of home-based rehabilitation, (2) provision of the new fatigue inference system with the highest accuracy in comparison with the existing approaches in literature, and (3) generation of the new knowledge regarding the difference between the individual-based inference and group-based inference approaches.

Aquest estudi pretén examinar les estratègies i la producció oral d’aprenents d’anglès com a llengua estrangera en tres tasques comunicatives a l’aula. Els objectius van ser investigar: a) diferències entre les tasques en la producció oral d’aprenents de competència alta i baixa b) diferències entre les tasques en l’ús de les estratègies d’aprenents de competència alta i baixa c) diferències en la producció oral i l’ús de les estratègies entre els grups d’alta i baixa competència d) si les estratègies que els aprenents creuen que fan servir reflecteixen les estratègies que realment fan servir i e) fins a quin punt el qüestionari d’estratègies pot preveure la producció oral. En aquest estudi, hi van participar 48 estudiants universitaris catalans i espanyols amb un nivell de competència oral baix (N=24) i alt (N=24), a més a més de quatre nadius que van actuar com a punts de referència per a la producció oral. Els participants es van gravar en vídeo en parelles, fent les tres tasques: explicant una història en dibuixos (Picture Story), descrivint una obra d’art (Art Description) i omplint buits en la informació proporcionada (Information Gap). Les tasques es van fer en un espai de dues setmanes entre cadascuna. Immediatament després de cada tasca els aprenents van assenyalar les estratègies que havien percebut (PSU) en un qüestionari de 44 ítems (SQ). Una anàlisi factorial va agrupar PSU en cinc categories: les estratègies Interaccionals, Compensatòries, de Manteniment de flux de la conversa, Planificació i Avaluació, que es van utilitzar per a l’anàlisi de diferències en l’ús de les estratègies. Així mateix, una submostra dels grups d’alta (N=4) i baixa (N=4) competència van participar en sessions retrospectives per reflexionar sobre com van actuar durant les tasques. Entre les tres tasques, segons les proves Friedman-Wilcoxon, es van trobar diferències significatives en totes les mesures de producció oral, amb l'excepció de la velocitat de la parla, i l'autoreparació en el grup alt. La complexitat estructural va ser més alta en el Picture Story, la complexitat lexical va ser més alta en l'Art Description i la precisió i fluïdesa van ser més altes en l'Information Gap. Pel que fa a PSU, el 34% de les estratègies van ser diferents en el grup alt i el 9% en el grup baix. En el grup alt la majoria de les diferències van ser en les estratègies Interaccionals i Compensatòries, que es van utilitzar més en l'Information Gap. Entre els grups de competència baixa i alta, segons les proves Mann-Whitney, hi havia diferències significatives en la producció oral. La precisió, la complexitat lèxica, la complexitat estructural i la velocitat de la parla van ser constantment més altes en el grup alt a cada tasca. No obstant, hi havia poques diferències entre les altres mesures de fluïdesa (pauses llargues, repetició, reformulació i autocorreció). En el cas de PSU, hi havia poques (18% max.) diferències entre els grups de diferents competències en cadascuna de les tasques. En general, el grup baix feia servir unes quantes estratègies Compensatòries més. Per validar el qüestionari, el PSU va ser contrastat amb l’ús real de les estratègies (ASU) identificades en les actuacions en les tasques i en els comentaris durant les sessions retrospectives. Primer, el grup alt va ser constant a l’hora de concretar el nivell d’ús d’un 63% de les estratègies comparada amb un 48% en el grup baix. Segon, el grup alt va ser capaç de indicar les diferències en l’ús de les estratègies entre les tres tasques amb més precisió. No obstant, es van trobar més diferències en ASU entre les tasques que les indicades pels dos grups en PSU. Finalment, l’anàlisi de regressió múltiple de les cinc categories d’estratègies i les nou mesures de producció va mostrar que la SQ podria preveure moderadament la precisió (23%), la complexitat lexical (36%) i la velocitat de la parla (31%). Concretament, com més estratègies Compensatòries indicaven els aprenents, més baixa era la precisió, la complexitat lexical i la velocitat de la parla. Els resultats s’expliquen segons els mecanismes de processament de la parla i els trets cognitius i interaccionals de les tasques, i s‘arriba a les conclusions següents: a) l’ús de les estratègies i la producció oral es pot preveure a partir de les característiques de les tasques, b) la tasca influeix en l’ús de les estratègies més que la competència, c) és possible que no es vegin diferències en l’ús de les estratègies entre grups de diferentes competències perquè fan servir les mateixes estratègies però de maneres diferents, d) els qüestionaris sobre estratègies són indicadors moderats de l’ús real de les estratègies i e) la relació entre les estratègies i la producció oral és no-lineal. / Este estudio pretende examinar las estrategias y la producción oral de los estudiantes de inglés como lengua extranjera en tres tareas comunicativas en el aula. Los objetivos fueron investigar: a) diferencias entre las tareas en la producción oral tanto en estudiantes de competencia alta como baja b) diferencias entre el uso de las estrategias tanto en estudiantes de competencia alta como baja c) diferencias en la producción oral y el uso de las estrategias entre los grupos de competencia alta i baja d) si hay correspondencia entre las estrategias que creen que utilizan y las que realmente utilizan y e) hasta qué punto un cuestionario sobre las estrategias refleja la producción oral. En este estudio participaron una muestra de 48 estudiantes universitarios catalanes y españoles con competencia oral baja (N=24) y alta (N=24), además de 4 nativos que actuaron como puntos de referencia para la producción oral. Los participantes se grabaron en video por parejas, realizando tres tareas: explicando una historia en dibujos (Picture Story), describiendo una obra de arte (Art Description) y rellenando lagunas en la información proporcionada (Information Gap). Las tareas se realizaron dejando un periodo de dos semanas entre cada una. Inmediatamente después de cada tarea los estudiantes señalaron las estrategias que creyeron utilizar (PSU) en un cuestionario de 44 artículos (SQ). Un análisis factorial agrupó las estrategias en cinco categorías: Interaccionales, Compensatorias, Mantenimiento del flujo de la Conversación, Planificación y Evaluación, que se utilizó para analizar las diferencias en el uso de las estrategias. Así mismo, una submuestra de los grupos de competencia alta (N=4) y baja (N=4) participaron en sesiones retrospectivas para reflexionar sobre su actuación durante las tareas. Entre las tres tareas, de acuerdo con las pruebas Friedman-Wilcoxon, encontramos diferencias significativas en todas las medidas de producción oral, con la excepción de la velocidad de habla y la autocorrección en el grupo alto. La complejidad estructural fue más alta en el Picture Story, la complejidad léxica fue más alta en el Art Description y la precisión y la fluidez fueron más altas en el Information Gap. Por lo que se refiere a PSU, se encontraron diferencias en un 34% de las estrategias en el grup alto y un 9% en el grupo bajo. En el grupo alto, la mayoría de las diferencias se dieron en las estrategias Interaccionales y Compensatorias, más utilizadas en el Information Gap. Entre los grupos de competencia baja y alta, de acuerdo con las pruebas Mann-Whitney, hubo diferencias significativas en la producción oral. La precisión, la complejidad léxica, la complejidad estructural y la velocidad de habla fueron coherentemente más altas en el grupo alto en todas las tareas. No obstante, hubo pocas diferencias entre las otras medidas de fluidez (pausas largas, repetición, reformulación y autocorrección). En el caso de PSU, hubo pocas diferencias (18% max.) entre los grupos de diferente competencia en cada una de las tareas. En general, el grupo bajo utilizó unas cuantas estrategias Compensatorias más. Para validar el cuestionario se contrastó el PSU con el uso real de las estrategias (ASU), identificadas en las tareas y a partir de los comentarios durante las sesiones retrospectivas. Primero, el grupo alto concretó de forma precisa un 63% de las estrategias frente a un 48% en el caso del grupo bajo. Segundo, el grupo alto fue capaz de indicar con más precisión las diferencias de uso de las estrategias en cada una de las tres tareas. No obstante, encontramos más diferencias entre las tareas en ASU que las identificadas en el PSU por los dos grupos. Finalmente, el análisis de regresión múltiple de las cinco categorías de estrategias y las nueve medidas de producción indicó que la SQ podría predecir un 23 % de la precisión, 36% de la complejidad léxica y 31% de la velocidad de habla. A más estrategias Compensatorias, menor precisión y complejidad léxica y menor velocidad de habla. Los resultados se explican de acuerdo con los mecanismos de procesamiento de habla y los rasgos cognitivos e interaccionales de las tareas, y se llega a las conclusiones siguientes: a) el uso de las estrategias y la producción oral se puede predecir a partir de las características de las tareas, b) la tarea influye en el uso de las estrategias más que la competencia, c) es posible que no se reflejen las diferencias en el uso de las estrategias entre grupos de diferente competencia, debido a que, aunque utilizan las mismas estrategias, las utilizan de forma diferente, d) los cuestionarios sobre estrategias son indicadores moderados del uso real de las estrategias y e) la relación entre las estrategias y la producción oral es no-lineal. / This study sought to examine EFL learners’ strategies and spoken production on different types of oral communication tasks in the classroom. The objectives were to investigate: a) across-task differences in spoken production for high and low proficiency learners, b) across-task differences in strategy use for high and low proficiency learners, c) differences in both spoken production and strategy use between high and low proficiency groups, d) whether strategies learners perceived using reflected the strategies they actually used and e) to what extent a strategy questionnaire could predict spoken production. In this study 48 Catalan and Spanish undergraduates participated with low (N=24) and high (N=24) oral proficiency, as well as four native speakers who acted as benchmarks for spoken production. Participants were recorded on video in pairs performing three oral tasks: a Picture Story, an Art Description and an Information Gap. Tasks were carried out two weeks apart. Immediately after each task learners reported their perceived strategy use (PSU) on a 44–item strategy questionnaire (SQ). Factor analysis grouped PSU into five categories: Interactional, Compensation, Conversation flow maintenance, Planning and Evaluating strategies, which were used to analyze differences in strategy use. In addition, a sub-sample from the high (N=4) and low (N=4) proficiency groups participated in stimulated recall sessions to reflect on their task performance. Across the three tasks, according to Friedman-Wilcoxon tests, significant differences in all spoken production measures were found for both proficiency groups, except for speech rate, and self repair for the high group. Structural complexity was highest on the Picture Story, lexical complexity was highest on the Art Description and accuracy and fluency were highest on the Information Gap. As for PSU, 34% strategies differed for the high group compared to 9% for the low group. For the high group most differences were in Interactional and Compensation strategy use, which was higher on the Information Gap. Between low and high proficiency groups, according to Mann-Whitney tests, there were significant differences in spoken production. Accuracy, lexical complexity, structural complexity and speech rate were consistently higher for the high group on every task. However, there were few differences in other fluency measures (long pauses, repetition, reformulation and self- repair). As for PSU, there were few (18% max.) significant differences between proficiency groups on any one task. In general, the low group used a few more Compensation strategies. To validate the questionnaire PSU was contrasted with actual strategy use (ASU), identified in task performances and stimulated recall comments. Firstly, results showed that the high group were consistent in reporting 63% strategies compared to 48% for the low group. Secondly, the high group was able to gauge differences in strategy use across the tasks more accurately. However, more differences in ASU were found across tasks than identified by both groups as PSU. Finally, multiple regression analysis with the five strategy categories and nine production measures revealed that the SQ was a weak predictor of accuracy (23%), lexical complexity (36%) and speech rate (31%). In particular, the more Compensation strategies learners perceived using, the lower their accuracy, lexical complexity and speech rate. Results are discussed in terms of speech processing mechanisms and cognitive and interactional task features with the following conclusions: a) strategy use and spoken production can be predicted from task characteristics, b) task influences strategy use more than proficiency, c) differences in strategy use may not be observed between proficiency groups, because they may use the same strategies but in different ways, d) strategy questionnaires are moderate indicators of actual strategy use and e) the relationship between strategies and spoken production is non-linear.

In recent years parallel computing has become ubiquitous. Lead by the spread of commodity multicore processors, parallel programming is not anymore an obscure discipline only mastered by a few.Unfortunately, the amount of able parallel programmers has not increased at the same speed because is not easy to write parallel codes.Parallel programming is inherently different from sequential programming. Programmers must deal with a whole new set of problems: identification of parallelism, work and data distribution, load balancing, synchronization and communication.Parallel programmers have embraced several languages designed to allow the creation of parallel applications. In these languages, the programmer is not only responsible of identifying the parallelism but also of specifying low-level details of how the parallelism needs to exploited (e.g. scheduling, thread distribution ...). This is a burden than hampers the productivity of the programmers.We demonstrate that is possible for the runtime component of a parallel environment to adapt itself to the application and the execution environment and thus reducing the burden put into the programmer. For this purpose we study three different parameters that are involved in the parallel exploitation of the OpenMP parallel language: parallel loop scheduling, thread allocation in multiple levels of parallelism and task granularity control.In all the cases, we propose a self-tuned algorithm that will first perform an on-line profiling of the application and based on the information gathered it will adapt the value of the parameter to the one that maximizes the performance of the application.Our goal is not to develop methods that outperform a hand-tuned application for a specific scenario, as this is probably just as difficult as compiler code outperforming hand-tuned assembly code, but methods that get close to that performance with a minimum effort from the programmer. In other words, what we want to achieve with our self-tuned algorithms is to maximize the ratio performance over effort so the entry level to the parallelism is lower. The evaluation of our algorithms with different applications shows that we achieve that goal.

Previous research has identified commitment as one of the major contributing factors in group conflict and this study extends this line of research by considering group conflict as a result of varying members’ prior commitment for groups working on judgmental tasks with a unanimous decision rule. The goal of the group working on a judgmental task is to reach consensus but unanimous decision rule further complicates the decision making process as it requires every member to agree with group’s decision. Given this setting, prior commitment was expected to result in higher group conflict and this elevated level of conflict to have more negative affect on the group. With three judgmental cases formulated, subjects were divided into post-decision and pre-decision groups. In the post-decision groups, each subject was individually presented with each case first and committed to his own decision. With these prior commitments, subjects reached a unanimous decision through a group discussion. In the pre-decision groups, subjects reached a group decision without any prior commitment to their own decisions. Results showed signs of prior commitment contributing to group conflict. Furthermore, the post-experimental questionnaire showed that higher group conflict translates to more negative affects on the perceived performance, the extent to which members agree with the group’s decision, feeling toward other members and willingness to work with the same group.

Previous research has identified commitment as one of the major contributing factors in group conflict and this study extends this line of research by considering group conflict as a result of varying members’ prior commitment for groups working on judgmental tasks with a unanimous decision rule. The goal of the group working on a judgmental task is to reach consensus but unanimous decision rule further complicates the decision making process as it requires every member to agree with group’s decision. Given this setting, prior commitment was expected to result in higher group conflict and this elevated level of conflict to have more negative affect on the group. With three judgmental cases formulated, subjects were divided into post-decision and pre-decision groups. In the post-decision groups, each subject was individually presented with each case first and committed to his own decision. With these prior commitments, subjects reached a unanimous decision through a group discussion. In the pre-decision groups, subjects reached a group decision without any prior commitment to their own decisions. Results showed signs of prior commitment contributing to group conflict. Furthermore, the post-experimental questionnaire showed that higher group conflict translates to more negative affects on the perceived performance, the extent to which members agree with the group’s decision, feeling toward other members and willingness to work with the same group.

Probability matching is the tendency to predict outcomes in accordance with their actual contingencies in a binary choice task. It is, however, a suboptimal response if the aim is to maximize correct predictions. I review two theories that attempt to explain why probability matching occurs: the pattern-search hypothesis and dual-systems theory. These theories are tested in two studies which suggest that dual-systems theory provides a better account of probability matching behavior. Studies 3, 4, and 5 then provide evidence for an extension of the dual-systems theory, called expectation matching, which is intended to explain why probability matching is the intuitive response to a binary choice problem.

Stroke rehabilitation is to train the motor function of a patients limb. In this process, functional assessment is of importance, and it is primarily based on a patients task performance. The context of the rehabilitation discussed in this thesis is such that functional assessment is conducted through a computer system and the Internet. In particular, a patient performs the task at home in a haptic virtual environment, and the task performance is transmitted to the therapist over the Internet. One problem with this approach to functional assessment is that a patients mind state is little known to the therapist. This immediately leads to one question, that is, whether an elevated mind state will have some significant effect on the patients task performance? If so, this approach can result in a considerable error. The overall objective of this thesis study was to generate an answer to the aforementioned question. The study focused on a patients elevated fatigue state. The specific objectives of the study include: (i) developing a haptic virtual environment prototype system for functional assessment, (ii) developing a physiological-based inference system for fatigue state, and (iii) performing an experiment to generate knowledge regarding the fatigue effect on task performance. With a limited resource in recruiting patients in the experiment, the study conducted few experiments on patients but mostly on healthy subjects. The study has concluded: (1) the proposed haptic virtual environment system is effective for the wrist coordination task and is likely promising to other tasks, (2) the accuracy of proposed fatigue inference system achieves 89.54%, for two levels of fatigue state, which is promising, (3) the elevated fatigue state significantly affects task performance in the context of wrist coordination task, and (4) the accuracy of the individual-based inference approach is significantly higher than that of the group-based inference approach. The main contributions of the thesis are (1) generation of the new knowledge regarding the fatigue effect on task performance in the context of home-based rehabilitation, (2) provision of the new fatigue inference system with the highest accuracy in comparison with the existing approaches in literature, and (3) generation of the new knowledge regarding the difference between the individual-based inference and group-based inference approaches.