The effects of word prediction and text-to-speech technologies on the narrative writing skills of students with specific learning disabilities

This study investigated the effects of word prediction and text-to-speech on the narrative composition writing skills of 6, fifth-grade Hispanic boys with specific learning disabilities (SLD). A multiple baseline design across subjects was used to explore the efficacy of word prediction and text-to-speech alone and in combination on four dependent variables: writing fluency (words per minute), syntax (T-units), spelling accuracy, and overall organization (holistic scoring rubric). Data were collected and analyzed during baseline, assistive technology interventions, and at 2-, 4-, and 6-week maintenance probes. ^ Participants were equally divided into Cohorts A and B, and two separate but related studies were conducted. Throughout all phases of the study, participants wrote narrative compositions for 15-minute sessions. During baseline, participants used word processing only. During the assistive technology intervention condition, Cohort A participants used word prediction followed by word prediction with text-to-speech. Concurrently, Cohort B participants used text-to-speech followed by text-to-speech with word prediction. ^ The results of this study indicate that word prediction alone or in combination with text-to-speech has a positive effect on the narrative writing compositions of students with SLD. Overall, participants in Cohorts A and B wrote more words, more T-units, and spelled more words correctly. A sign test indicated that these perceived effects were not likely due to chance. Additionally, the quality of writing improved as measured by holistic rubric scores. When participants in Cohort B used text-to-speech alone, with the exception of spelling accuracy, inconsequential results were observed on all dependent variables. ^ This study demonstrated that word prediction alone or in combination assists students with SLD to write longer, improved-quality, narrative compositions. These results suggest that word prediction or word prediction with text-to-speech be considered as a writing support to facilitate the production of a first draft of a narrative composition. However, caution should be given to the use of text-to-speech alone as its effectiveness has not been established. Recommendations for future research include investigating the use of these technologies in other phases of the writing process, with other student populations, and with other writing styles. Further, these technologies should be investigated while integrated into classroom composition instruction. ^

An incremental multilinear system for human face learning and recognition

This dissertation establishes a novel system for human face learning and recognition based on incremental multilinear Principal Component Analysis (PCA). Most of the existing face recognition systems need training data during the learning process. The system as proposed in this dissertation utilizes an unsupervised or weakly supervised learning approach, in which the learning phase requires a minimal amount of training data. It also overcomes the inability of traditional systems to adapt to the testing phase as the decision process for the newly acquired images continues to rely on that same old training data set. Consequently when a new training set is to be used, the traditional approach will require that the entire eigensystem will have to be generated again. However, as a means to speed up this computational process, the proposed method uses the eigensystem generated from the old training set together with the new images to generate more effectively the new eigensystem in a so-called incremental learning process. In the empirical evaluation phase, there are two key factors that are essential in evaluating the performance of the proposed method: (1) recognition accuracy and (2) computational complexity. In order to establish the most suitable algorithm for this research, a comparative analysis of the best performing methods has been carried out first. The results of the comparative analysis advocated for the initial utilization of the multilinear PCA in our research. As for the consideration of the issue of computational complexity for the subspace update procedure, a novel incremental algorithm, which combines the traditional sequential Karhunen-Loeve (SKL) algorithm with the newly developed incremental modified fast PCA algorithm, was established. In order to utilize the multilinear PCA in the incremental process, a new unfolding method was developed to affix the newly added data at the end of the previous data. The results of the incremental process based on these two methods were obtained to bear out these new theoretical improvements. Some object tracking results using video images are also provided as another challenging task to prove the soundness of this incremental multilinear learning method.

Document understanding using data mining and machine learning techniques

With the explosive growth of the volume and complexity of document data (e.g., news, blogs, web pages), it has become a necessity to semantically understand documents and deliver meaningful information to users. Areas dealing with these problems are crossing data mining, information retrieval, and machine learning. For example, document clustering and summarization are two fundamental techniques for understanding document data and have attracted much attention in recent years. Given a collection of documents, document clustering aims to partition them into different groups to provide efficient document browsing and navigation mechanisms. One unrevealed area in document clustering is that how to generate meaningful interpretation for the each document cluster resulted from the clustering process. Document summarization is another effective technique for document understanding, which generates a summary by selecting sentences that deliver the major or topic-relevant information in the original documents. How to improve the automatic summarization performance and apply it to newly emerging problems are two valuable research directions. To assist people to capture the semantics of documents effectively and efficiently, the dissertation focuses on developing effective data mining and machine learning algorithms and systems for (1) integrating document clustering and summarization to obtain meaningful document clusters with summarized interpretation, (2) improving document summarization performance and building document understanding systems to solve real-world applications, and (3) summarizing the differences and evolution of multiple document sources.

Exploring the goals, content, and methods of entrepreneurship professors: A multiple case study

Along with the accumulation of evidence supporting the role of entrepreneurship in economic development (Acs & Armington, 2006; Kuratko, 2005, Reynolds, 2007), governments have persisted in encouraging people to become entrepreneurs (Acs & Stough, 2008; Brannback & Carsrud, 2008). These efforts have tried to reproduce the conditions under which entrepreneurship emerges. One of these conditions is to develop entrepreneurial skills among students and scientists (Fan & Foo, 2004). Entrepreneurship education within higher education has experienced a remarkable expansion in the last 20 years (Green, 2008). To develop entrepreneurial skills among students, scholars have proposed different teaching approaches. However, no clear relationship has been demonstrated between entrepreneurship education, learning outcomes, and business creation (Hostager & Decker, 1999). Despite policy makers demands for more accountability from educational institutions (Klimoski, 2007) and entrepreneurship instructors demands for consistency about what should be taught and how (Maidment, 2009), the appropriate content for entrepreneurship programs remains under constant discussion (Solomon, 2007). Entrepreneurship education is still in its infancy, professors propose diverse teaching goals and radically different teaching methods. This represents an obstacle to development of foundational and consistent curricula across the board (Cone, 2008). Entrepreneurship education is in need of a better conceptualization of the learning outcomes pursued in order to develop consistent curriculum. Many schools do not have enough qualified faculty to meet the growing student demand and a consistent curriculum is needed for faculty development. Entrepreneurship instructors and their teaching practices are of interest because they have a role in producing the entrepreneurs needed to grow the economy. This study was designed to understand instructors’ perspectives and actions related to their teaching. The sample studied consisted of eight college and university entrepreneurship instructors. Cases met predetermined criteria of importance followed maximum variation strategies. Results suggest that teaching content were consistent across participants while different teaching goals were identified: some instructors inspire and develop general skills of students while others envision the creation of a real business as the major outcome of their course. A relationship between methods reported by instructors and their disciplinary background, teaching perspective, and entrepreneurial experience was found.

The use of computer graphic organizers for narrative writing by elementary school students with Specific Learning Disabilities

Writing is an academic skill critical to students in today's schools as it serves as a predominant means for demonstrating knowledge during school years (Graham, 2008). However, for many students with Specific Learning Disabilities (SLD), learning to write is a challenging, complex process (Lane, Graham, Harris, & Weisenbach, 2006). Students SLD have substantial writing challenges related to the nature of their disability (Mayes & Calhoun, 2005). ^ This study investigated the effects of computer graphic organizer software on the narrative writing compositions of four, fourth- and fifth-grade, elementary-level boys with SLD. A multiple baseline design across subjects was used to explore the effects of the computer graphic organizer software on four dependent variables: total number of words, total planning time, number of common story elements, and overall organization. ^ Prior to baseline, participants were taught the fundamentals of narrative writing. Throughout baseline and intervention, participants were read a narrative writing prompt and were allowed up to 10 minutes to plan their writing, followed by 15 minutes for writing, and 5 minutes of editing. During baseline, all planning was done using paper and pencil. During intervention, planning was done on the computer using a graphic organizer developed from the software program Kidspiration 3.0 (2011). All compositions were written and editing was done using paper and pencil during baseline and intervention. ^ The results of this study indicated that to varying degrees computer graphic organizers had a positive effect on the narrative writing abilities of elementary aged students with SLD. Participants wrote more words (from 54.74 to 96.60 more), planned for longer periods of time (from 4.50 to 9.50 more minutes), and included more story elements in their compositions (from 2.00 to 5.10 more out of a possible 6). There were nominal to no improvements in overall organization across the 4 participants. ^ The results suggest that teachers of students with SLD should considering use computer graphic organizers in their narrative writing instruction, perhaps in conjunction with remedial writing strategies. Future investigations can include other types of writing genres, other stages of writing, participants with varied demographics and their use combined with remedial writing instruction. ^

The Effect of Mobile Technology as an Active Student Response System on the Acquisition of U.S. History Content of Secondary Students with Specific Learning Disabilities

Students with specific learning disabilities (SLD) typically learn less history content than their peers without disabilities and show fewer learning gains. Even when they are provided with the same instructional strategies, many students with SLD struggle to grasp complex historical concepts and content area vocabulary. Many strategies involving technology have been used in the past to enhance learning for students with SLD in history classrooms. However, very few studies have explored the effectiveness of emerging mobile technology in K-12 history classrooms. This study investigated the effects of mobile devices (iPads) as an active student response (ASR) system on the acquisition of U.S. history content of middle school students with SLD. An alternating treatments single subject design was used to compare the effects of two interventions. There were two conditions and a series of pretest probesin this study. The conditions were: (a) direct instruction and studying from handwritten notes using the interactive notebook strategy and (b) direct instruction and studying using the Quizlet App on the iPad. There were three dependent variables in this study: (a) percent correct on tests, (b) rate of correct responses per minute, and (c) rate of errors per minute. A comparative analysis suggested that both interventions (studying from interactive notes and studying using Quizlet on the iPad) had varying degrees of effectiveness in increasing the learning gains of students with SLD. In most cases, both interventions were equally effective. During both interventions, all of the participants increased their percentage correct and increased their rate of correct responses. Most of the participants decreased their rate of errors. The results of this study suggest that teachers of students with SLD should consider a post lesson review in the form of mobile devices as an ASR system or studying from handwritten notes paired with existing evidence-based practices to facilitate students’ knowledge in U.S. history. Future research should focus on the use of other interactive applications on various mobile operating platforms, on other social studies subjects, and should explore various testing formats such as oral question-answer and multiple choice.

Effects of Repeated Readings on Reading Abilities of English Language Learners with Specific Learning Disabilities

This study investigated the effects of repeated readings on the reading abilities of 4, third-, fourth-, and fifth-grade English language learners (ELLs) with specific learning disabilities (SLD). A multiple baseline probe design across subjects was used to explore the effects of repeated readings on four dependent variables: reading fluency (words read correctly per minute; wpm), number of errors per minute (epm), types of errors per minute, and answer to literal comprehension questions. Data were collected and analyzed during baseline, intervention, generalization probes, and maintenance probes. Throughout the baseline and intervention phases, participants read a passage aloud and received error correction feedback. During baseline, this was followed by fluency and literal comprehension question assessments. During intervention, this was followed by two oral repeated readings of the passage. Then the fluency and literal comprehension question assessments were administered. Generalization probes followed approximately 25% of all sessions and consisted of a single reading of a new passage at the same readability level. Maintenance sessions occurred 2-, 4-, and 6-weeks after the intervention ended. The results of this study indicated that repeated readings had a positive effect on the reading abilities of ELLs with SLD. Participants read more wpm, made fewer epm, and answered more literal comprehension questions correctly. Additionally, on average, generalization scores were higher in intervention than in baseline. Maintenance scores were varied when compared to the last day of intervention, however, with the exception of the number of hesitations committed per minute maintenance scores were higher than baseline means. This study demonstrated that repeated readings improved the reading abilities of ELLs with SLD and that gains were generalized to untaught passages. Maintenance probes 2-, 4-, and 6- weeks following intervention indicated that mean reading fluency, errors per minute, and correct answers to literal comprehensive questions remained above baseline levels. Future research should investigate the use of repeated readings in ELLs with SLD at various stages of reading acquisition. Further, future investigations may examine how repeated readings can be integrated into classroom instruction and assessments.

The Use of a Computer Graphic Organizer for Persuasive Composition Writing by Hispanic Students with Specific Learning Disabilities

Many culturally and linguistically diverse (CLD) students with specific learning disabilities (SLD) struggle with the writing process. Particularly, they have difficulties developing and expanding ideas, organizing and elaborating sentences, and revising and editing their compositions (Graham, Harris, & Larsen, 2001; Myles, 2002). Computer graphic organizers offer a possible solution to assist them in their writing. This study investigated the effects of a computer graphic organizer on the persuasive writing compositions of Hispanic middle school students with SLD. A multiple baseline design across subjects was used to examine its effects on six dependent variables: number of arguments and supporting details, number and percentage of transferred arguments and supporting details, planning time, writing fluency, syntactical maturity (measured by T-units, the shortest grammatical sentence without fragments), and overall organization. Data were collected and analyzed throughout baseline and intervention. Participants were taught persuasive writing and the writing process prior to baseline. During baseline, participants were given a prompt and asked to use paper and pencil to plan their compositions. A computer was used for typing and editing. Intervention required participants to use a computer graphic organizer for planning and then a computer for typing and editing. The planning sheets and written composition were printed and analyzed daily along with the time each participant spent on planning. The use of computer graphic organizers had a positive effect on the planning and persuasive writing compositions. Increases were noted in the number of supporting details planned, percentage of supporting details transferred, planning time, writing fluency, syntactical maturity in number of T-units, and overall organization of the composition. Minimal to negligible increases were noted in the mean number of arguments planned and written. Varying effects were noted in the percent of transferred arguments and there was a decrease in the T-unit mean length. This study extends the limited literature on the effects of computer graphic organizers as a prewriting strategy for Hispanic students with SLD. In order to fully gauge the potential of this intervention, future research should investigate the use of different features of computer graphic organizer programs, its effects with other writing genres, and different populations.

Development of a program of radiopaque lesions images via radiographic cases for active learning in oral medicine

Through the creation of this project in English, we have made a file of radiographic images that will be used by third year dental students in order to improve the practical teaching part of the subject of Oral Medicine, essentially by incorporating these files to the Virtual Campus. We have selected the most representative radiopaque radiographic images studied in pathology lectures given. We have prepared a file with 59 radiopaque radiographic images. These lesions have been divided according to their relationship and number with the tooth, into the following groups: “Anatomic radiopacities”, “Periapical radiopacities”, “Solitary radiopacities not necessarily contacting teeth”,“Multiple separate radiopacities”, and “Generalized radiopacities”. We created 4 flowcharts synthesizing the mayor explanatory bases of each pathological process in relation to other pathologies within each location. We have focused primarily in those clinical and radiographic features that can help us differentiate one pathology from another. We believe that by giving the student a knowledge base through each flowchart, as well as provide clinical cases, will start their curiosity to seek new cases on the Internet or try to look for images that we have not been able to locate due to low frequency. In addition, as this project has been done in English, it will provide the students with necessary tools to do a literature search, as most of the medical and dental literature is in English; thus far, providing the student with this material necessary to make the appropriate searched using keywords in English.

An alternative to BlackboardTM? Student use of web-based textbook teaching and learning material

This paper investigates the use of web-based textbook supplementary teaching and learning materials which include multiple choice test banks, animated demonstrations, simulations, quizzes and electronic versions of the text. To gauge their experience of the web-based material students were asked to score the main elements of the material in terms of usefulness. In general it was found that while the electronic text provides a flexible platform for presentation of material there is a need for continued monitoring of student use of this material as the literature suggests that digital viewing habits may mean there is little time spent in evaluating information, either for relevance, accuracy or authority. From a lecturer perspective these materials may provide an effective and efficient way of presenting teaching and learning materials to the students in a variety of multimedia formats, but at this stage do not overcome the need for a VLE such as Blackboard™.

New Advancements of Scalable Statistical Methods for Learning Latent Structures in Big Data

Constant technology advances have caused data explosion in recent years. Accord- ingly modern statistical and machine learning methods must be adapted to deal with complex and heterogeneous data types. This phenomenon is particularly true for an- alyzing biological data. For example DNA sequence data can be viewed as categorical variables with each nucleotide taking four different categories. The gene expression data, depending on the quantitative technology, could be continuous numbers or counts. With the advancement of high-throughput technology, the abundance of such data becomes unprecedentedly rich. Therefore efficient statistical approaches are crucial in this big data era.

Previous statistical methods for big data often aim to find low dimensional struc- tures in the observed data. For example in a factor analysis model a latent Gaussian distributed multivariate vector is assumed. With this assumption a factor model produces a low rank estimation of the covariance of the observed variables. Another example is the latent Dirichlet allocation model for documents. The mixture pro- portions of topics, represented by a Dirichlet distributed variable, is assumed. This dissertation proposes several novel extensions to the previous statistical methods that are developed to address challenges in big data. Those novel methods are applied in multiple real world applications including construction of condition specific gene co-expression networks, estimating shared topics among newsgroups, analysis of pro- moter sequences, analysis of political-economics risk data and estimating population structure from genotype data.

National, inter-institutional, graduate modules – blended learning approaches for PhD education in Ireland

This paper is a case study that describes the design and delivery of national PhD lectures with 40 PhD candidates in Digital Arts and Humanities in Ireland simultaneously to four remote locations, in Trinity College Dublin, in University College Cork, in NUI Maynooth and NUI Galway. Blended learning approaches were utilized to augment traditional teaching practices combining: face-to-face engagement, video-conferencing to multiple sites, social media lecture delivery support – a live blog and micro blogging, shared, open student web presence online. Techniques for creating an effective, active learning environment were discerned via a range of learning options offered to students through student surveys after semester one. Students rejected the traditional lecture format, even through the novel delivery method via video link to a number of national academic institutions was employed. Students also rejected the use of a moderated forum as a means of creating engagement across the various institutions involved. Students preferred a mix of approaches for this online national engagement. The paper discusses successful methods used to promote interactive teaching and learning. These included Peer to peer learning, Workshop style delivery, Social media. The lecture became a national, synchronous workshop. The paper describes how allowing students to have a voice in the virtual classroom they become animated and engaged in an open culture of shared experience and scholarship, create networks beyond their institutions, and across disciplinary boundaries. We offer an analysis of our experiences to assist other educators in their course design, with a particular emphasis on social media engagement.

Learning and Exploiting Camera Geometry for Computer Vision

This work explores the use of statistical methods in describing and estimating camera poses, as well as the information feedback loop between camera pose and object detection. Surging development in robotics and computer vision has pushed the need for algorithms that infer, understand, and utilize information about the position and orientation of the sensor platforms when observing and/or interacting with their environment.

The first contribution of this thesis is the development of a set of statistical tools for representing and estimating the uncertainty in object poses. A distribution for representing the joint uncertainty over multiple object positions and orientations is described, called the mirrored normal-Bingham distribution. This distribution generalizes both the normal distribution in Euclidean space, and the Bingham distribution on the unit hypersphere. It is shown to inherit many of the convenient properties of these special cases: it is the maximum-entropy distribution with fixed second moment, and there is a generalized Laplace approximation whose result is the mirrored normal-Bingham distribution. This distribution and approximation method are demonstrated by deriving the analytical approximation to the wrapped-normal distribution. Further, it is shown how these tools can be used to represent the uncertainty in the result of a bundle adjustment problem.

Another application of these methods is illustrated as part of a novel camera pose estimation algorithm based on object detections. The autocalibration task is formulated as a bundle adjustment problem using prior distributions over the 3D points to enforce the objects' structure and their relationship with the scene geometry. This framework is very flexible and enables the use of off-the-shelf computational tools to solve specialized autocalibration problems. Its performance is evaluated using a pedestrian detector to provide head and foot location observations, and it proves much faster and potentially more accurate than existing methods.

Finally, the information feedback loop between object detection and camera pose estimation is closed by utilizing camera pose information to improve object detection in scenarios with significant perspective warping. Methods are presented that allow the inverse perspective mapping traditionally applied to images to be applied instead to features computed from those images. For the special case of HOG-like features, which are used by many modern object detection systems, these methods are shown to provide substantial performance benefits over unadapted detectors while achieving real-time frame rates, orders of magnitude faster than comparable image warping methods.

The statistical tools and algorithms presented here are especially promising for mobile cameras, providing the ability to autocalibrate and adapt to the camera pose in real time. In addition, these methods have wide-ranging potential applications in diverse areas of computer vision, robotics, and imaging.

Essays in Empirical Operations Management: Bayesian Learning of Service Quality and Structural Estimation of Complementary Product Pricing and Inventory Management

This dissertation contributes to the rapidly growing empirical research area in the field of operations management. It contains two essays, tackling two different sets of operations management questions which are motivated by and built on field data sets from two very different industries --- air cargo logistics and retailing.

The first essay, based on the data set obtained from a world leading third-party logistics company, develops a novel and general Bayesian hierarchical learning framework for estimating customers' spillover learning, that is, customers' learning about the quality of a service (or product) from their previous experiences with similar yet not identical services. We then apply our model to the data set to study how customers' experiences from shipping on a particular route affect their future decisions about shipping not only on that route, but also on other routes serviced by the same logistics company. We find that customers indeed borrow experiences from similar but different services to update their quality beliefs that determine future purchase decisions. Also, service quality beliefs have a significant impact on their future purchasing decisions. Moreover, customers are risk averse; they are averse to not only experience variability but also belief uncertainty (i.e., customer's uncertainty about their beliefs). Finally, belief uncertainty affects customers' utilities more compared to experience variability.

The second essay is based on a data set obtained from a large Chinese supermarket chain, which contains sales as well as both wholesale and retail prices of un-packaged perishable vegetables. Recognizing the special characteristics of this particularly product category, we develop a structural estimation model in a discrete-continuous choice model framework. Building on this framework, we then study an optimization model for joint pricing and inventory management strategies of multiple products, which aims at improving the company's profit from direct sales and at the same time reducing food waste and thus improving social welfare.

Collectively, the studies in this dissertation provide useful modeling ideas, decision tools, insights, and guidance for firms to utilize vast sales and operations data to devise more effective business strategies.

Bayesian Learning with Dependency Structures via Latent Factors, Mixtures, and Copulas

Bayesian methods offer a flexible and convenient probabilistic learning framework to extract interpretable knowledge from complex and structured data. Such methods can characterize dependencies among multiple levels of hidden variables and share statistical strength across heterogeneous sources. In the first part of this dissertation, we develop two dependent variational inference methods for full posterior approximation in non-conjugate Bayesian models through hierarchical mixture- and copula-based variational proposals, respectively. The proposed methods move beyond the widely used factorized approximation to the posterior and provide generic applicability to a broad class of probabilistic models with minimal model-specific derivations. In the second part of this dissertation, we design probabilistic graphical models to accommodate multimodal data, describe dynamical behaviors and account for task heterogeneity. In particular, the sparse latent factor model is able to reveal common low-dimensional structures from high-dimensional data. We demonstrate the effectiveness of the proposed statistical learning methods on both synthetic and real-world data.