Search

Providing early predictions of academic performance is necessary for identifying at-risk students and subsequently providing them with timely intervention for critical factors affecting their academic performance. Although e-book systems are often used to provide students with teaching/learning materials in university courses, seldom has research made the early prediction based on their online reading behaviours by implementing machine learning classifiers. This study explored to what extent university students’ academic achievement can be predicted, based on their reading behaviours in an e-book supported course, using the classifiers. It further investigated which of the features extracted from the reading logs influence the predictions. The participants were 100 first-year undergraduates enrolled in a compulsory course at a university in Taiwan. The results suggest that logistic regression supports vector classification, decision trees, and random forests, and neural networks achieved moderate prediction performance with accuracy, precision, and recall metrics. The Bayes classifier identified almost all at-risk students. Additionally, student online reading behaviours affecting the prediction models included: turning pages, going back to previous pages and jumping to other pages, adding/deleting markers, and editing/removing memos. These behaviours were significantly positively correlated to academic achievement and should be encouraged during courses supported by e-books. Implications for practice or policy: For identifying at-risk students, educators could prioritise using Gaussian naïve Bayes in an e-book supported course, as it shows almost perfect recall performance. Assessors could give priority to logistic regression and neural networks in this context because they have stable achievement prediction performance with different evaluation metrics. The prediction models are strongly affected by student online reading behaviours, in particular by locating/returning to relevant pages and modifying markers.

“Making” has its origin in the maker movement, emphasizes novel, innovative creations by adopting applicable technologies, and fosters students’ creativity or encourages them to engage in creative ...

The goal of this study was to investigate the determinants affecting university students’ attitude and intention towards using a learning management system from the perspective of self-determinatio...

Block-based visual programming tools, such as Scratch, Alice, and MIT App Inventor, provide an intuitive and easy-to-use editing interface through which to promote programming learning for novice students of various ages. However, very little attention has been paid to investigating these tools’ overall effects on students’ academic achievement and the study features that may moderate the effects of block-based visual programming from a comprehensive perspective. Thus, the present study carried out a meta-analysis to systemically examine 29 empirical studies (extracting 34 effect sizes) using experimental or quasi-experiments involving the programming learning effects of employing block-based visual programming tools to date (until the end of 2019). The results showed a small to medium significant positive overall mean effect size (fixed-effect model g = 0.37; random-effects model g = 0.47) of the use of these block-based visual programming tools with respect to students’ academic achievement. Furthermore, the overall mean effect size was significantly affected by the educational stage, programming tool used, experimental treatment, and school location. Discussions and implications based on the findings are provided.

Project-based learning is generally considered an alternative to traditional, teacher-led instruction. However, there is a noticeable lack of meta-analyses with regard to determining its overall effects on students' academic achievement, and what study features may moderate the impacts of project-based learning. This study thus performed a meta-analysis to synthesize existing research that compared the effects of project-based learning and those of traditional instruction on student academic achievement. Forty-six effect sizes (comparisons) extracted from 30 eligible journal articles published from 1998 to 2017 were analyzed, representing 12,585 students from 189 schools in nine countries. The results showed that the overall mean weighted effect size (d+) was 0.71, indicating that project-based learning has a medium to large positive effect on students' academic achievement compared with traditional instruction. In addition, the mean effect size was affected by subject area, school location, hours of instruction, and information technology support, but not by educational stage and small group size.

Spatial light modulator is the key component for dynamic computer generated hologram (CGH) and phase type liquid crystal on silicon (LCoS) device is among the most commercialized. Due to the consideration on efficiency or the requirement from system architecture in the holographic displays, the reconstruction illumination could obliquely incident on the LCoS, which will cause distortion of signal window, hence the image within the signal window. Therefore, the target image needs pre-distortion before the calculation and optimization for the corresponding phase pattern on LCoS in order to obtain the originally desired pattern. This paper illustrates the mechanism of image distortion and provides quantitative analysis for arbitrary azimuthal and radial angle of reconstruction illumination on LCoS device. The required correction is also derived with the corresponding experiment showing the effectiveness of the proposed scheme.

Holographic displays have the feature to show images out of the plane of the device itself, which is especially favored for augmented reality (AR) applications where the images need to be merged with the real world. In existing cases of AR holographic display, a combiner is used to converge the light path of the display image and surrounding scene toward the viewer's eye. In this paper, the idea of combining the holographic device and the combiner has been proposed, resulting in a see-through holographic display. In order to maintain the see-through quality of the device, the concept of partial hologram has been introduced, which means only a part of the area on the device has the holographic fringe pattern while leaving the rest fully transparent. Experiment and theoretical investigation shows that an evenly yet randomly distributed partial hologram provides the best holographic image quality assuming a fixed percentage of the holographic area on the device. A passive computer generated hologram (CGH) with two phase levels has been designed and fabricated for the verification. With partial hologram sharing 25% of the whole area, the CGH exhibits 90.9% of total transmission and 72.2% of parallel transmission. The demonstration shows a high see-through quality while providing a clear holographic image.

Head-up displays (HUDs) have already penetrated into vehicle applications and demand keeps growing. Existing head-up displays have their image fixed at a certain distance in front of the windshield. New development could have two images displayed at two different yet fixed distances simultaneously or switchable upon request. The physical distance of HUD image is associated with the accommodation delay as a safety issue in driving, and could also be a critical parameter for augmented reality (AR) function. In this paper, a novel architecture for HUD has been proposed to make the image distance continuously tunable by exploiting the merit of both holographic and geometrical imaging. Holographic imaging is capable of changing image position by varying the modulation on a spatial light modulator (SLM) without any mechanical movement. Geometrical imaging can easily magnify longitudinal image position with short depth of focus by using large aperture components. A prototype based on liquid crystal on silicon (LCoS) SLM has demonstrated the capability of changing image position from 3 m to 30 m verified with parallax method.

Computer-generated holograms (CGHs) have their phase and/or amplitude modulation pattern calculated rather than recorded as for traditional holograms. In practice, the CGH devices are normally pixelated, no matter if they are passive or active ones. In many cases, the reconstruction light illuminates on the CGH devices obliquely, and the pattern generated on the target plane will be distorted from the originally desired one, even if the modulation on the CGH devices has been calibrated for the corresponding illumination angle in CGH calculation and optimization. The distortion is purely related to the diffraction behavior resulting from the geometry of the pixel on the CGH, and therefore diffractive distortion has been coined for this specific phenomenon. In this paper, quantitative analysis of diffractive distortion and a corresponding scheme for correction have been given based on scalar diffraction theory. The proposed concept is that the distortion of the reconstructed image is proportional to the distortion of the signal window. An experiment has been conducted with a phase-type liquid crystal on silicon spatial light modulator (SLM). Both the distortion analysis and the correction scheme have been verified quantitatively for various illumination angles and the direction of the reconstruction light.

The Ritchey&ndash, Chr&eacute, tien telescope has been the key optical module for remote sensing instruments (RSI), in which the root mean square (RMS) random surface wavefront error and the alignment error of the primary and the secondary mirror takes the highest weighting in the tolerance analysis for the fabrication and assembly of the telescope. Therefore, the higher tolerance of those items becomes preferable for higher efficiency of RSI manufacturing. In this paper, the correlation between those tolerance items and the f-number of the telescope has been investigated. Although the f-number is normally a system parameter well specified in the beginning of the design process, it is not very rigid in practice and has a certain amount of allowable range. The optimal f-number can then be chosen based on the consideration of those key tolerance items. The proposed concept can be generalized as a novel methodology of design for tolerance.

Built-in autonomous stereo vision devices play a critical role in the autonomous docking instruments of space vehicles. Traditional stereo cameras for space autonomous docking use charge-coupled device (CCD) image sensors, and it is difficult for the overall size to be reduced due to the size of the CCD. In addition, only the few outermost elements of the camera lens use radiation-resistant optical glass material. In this paper, a complementary metal&ndash, oxide semiconductor (CMOS) device is used as the image sensor, and radiation-resistant optical glass material is introduced to all lens elements in order to make a compact and highly reliable space grade instrument. Despite the limited available material, a fixed focus module with 7 lens elements and overall length of 42 mm has been achieved, while meeting all the required performance demands for the final vision-guided docking process.

The low birth rate in Taiwan has led to a severe challenge for many universities to enroll a sufficient number of students. Consequently, a large number of students have been admitted to universities regardless of whether they have an aptitude for academic studies. Early diagnosis of students with a high dropout risk enables interventions to be provided early on, which can help these students to complete their studies, graduate, and enhance their future competitiveness in the workplace. Effective prelearning interventions are necessary, therefore students’ learning backgrounds should be thoroughly examined. This study investigated how big data and artificial intelligence can be used to help universities to more precisely understand student backgrounds, according to which corresponding interventions can be provided. For this study, 3552 students from a university in Taiwan were sampled. A statistical learning method and a machine learning method based on deep neural networks were used to predict their probability of dropping out. The results revealed that student academic performance (regarding the dynamics of class ranking percentage), student loan applications, the number of absences from school, and the number of alerted subjects successfully predicted whether or not students would drop out of university with an accuracy rate of 68% when the statistical learning method was employed, and 77% for the deep learning method, in the case of giving first priority to the high sensitivity in predicting dropouts. However, when the specificity metric was preferred, then the two approaches both reached more than 80% accuracy rates. These results may enable the university to provide interventions to students for assisting course selection and enhancing their competencies based on their aptitudes, potentially reducing the dropout rate and facilitating adaptive learning, thereby achieving a win-win situation for both the university and the students. This research offers a feasible direction for using artificial intelligence applications on the basis of a university’s institutional research database.

Students’ prior knowledge contributes to their e-book reading comprehension and might affect the usability of an e-book system. This study explored the influence of university students’ prior knowledge on the usability of an e-book system termed “BookRoll,” which was introduced in the context of an information and technology course in this study. One hundred first-year students (36 male and 64 female students) at a university in central Taiwan participated in this study. Students previewed the lecture slides and used e-book functions such as bookmarks, highlights, and memos during seven weekly lessons. Students’ prior knowledge in the domain of information and technology and the Usefulness, Satisfaction, and Ease of use (USE) dimensions of the BookRoll system were collected before and after the treatment course, respectively. This study revealed that students’ prior domain knowledge had no significant effect on the usability of the e-book system. In general, students’ ratings of BookRoll’s usability were high. Positive qualitative feedback across categories was given by students with high and low prior knowledge; however, some negative but constructive qualitative feedback was also provided. In addition, a short questionnaire to measure the USE dimensions of an e-book system for undergraduate users was developed and provided.

Recent research has been very attentive to the examination of learners’ behavioural patterns of using learning management systems (LMS), but these studies seldom address the diversity in the LMS usage behaviours of teachers. This study aimed to discover the behavioural patterns of university instructors regarding the use of an LMS by using sequential and clustering analysis techniques. The usage behaviours of 268 teachers at a public university in China were extracted from the Blackboard platform over the course of one-semester. These behaviours were classified according to five different LMS behavioural types: (1) course and content; (2) assignment; (3) communication and collaboration; (4) assessment; and (5) administration. The results indicated that the most frequent teachers’ LMS usage behaviour was course and content followed by assessment and administration. The results of the sequential analysis indicated that most of the instructors are used to adopting communication and collaboration and assignment when they finish using course and content. In addition, three distinct usage behavioural pattern subgroups were named as teachers preferred assessment, teachers of regular use, and teachers of less use. Implications of these findings are discussed in the light of the university teachers’ behavioural patterns toward using the LMS.

Aim/Purpose: The present study investigated and compared students’ attention in terms of time-on-task and number of distractors between using a touchscreen and a pen tablet in mathematical problem-solving activities with virtual manipulatives. Background: Although there is an increasing use of these input devices in educational practice, little research has focused on assessing student attention while using touchscreens or pen tablets in a mathematics classroom. Methodology: A qualitative exploration was conducted in a public elementary school in New Taipei, Taiwan. Six fifth-grade students participated in the activities. Video recordings of the activities and the students’ actions were analyzed. Findings: The results showed that students in the activity using touchscreens maintained greater attention and, thus, had more time-on-task and fewer distractors than those in the activity using pen tablets. Recommendations for Practitioners: School teachers could employ touchscreens in mathematics classrooms to support activities that focus on students’ manipulations in relation to the attention paid to the learning content. Recommendation for Researchers: The findings enhance our understanding of the input devices used in educational practice and provide a basis for further research. Impact on Society: The findings may also shed light on the human-technology interaction process involved in using pen and touch technology conditions. Future Research: Activities similar to those reported here should be conducted using more participants. In addition, it is important to understand how students with different levels of mathematics achievement use the devices in the activities.

Two or more focal planes are required in augmented reality head-up displays (AR HUDs) to respectively present basic and interactive driving information to car drivers; whereas, solutions using two separated picture generation units (PGUs) with two sets of optics incur increased cost, reduced reliability, and expanded volume. To develop an AR HUD using a single PGU and a single curved mirror, we propose to set two logically separated regions on a single PGU and optically relay one of them to a new position to create two focal planes. A single freeform mirror acquired through careful optical and mechanical design optimization produces high image quality in an eyebox of 120 mm by 60 mm, simultaneously for a far image (9 m, 10° by 3°) and a near image (2.5 m, 6° by 2°). Finally, an AR HUD product with a compact volume of 8.5 L is fabricated to experimentally verify the design.

Traditional backlights are designed with new dot patterns, and then injection molding, laser beam fabrication, or UV (Ultraviolet) roll-to-plate imprinting is used to apply dot patterns to a light guide plate&mdash, the process consumes considerable time and resources. Therefore, we propose a novel light guide design that does not use a dot pattern. We designed an asymmetric intensity distribution of mini-LEDs (Light Emitting Diode) and a light guide plate with a fully printed diffusion reflection on the bottom surface for a planar illuminator. The design rules for the proposed architecture are described in this paper. The archetype design with a 152.4 mm circular down-light has a diameter of 143 mm for the planar light source module. The experiment achieved a total efficiency of 85% and uniformity of 92.6%.

A laser processing system consisting of two diffractive elements and one refractive element is proposed enabling a Gaussian laser beam to be transformed into two beams with a depth of focus of up to 150 µm and focal spot smaller than 5 µm. For specific laser processing, the two beams are rotatable when the beam-splitting diffractive element is rotated. The overall system is versatile for laser cutting and drilling.

High-brightness white-light-emitting diodes (w-LEDs) with excellent color quality is demonstrated by using nontoxic nanomaterials. Previously, we have reported the high color quality w-LEDs with heavy-metal phosphor and quantum dots (QDs), which may cause environmental hazards. In the present work, liquid-type white LEDs composed of nontoxic materials, named as graphene and porous silicon quantum dots are fabricated with a high color rendering index (CRI) value gain up to 95. The liquid-typed device structure possesses minimized surface temperature and 25% higher value of luminous efficiency as compare to dispensing-typed structure. Further, the as-prepared device is environment friendly and attributed to low toxicity. The low toxicity and high R9 (87) component values were conjectured to produce new or improve current methods toward bioimaging application.

This study demonstrates that flexible white LEDs, doped with diffusion particles and with a square column structure, have excellent luminosity, uniformity, and bending reliability. This large area (5 cm &times, 5 cm) square column flexible device had a smaller thickness (2 mm), and enhancements in both luminous efficiency (29.5%) and uniformity (44.6%) compared to the characteristics of the 6 mm reference sample. Optimization of the reflective layer coating for the square column, flexible white LED was achieved with a higher luminous efficiency (171 lm/w) and uniformity (92%). We designed a novel lightning bolt electrode to improve reliability and bendability. After the bending test, the blue flexible LED had a lower bending diameter (10 mm) but more bending circles (increased to 2000 times.

To realize high-efficiency AlGaN-based deep-ultraviolet light-emitting diodes (DUV-LEDs), enhancing their light-extraction efficiency (LEE) is crucial. This paper proposes an aluminum-based sidewall reflector structure that could replace the conventional ceramic-based packaging method. We design optimization simulations and experimental results demonstrated the light power output could be enhanced 18.38% of DUV-LEDs packaged with the aluminum-based sidewall.

This study presents a novel method for estimating the phosphor conversion efficiency of white light-emitting diodes (WLEDs) with different ratios of phosphors. Numerous attempts have been made for predicting the phosphor conversion efficiency of WLEDs using Monte Carlo ray tracing and the Mie scattering theory. However, because efficiency depends on the phosphor concentration, obtaining a tight match between this model and the experimental results remains a major challenge. An accurate prediction depends on various parameters, including particle size, morphology, and packaging process criteria. Therefore, we developed an efficient model that can successfully correlate the total absorption ratio to the phosphor concentration using a simple equation for estimating the spectra and lumen output. The novel and efficient method proposed here can accelerate WLED development by reducing costs and saving fabrication time.