Tracking a multitude of abilities as they develop.

Recently, the Urnings algorithm (Bolsinova et al., 2022, J. R. Stat. Soc. Ser. C Appl. Statistics, 71, 91) has been proposed that allows for tracking the development of abilities of the learners and the difficulties of the items in adaptive learning systems. It is a simple and scalable algorithm which is suited for large-scale applications in which large streams of data are coming into the system and on-the-fly updating is needed. Compared to alternatives like the Elo rating system and its extensions, the Urnings rating system allows the uncertainty of the ratings to be evaluated and accounts for adaptive item selection which, if not corrected for, may distort the ratings. In this paper we extend the Urnings algorithm to allow for both between-item and within-item multidimensionality. This allows for tracking the development of interrelated abilities both at the individual and the population level. We present formal derivations of the multidimensional Urnings algorithm, illustrate its properties in simulations, and present an application to data from an adaptive learning system for primary school mathematics called Math Garden.

SYMBALS: A Systematic Review Methodology Blending Active Learning and Snowballing.

Research output has grown significantly in recent years, often making it difficult to see the forest for the trees. Systematic reviews are the natural scientific tool to provide clarity in these situations. However, they are protracted processes that require expertise to execute. These are problematic characteristics in a constantly changing environment. To solve these challenges, we introduce an innovative systematic review methodology: SYMBALS. SYMBALS blends the traditional method of backward snowballing with the machine learning method of active learning. We applied our methodology in a case study, demonstrating its ability to swiftly yield broad research coverage. We proved the validity of our method using a replication study, where SYMBALS was shown to accelerate title and abstract screening by a factor of 6. Additionally, four benchmarking experiments demonstrated the ability of our methodology to outperform the state-of-the-art systematic review methodology FAST2.

Combining Supervised and Unsupervised Machine Learning Methods for Phenotypic Functional Genomics Screening.

There has been an increase in the use of machine learning and artificial intelligence (AI) for the analysis of image-based cellular screens. The accuracy of these analyses, however, is greatly dependent on the quality of the training sets used for building the machine learning models. We propose that unsupervised exploratory methods should first be applied to the data set to gain a better insight into the quality of the data. This improves the selection and labeling of data for creating training sets before the application of machine learning. We demonstrate this using a high-content genome-wide small interfering RNA screen. We perform an unsupervised exploratory data analysis to facilitate the identification of four robust phenotypes, which we subsequently use as a training set for building a high-quality random forest machine learning model to differentiate four phenotypes with an accuracy of 91.1% and a kappa of 0.85. Our approach enhanced our ability to extract new knowledge from the screen when compared with the use of unsupervised methods alone.

Tracking with (Un)Certainty.

One of the highest ambitions in educational technology is the move towards personalized learning. To this end, computerized adaptive learning (CAL) systems are developed. A popular method to track the development of student ability and item difficulty, in CAL systems, is the Elo Rating System (ERS). The ERS allows for dynamic model parameters by updating key parameters after every response. However, drawbacks of the ERS are that it does not provide standard errors and that it results in rating variance inflation. We identify three statistical issues responsible for both of these drawbacks. To solve these issues we introduce a new tracking system based on urns, where every person and item is represented by an urn filled with a combination of green and red marbles. Urns are updated, by an exchange of marbles after each response, such that the proportions of green marbles represent estimates of person ability or item difficulty. A main advantage of this approach is that the standard errors are known, hence the method allows for statistical inference, such as testing for learning effects. We highlight features of the Urnings algorithm and compare it to the popular ERS in a simulation study and in an empirical data example from a large-scale CAL application.

Dynamic estimation in the extended marginal Rasch model with an application to mathematical computer-adaptive practice.

We introduce a general response model that allows for several simple restrictions, resulting in other models such as the extended Rasch model. For the extended Rasch model, a dynamic Bayesian estimation procedure is provided, which is able to deal with data sets that change over time, and possibly include many missing values. To ensure comparability over time, a data augmentation method is used, which provides an augmented person-by-item data matrix and reproduces the sufficient statistics of the complete data matrix. Hence, longitudinal comparisons can be easily made based on simple summaries, such as proportion correct, sum score, etc. As an illustration of the method, an example is provided using data from a computer-adaptive practice mathematical environment.

Improving Comprehension Efficiency of High Content Screening Data Through Interactive Visualizations.

In this study, an experiment is conducted to measure the performance in speed and accuracy of interactive visualizations. A platform for interactive data visualizations was implemented using Django, D3, and Angular. Using this platform, a questionnaire was designed to measure a difference in performance between interactive and noninteractive data visualizations. In this questionnaire consisting of 12 questions, participants were given tasks in which they had to identify trends or patterns. Other tasks were directed at comparing and selecting algorithms with a certain outcome based on visualizations. All tasks were performed on high content screening data sets with the help of visualizations. The difference in time to carry out tasks and accuracy of performance was measured between a group viewing interactive visualizations and a group viewing noninteractive visualizations. The study shows a significant advantage in time and accuracy in the group that used interactive visualizations over the group that used noninteractive visualizations. In tasks comparing results of different algorithms, a significant decrease in time was observed in using interactive visualizations over noninteractive visualizations.

Distinguishing Fast and Slow Processes in Accuracy - Response Time Data.

We investigate the relation between speed and accuracy within problem solving in its simplest non-trivial form. We consider tests with only two items and code the item responses in two binary variables: one indicating the response accuracy, and one indicating the response speed. Despite being a very basic setup, it enables us to study item pairs stemming from a broad range of domains such as basic arithmetic, first language learning, intelligence-related problems, and chess, with large numbers of observations for every pair of problems under consideration. We carry out a survey over a large number of such item pairs and compare three types of psychometric accuracy-response time models present in the literature: two 'one-process' models, the first of which models accuracy and response time as conditionally independent and the second of which models accuracy and response time as conditionally dependent, and a 'two-process' model which models accuracy contingent on response time. We find that the data clearly violates the restrictions imposed by both one-process models and requires additional complexity which is parsimoniously provided by the two-process model. We supplement our survey with an analysis of the erroneous responses for an example item pair and demonstrate that there are very significant differences between the types of errors in fast and slow responses.