We are excited to support the innovative work done by researchers connected to the Digital Innovation Lab. Read more about the impact of these projects and how you can become involved.
The following are research projects supported in various ways through the Digital Innovation Lab. Please scroll down or click Read More for further details about each of these innovative projects.
Principal investigator: Dr. Gregory Tweedie
Principal Investigator: Dr. Gregory Tweedie
Principal investigator: Dr. Gregory Tweedie
Principal investigator: Dr. Gregory Tweedie
Principal Investigator: Dr. Barbara Brown
Principal investigator: Dr. Gregory Tweedie
Principal Investigator: Dr. Barbara Brown
Principal Investigator: Dr. Roswita Dressler
Co-Investigator: Dr. Katherine Mueller
Collaborator: Dr. Bernd Nuss
Read more about the impact of new research projects that are connected to the Digital Innovation Lab.
Mitacs Accelerate Entrepreneur $60,000
Principal investigator: Dr. Gregory Tweedie
Project description: Immigration-receiving countries rely on language proficiency tests in determining suitability for high-stakes decisions like study program admission, employment, residency, and citizenship. Increasingly, such tests are administered by automated scoring systems (Foltz et al., 2020), employing artificial intelligence (AI) machine learning (ML) models. In arguing for socially just educational measurement, Stein (2016) asserts that participants in measurement practices have a right to determine measures most relevant to their needs; therefore in just measurement, “measurement infrastructures are a rightful subject for deliberative democratic decision-making” (p. 96). In contrast, AI and ML are critiqued for “black box” elements leading to inscrutable and unexplainable models (Pardos et al., 2019). What is needed is a machine learning application for language testing aligning with levels of individual autonomy characteristic of just educational measurement, in transparent and scrutable ways. In this project, we make the case for a more deliberative democratisation of the language testing process from the perspective of test-takers, through the application of Value-Sensitive Algorithm Design (Zhu et al, 2018) to a predictive machine learning technology.
For further information or to become involved:
Google Cloud Research Credits Program $6,452
Principal investigator: Dr. Gregory Tweedie
Project description: Language testing is a multi-billion-dollar business. 3 million IELTS tests are held yearly, at an average cost of USD $250/test, with annual revenues of USD $750 million. ETS (which administers the TOEFL and TOEIC tests) generates annual revenues of $USD 1 billion. Canada’s 530,000 international students take these language tests, as do many of its 300,000 new immigrants annually. Combining research expertise from data science, artificial intelligence, machine learning and applied linguistics, the researchers made use of the Google Cloud Research Credits program to access the Vertex AI Machine Learning platform. With the assistance of Innovate Calgary, the researchers have created through Vertex AI a proprietary algorithm to predict the results of second language tests. Research shows that test takers make an average of 3 attempts (at USD $250 per attempt) before achieving their required score and resulting in delays of 18-24 months to access employment or programs of study, with the attendant costs of unemployment or underemployment. The platform will provide an accurate (presently >86%) prediction of test results, for 15% of the cost of a single language test. The beta version launch of the predictive platform is set for June 30, 2022.
For further information or to become involved:
Principal investigator: Dr. Gregory Tweedie
Project description: Canada endures a chronic nursing shortage, but an inefficient and inequitable English language testing process unnecessarily delays workforce access of otherwise qualified nurses. The aim of this research is to investigate the potential of automated scoring (AS) methods in enhancing and assessing the language proficiency of internationally educated nurses (IENs), using the descriptive scale of the CELBAN nursing English language test (Canadian English Language Benchmarks Assessment for Nurses), to streamline their deployment to the Canadian healthcare workforce. The findings from this research will contribute directly to the health of Canadians through streamlining integration of internationally educated nurses into Canada’s healthcare workforce, by examining the affordances of technology in reducing the current language testing bottleneck in IENs’ employment access.
For further information or to become involved as:
Principal investigator: Dr. Gregory Tweedie
Project description: ASDSS is a framework for decision support of English language proficiency assessment and English language teaching developed using artificial intelligence, machine learning and automated scoring systems. The ASDSS web-based application will draw upon the Alberta K-12 ESL Proficiency Benchmarks descriptive scale, given that the Benchmarks is widely respected and cited as a model framework for K-12 English language proficiency assessment and monitoring. It is proposed that the ASDSS would utilize the current version of the Alberta K-12 ESL Proficiency Benchmarks (2010) and/or any newer versions of the Benchmarks as available. The ASDSS tool would provide teachers with an automated Benchmarks level to inform teachers’ benchmarking decision, highlighting features of student language performance which meet/do not meet Benchmarks thresholds. Additionally, the ASDSS would suggest teaching points and language learning activities based on student performance. ASDSS would link compatibly with existing school grading and assessment systems to streamline assessment, record-keeping and the ongoing monitoring of student language learning progress.
For further information or to become involved as:
Principal Investigator: Dr. Barbara Brown
Alberta Education Funded
Project Description: This study has been designed as a research-practice partnership with a local school district. The purpose of this study is to examine the impact of technology-enhanced assessment strategies on students’ learning in a flexible high school learning environment. Through a Google Classroom professional learning forum and series, the collaborative team of researchers and practitioners will use a proven synchronous and asynchronous approach to engage teachers at a high school level in a cyclic design process with an emphasis on creating technology-enhanced strategies for formative and summative assessments. Design research in education offers a practical research approach for developing assessment practices with a unique high school population offering continuous progress in a personalized and individualized learning environment.
For further information or to become involved as:
Principal investigator: Dr. Gregory Tweedie
Project description: English 30-1 is a requirement for postsecondary admission for all students, and newcomer English Language Learners find the written diploma component of this compulsory subjects a particular challenge, given the cultural background required (e.g., knowledge of Shakespearean literature, classic poetry, Western rhetorical patterns, etc.). This proposed project will explore the predictive capabilities of machine learning (ML) to identify linguistic and textual features of English as a Second/Additional Language writing which are predictive of diploma exam success for English 30-1. In collaboration with Alberta Education, and a school district with a large number of English Language Learners, anonymized English 30-1 responses will be analyzed using ML algorithms for linguistic pattern recognition of student scores. The findings will then be used to develop a learning-focused application which applies the research findings to enhance learner diploma success - by supporting both teacher and student user interfaces.
For further information or to become involved as:
Principal Investigator: Dr. Barbara Brown
SSHRC Funded
Project Description: This project seeks to explore novel ways of designing group work in online courses to promote professional collaboration and connectedness. Findings in the learning sciences are demonstrating that the design of collaborative knowledge-building learning environments can shape how students learn and develop professional collaboration skills (Barkley et al., 2014; Brouwer & Jansen, 2019; Friesen, 2009; Rios et al., 2020; Scardamalia & Bereiter, 2014). However, instructors and students often avoid group work due to challenges such as inequitable work contributions, difficulty in scheduling group meetings, and unfair group assessment (Berlin & White, 2012; Brown et al., 2018; Thom, 2020; Thomas & Brown, 2017), which can result in designing learning activities in online courses that perpetuate individual work. In this case study, researchers are examining how instructors design group work in post-secondary online teacher education courses.
For further information or to become involved:
VPR Catalyst Grant $10,730
Principal Investigator: Dr. Roswita Dressler
Co-Investigator: Dr. Katherine Mueller
Collaborator: Dr. Bernd Nuss
Project description: Canadian educational programs are known globally for the language learning opportunities that are available to school children. However, in some languages (e.g., German), teachers have identified a challenge in finding suitable books and digital resources for teaching reading in the second language (Dressler, 2012; 2018). A lack of suitable instructional materials impacts students’ potential to achieve the high levels of language proficiency needed to compete in a global marketplace. The aim of this research is to investigate and develop readability rating methods to inform the choice of texts for teaching reading in German. The findings from this research will contribute toward strengthening the quality of second language education by applying linguistic research and practitioner expertise to solving the ongoing problem of how to efficiently assess the suitability of German language texts for use in bilingual education.
For further information or to assist with:
Transdisciplinary Scholarship Connector: $8,491
Principal Investigator: Dr Gregory Tweedie
Co-investigator: Dr Meagan LaRiviere
Co-investigator: Crystal Lawrence
Project description: Canada endures a chronic shortage of nurses – a reality especially evident during the COVID-19 pandemic (Lopez et al, 2022) – and the recruitment and deployment of internationally educated nurses (IENs) as one means of filling this gap is a feature of health policy across provinces, including Alberta (Nordstrom et al. 2018). However, an inconsistent, unclear, haphazard and bureaucratic process of credentialing, particularly the procedure for determining equivalency of nursing competencies and qualifications, often results in long periods of workforce deployment: one study showed an average length of 656 days for the registration process (Giblin et al, 2016). Even after a substantial study and policy interventions, the total length of time did not substantially change except for those applicants who opted to bypass the assessment process and move right into a nurse education bridging program (Giblin, et al. 2016).
The process of determining IENs’ eligibility for nursing practice in Canada involves substantial equivalent competency assessment (SEC; see Kwan et al, 2017) that compares prior education and experience in the IEN’s home country to entry-level Canadian nursing competencies.
Artificial intelligence / machine learning technology holds promise to significantly reduce delays in the SEC process. A machine learning model trained to map prior qualifications to Alberta nursing competencies, presented in an accessible user interface which IENs can access in their home country at a fraction of the cost, will significantly reduce time to workforce deployment of qualified IENs, reducing the strain on our healthcare system, and on IENs themselves, who as newcomers to Canada are vulnerable to under-/unemployment and the accompanying challenges, while awaiting the SEC process.