Creating A Recognition Eco-System

Abstract

This paper describes an Australian initiative to develop industry utility of recognition of prior learning (RPL), based on the application of new competency-based data ontologies that enable granular accreditation and RPL assessments, leveraging Artificial Intelligence (AI) technologies. The platform is a joint venture between research and development consultancy Incept Labs and the Institute of Public Accountants (IPA) to support the IPA's Global Certificate of Public Accounting (GCPA), the first professional accounting qualification achieved solely through the completion of stackable micro credentials. 

A key design objective was to support increased RPL mobility for professionals. Variability in post-secondary education frameworks, inconsistent course documentation, and changing professional requirements limit the efficient application of RPL. To support IPA's new professional qualification, Incept Labs formed a joint venture with the IPA to develop industry utility for efficient RPL assessments. 

Called Ripltec, the platform enables granular accreditation assessments of other providers' educational programs and detailed RPL assessments for learners. While the IPA is the first system user, it aims to accommodate a diverse range of RPL and accreditation assessments for professional bodies and education providers. Unsurprisingly, designing involves numerous philosophical, methodological, and technological challenges. Providing a learner with RPL is a statement of trust or belief in the capabilities a learner has evidenced relative to a predefined educational requirement. This is usually based on statements of prior experience or evidence of educational attainment. In both cases, the awarding body needs to decide on two main factors: 

1. Scope alignment between prior learning and the educational requirement; 
2. The quality and level of that prior learning compared to the educational requirement. 

Creating a system must accommodate paradigmatically different approaches. Most higher education programs are structured around 'volume of learning' or study load definitions. In these systems, learners achieve learning objectives proportionally (e.g., a  50 percent pass) by completing a prescribed number of study hours. This fundamentally contrasts with competency-based systems (common in technical and vocational educational and training), where learners must demonstrate achievement of all learning outcomes to be judged competent, regardless of the study hours. 

Higher education providers and professional bodies independently determine the proportion and granularity of RPL they provide. For example, the IPA has no limit to the proportion of RPL that can be granted in the GCPA Framework; RPL is granted against individual competencies rather than whole credentials or units. Conversely, higher education and professional bodies often accredit and credit whole credentials or units. Methodologically, this is problematic as units from two different degrees (with similar topics) rarely provide the same scope or coverage. Technically, these systems are input-based rather than learning-based. As a result, learners historically receive less RPL than their learning merits, or education providers and bodies have provided more RPL than is really justified. Accounting for the level of granularity in the calculation is therefore crucial. 

The system leverages AI technologies to support efficiency and speed in processing learning artifacts Initial AI applications relate to document processing, for example: making transcripts machine-readable, or matching similarly worded competency and learning outcome statements. The ultimate judgment of the scope and quality of a learner's experience, however, remains the task of human subject-matter experts.  RPL assessments are statements of belief based on evidence assessment. Managing diverse learning evidence for industry bodies and education providers demands sustained effort and intelligent application of AI technologies.  

Keywords: Recognition of Prior Learning, AI-enabled assessment, competency-based accreditation, micro-credentials, stackable qualifications