Research on deep learning was initiated in the 1970's. Winje and Løndal's (2020) mapping review of definitions of deep learning from 1970-2018 shows two conceptualizations of deep learning both based on the cognitive learning perspective:
(1) meaningful learning and
(2) transfer of learning.

Other theories on deep learning (e.g., Warburton, 2003; Tochon, 2010) describe it as holistic, meaningful, and transformative learning, and stress the importance of interdisciplinarity. In the Norwegian context, when The Knowledge Promotion Reform of 2020 was introduced, deep learning was presented with the emphasis on a holistic understanding of a subject and the relationships between different subject areas, as well as applying the knowledge to solve problems and tasks in new situations. Deep learning was presented in opposition to superficial learning, which implies learning of factual information without setting the knowledge in a context (Meld. St. 28, (2015-2016)). The focus of this study is on deep learning in science subjects in vocational education and training in Norway, and the data were collected in a high school in Southern Norway.

The paper addresses two research questions (RQ):
(1) How do teaching staff understand the concept of deep learning in their subject? and
(2) How do teaching staff assess students’ deep learning in their subject?

Focus group interviews with 35 teachers were carried out in January, 2022. The participants were split in six groups, and the interviews were carried out in three rounds. First, the participants started in groups based on their subject (i.e., "home groups"), then they were mixed into cross-disciplinary groups, and finally, they returned to their home groups for the final reflection around the insights they got in the cross-disciplinary groups. The interviews were audio-recorded and transcribed, and the inductive content analysis method was employed to analyse the data. The teachers identified the following key categories to describe what deep learning entails: cross-disciplinarity, both “depth” and “broadness”, applicability, relevance, and connections both between the different subjects and work life. The objective of deep learning is to reach a better understanding through inquiry, critical thinking, and variation. The teachers mentioned the following approaches to assess deep learning: use students' study logs; assess the process continuously in both practical, written, and oral activites; assess the quality of reflection, independence, and understanding; consider students' report-writing and the ability to convey the essentials; follow-up students' progress and development over time; and use rubrics (where it's possible to cross out on various criteria). The participants concluded that it is challenging to formulate specific assessment criteria for deep learning.

References:
[1] Meld. St. 28 (2015–2016). Fag-Fordypning-Forståelse. En fornyelse av Kunnskapsløftet. Kunnskapsdepartementet. https://www.regjeringen.no/no/dokumenter/meld.-st.-28-20152016/id2483955/
[2] Tochon, F. (2010). Deep Education. Journal for Educators, Teachers and Trainers JETT, Vol. 1, s. 1-12.
[3] Warburton, K. (2003). Deep learning and education for sustainability. International Journal of Sustainability in Higher Education. Vol. 4 Nr. 1., s. 44-56.
[4] Winje, Ø., Løndal, K. Bringing deep learning to the surface: A systematic mapping review of 48 years of research in primary and secondary education. NJCIE 2020, Vol. 4(2), 25-41.

Keywords: Deep learning, vocational education, assessment.