This thesis has three parts: development of theory, development of a learning environment that is consistent with theory, and the naturalistic trial of this environment.

First, the focus is primarily on integrating some of the theoretical foundations of neuroscience, cognitive psychology and education. It is argued that, at a fundamental level, they all agree on the basic tenet of human learning: each individual constructs his or her own knowledge; there is no alternative. It is subsequently discussed how this informs teaching, arguing for stronger scaffolding for physics novices, emphasising the importance of prior knowledge with an explanation of why this is essential.

Based on this theoretical understanding, a learning aid for first year physics students was developed: Link Maps. These resemble concept maps and knowledge maps, but were developed specifically for physics based on its characteristic integrated knowledge structure. To test the pedagogical effectiveness of Link Maps, they were implemented in Map Meetings - relatively scaffolding tutorials. First year students in four different physics courses were allocated to either Map Meetings or the standard inquiry based physics tutorials. Link Maps were developed for each course, which differed in levels of assumed prior knowledge. Data on students' tutorial attendance, self-efficacy and examination performance were collected, and qualitative feedback was obtained through interviews and focus groups, short answer questions in questionnaires, tutorial observations by physics education experts and student-staff liaison committee meetings. Triangulation of results revealed that Map Meetings were considered more valuable by the students, both in terms of student attendance and qualitative feedback, had a more positive effect on students' self-efficacy, and resulted in fewer students at risk of failing the course with the lowest assumed prior knowledge – a result that was borderline significant (p =0.056).