Data Collection in the full study

Having completed two experimental phases, the actual data collection method drew on the lessons learnt. Full agreement was sought from the students and, in particular, it was ensured that the students were fully content to be recorded. Their interaction was limited to a single room and the other student PBL teams had little interaction, removing the problem of having to concentrate on only part of the group or to deal with the intervention of other students. Initially two PBL teams were recorded working on the same long PBL problem/task from different classes and with two different facilitators. One team was recorded for three weeks but then their attendance dropped and the group fragmented. After this, the data collection concentrated on just one group (Kevin’s team) and they were observed and recorded and used as the main data source for this study.

For the students this was their second semester at university and their second experience of PBL as a learning environment. Their task was to build a model bridge using supplied materials (paddle-pop sticks and glue) that would meet a preset stress test. The group consisted of five students, plus the inputs (structured and ad-hoc) of a tutor. Five sessions were video-recorded, subsequently referred to as Kevin Team 1, 2, 3, 4 and 5, of varying lengths. In theory each class should have been of the same length (i.e. an hour) but after the first class the facilitator left the room after a varying period of time. The students should then have carried on with self-directed learning but in practice they tended to leave the classroom shortly after the facilitator left.

However, unlike the first instance reported in 3.2.2.2, this particular class involved a substantive problem (build a model bridge) and thus offered the scope for more complex problem solving than had been the case with the theodolite PBL. The design task involved a combination of theoretical knowledge with a practical objective so allowed consideration of how meaning making might alter as the task progressed. Equally, the learning was substantive enough to mean that the facilitator would have to take an active role, either using scaffolding or with the direct input of key information. In addition, as discussed in the introduction, engineering design fits with the researcher’s interests in problem solving in science and engineering as it is based directly on resolving an ill-structured, but bounded, problem. Finally, the students had already had experience of one PBL class (in their first semester), so it was expected that they understood the implications and would know they were expected to adopt an active learning style.

Table 3-1: Outline of PBL sessions*

Date Duration (mins) Focus Participants (No.) Reference Appendix 16/09/2010 64 Student work on theoretical basis 5 Kevin Team 1 1 23/09/2010 45 Student work on background information for task 5 Kevin Team 2 2 07/10/2010 33 Preparation work, presentation of information on past designs by tutor

5 Kevin Team 3 3

11/10/2010 42 Discussion of how to build

model bridge

2 Kevin Team 4 4

21/10/2010 35 Presentation of initial

design

5 Kevin Team 5 5

*All sessions were recorded in the same classroom.

The class studied was designed to introduce concepts from mathematics and physics in a practical context using low cost materials that are easy to use.

Specifically, the students were studying solid mechanics and the lecturer’s philosophy for the approach is that students ‘experience and practice’ the approach of designing practical solutions for real problems. The design philosophy is summarised by the lecturer as:

things get difficult. It is the author’s aim through model making (and breaking) to create experiences that will deliver this sort of benefit to students as early as possible, and motivate them to finally succeed in their chosen Engineering degree”

To achieve this, the students were provided with ‘paddle pop’ sticks each of which was 100mm by 10mm by 2mm, originally designed as part of an iced confection. The students were given the criteria as (all information extracted from the class handbook):

“Objective:

To design, draw, construct and test a model truss bridge to achieve a maximum structural ‘efficiency’ over a 500mm span supporting a single central load within the specifications, rules and regulations as provided. A ‘professional’ report must accompany the model. …. Specifications and testing:

1. Model bridges are to be of the under-slung truss type. Any

shape/member configuration may be used provided the resultant model fits within the stated dimensional parameters.

2. Testing of models under a central load is in two parts whereby models that pass a

simple load test of supporting a freely suspended 30kgm mass at mid- span go into the

pool of ‘proofed models’ for final.

3. Final testing utilizes a calibrated Instron TM testing machine via a loading ‘head’

which is capable of limited rotation in the plane perpendicular to the model.

4. The load carried by each model at ‘collapse’ will be recorded and divided by the mass of the model to achieve a strength to weight efficiency ratio.

5. Collapse is defined when either the model cannot support any further increase in load or the deflection at the load point exceeds 30mm.

Material limitations:

• 275 standard, untreated, pine ‘paddle pop’ sticks as provided and a 250ml bottle of PVA adhesive as provided. No other materials may be used.

• No treatment may be applied to the sticks (Sanding may improve adhesive bond.)

• Sticks may be cut/glued in any shape/way to form the truss members and connections.

• The models must span a clear distance of 500mm with

approximately 20 to 25mm end support length. This means the total length of a model must be between 540 to 550mm exactly

• No part of a model may project up above the level of the support frame by 20mm nor down below the level of the testing frame top by more than 110mm.

• The width of a model must not exceed 110mm at any point.

• Contact between the model and the testing frame to be only at the support tops.”

The final assessment was to test the resulting model to the point where the bridge buckled or broke with this being seen as an important way in which the

students could learn about the strengths and weaknesses of a variety of model configurations and design strategies.