Article Preview
TopIntroduction
One of the main ideas of M-TEL (Mobile Technology Enhanced Learning) is that it can facilitate in situ learning. In this sense, M-TEL research and practice builds on situated perspectives on learning (Brown, Collins, & Duguid, 1989). The situated cognitive perspective is not exclusive to M-TEL, or even TEL, it is also a common viewpoint within HCI. But the study of M-TEL faces certain specific obstacles, such as the mobility of the learner and the mobility of the data collector (much like HCI for mobile artefacts).
In order to analyse M-TEL activities qualitatively it is common to collect data by audio and video recordings. As the learners in M-TEL are active, mobile, and often work in groups, there is a need to collect data from a number of learners. The collected data becomes vast, is time consuming to analyse thoroughly, and will include duplicates since several recorders are used. Because of the amount of data, studies can become fragmented and analysis and reporting is often done using only parts of the recorded data. There is also the risk of missing important data in the analysis. The question arises; how can we minimize the amount of collected data? Also, is it possible to give an estimation of the value of using a particular collection technique?
The aim of this article is to investigate how many recording devices are needed to record a sufficient amount of easily audible conversation, for later qualitative data analysis, for a group of learners, when for instance studying learning and collaboration in a real life setting in an outdoor context. The analysis reported is part of a larger M-TEL field study, MULLE (Math edUcation and pLayful LEarning), where two issues were studied where the speech of the learners were important for the analysis. The first was a study of interaction design through learners’ focus on devices via interaction analysis and the task model (Eliasson, Nouri, Ramberg, & Cerratto Pargman, 2010). The second was how different pedagogies may support collaboration and learning and was studied through conversation analysis (Nouri, Eliasson, Ramberg, & Rutz, 2010). The groups of learners used a designed mobile learning tool to measure areas in a real setting (in a field), where the learners would move as a group, but would also split up in order to solve tasks. The data was collected using video, to record how they moved, and digital audio recording devices with corded microphones attached to each learner (henceforth referred to as audio recorders), to record what they said. In the case of our short study (a little more than 1 hour with two groups of three learners in the field), we collected almost seven hours of data, which took several hours per collected hour to analyse. Gathering data through audio and video recorders is a common way to collect data, posing common problems for (at least) short term M-TEL studies. This article only discusses the gathered recordings from one of the two groups with three learners in each group.
When planning the study, the hypothesis was that it would be enough to attach an audio recorder to the learner who held the main M-TEL device and record non-verbal group interaction on video in order to minimize collected data. M-TEL is often a collaborative activity, requiring occasional grouping for discussion, and we thought that all interaction of interest, vocalised recordable interaction, would be recorded on either the single audio recorder attached to the learner in charge of the main mobile device, or be recorded on the close-up video camera. We wanted to know if it would be enough to attach a recorder to just one learner, but to investigate this; we attached audio recorders to all of them.