department, the head of project management for vehicle department, the vehicle development function manager, the head of human resources, the chief engineer of ‘packaging’ (the ‘packaging’ task will be explained in the following section), one ‘packaging’ team leader, three ‘packaging’ team members and the head of ‘project SUV’; (3) one researcher spent 10 months working at the research centre; during this time, he regularly took part in meetings, participated in some of the tasks, provided support, and so on. In order to increase the consistency of our data, we decided to focus our interviews and direct observations on one project. For confidentiality, we cannot use the real acronym of the project, so we have named it ‘project SUV’. The project was the most important one the research centre was carrying out during the observation. The method was designed to observe the same units of analysis from different angles, confirming the reliability of the data when they were consistent and inducing a new round of interviews when inconsistent. The comparison of quantitative and qualitative data, and the fact that managers from different departments and functions were interviewed using the same unit of analysis, made triangulation possible and extremely effective. Due to the intrinsic limitations of case study research (Miles and Huberman, 1994), the sample could be partially biased by the fact that only one project has been observed, which could represent a limit to the generalizability of the findings despite the fact managers interviewed confirmed that the project chosen was highly representative of their usual projects.
3.2. The task selected for observation: packaging
To analyze organizational routines, first, we must select a task in order to analyze how it is
accomplished by the organization. Following from the definition of a routine, it should be a task
that is recurrent, repeatedly accomplished in ‘the same’ (or very similar) way. For purposes of
observation, it would also be helpful if the frequency would be high rather than low so that numerous observations can be collected in a reasonable time frame. Furthermore, the task should be closely linked to the organization’s objectives in order to avoid sampling outliers such as activities that take place ‘at work’ but are not primarily concerned with accomplishing the organization’s objectives. Next, we must select the level of granularity at which to describe how tasks are accomplished, and this choice can present a problem. If the task chosen is too ‘large’ (macro, high-level), then