5.4 This would be a longitudinal study. Therefore, the potential of some of the threats to internal validity explained in Section 5.8 is greater simply because they have longer to develop. You would need to make sure that most of these threats were controlled as much as possible. For example, you would need to: • account for the possibility of a major event during the period of the research (widescale redundancies, which might affect employee attitudes) in one of the companies but not the other; • ensure that you used the same data collection devices in both companies; • be aware of the ‘mortality’ problem. Some of the sales assistants will leave. You would be advised to replace them with assistants with similar characteristics, as far as possible.