Table 2 Key issues in covert observational research

1. Fieldworker safety is paramount; fieldworkers should be aware of when, and how to abandon data collection. Lone worker protocols are also important (see text).

2. Detailed data collection protocols are essential to limit potential bias

3. Training on data collection, preferably involving role playing and visits to the site where observation will take place, is essential

4. Observers should be matched to the environment, for example by age and gender

5. Working in pairs may help fieldworkers feel safer, and less conspicuous, and may limit biases in data collection (though this will increase research costs)

6. Despite all possible precautions, covert observation may be noticed and queried; fieldworkers should therefore have a plausible reason for being where they are