Criteria for the Evaluation of Observational studies:
Evaluation of observational method performed by human beings should be done in terms of calibration reliability and validity.
1. Calibration is the capability of the observer to observe the situation under study. More training and expertise is required for an observer who is expected to perform with fine degree of calibration than those operating at a less detailed level.
2. Reliability is evaluated by the level of consistency in observation over time. Reliability will tend to be higher for simple counting types of observations and is lower when interpretation is also required. Overloading the capacity of the observer is likely to produce a non random loss of information with the more difficult aspects of the task being first to suffer.
3. Validity of observation refers to the question of whether rather observer observers and records all relevant present in the situation or only some biased portions of these aspects are recorded.