The survey is good for quickly needed information, disparate respondents; respondent possessing suppressed information, for opinions, review of past behaviour, the gathering of future intentions, keeping down costs and respondents are difficult to reach.
Disadvantages of the survey method are: respondents may not reveal the truth, motivation is required to answer, follow-ups and callbacks are generally required.
On the other hand, the observation method is good for collecting data when a natural setting is imperative, gathering information that may not be revealed by the respondent or not known by the subject (for example blood pressure values at one hour intervals), and where behavioural patterns are required.
Disadvantages of the observation method are: sensitive material is generally needed and persons are sometimes not told the complete truth behind the exercise, memory decay can be a major factor of incorrect reporting when devices do not exist for accurate recording, and it is generally more expensive to carry out.
Observation methods are useful when information needs to be gathered from subjects in cases where the subjects themselves cannot provide the information; for example, testing the reaction to certain drugs. Unobtrusive observation methods are useful when gathered information may need a cross-reference. Since observational means are usually more expensive that the communication method, observational methods are easier to justify when the possibility of attaining the said information is generally not possible or extremely difficult with the communication method.