Description, exploration and explanation are three main purposes of survey approach.
a) Description: Surveys are frequently conducted for the purposes of making discoveries on the distribution of certain traits or attributes. Here the researcher finds out "what the distribution is "and" not why the observed distribution exists".
For example, survey of hill tribes on occurrence of goitre where the survey results describe occurrences, severity, intensity and distribution of goitre in sub-Himalayan region. A number of nursing studies have been conducted for the purpose of describing the phenomena relating to nursing process.
For example, needs of cardiac patients, belief of mother on breast feeding, stress of pre-operative patients, etc.
b) Explanation: Survey explains phenomena. For this purpose, two or more variables are examined to find relationship between phenomena. For example survey of education and their knowledge on prevention of heart diseases may indicate as one of the factors that-influence knowledge.
c) Exploration: Survey can provide search devices, i.e. researcher can explore situations, new organization or a new group where there is very little existing information. Therefore, survey data helps to explore new possibilities, which can be later followed up in a more controlled survey.
Suppose for example, the researcher wishes to examine in depth the women's attitude toward Breast Self Examination (BSE) as a preventive health practice. Such a survey could involve a small sample group using probing questions. The finding of this study would give a better understanding of the problem in order to plan for a refined research project, for example, finding ways of developing
attitude towards regular practice of BSE.
Exploratory study can also be used for analytic purposes. For example, at the time attitudinal data are collected, other background information could also be obtained from a large population of women. As nothing much is known about relationship between individual's characteristics and her attitude, the researcher may explore these relationships through statistical analysis. In this case the sample size should be large.