Notice: the WebSM website has not been updated since the beginning of 2018.

Web Survey Bibliography

Title Reason analysis: an ambitious alternative for mixed‐mode survey design
Year 2009
Access date 22.11.2009

Reason analysis is a long sidelined method of data collection and analysis. I would like to draw attention to the applicability of the method today for addressing current complex social research problems. In my paper I will recount the roots and principles of the method and how it is applied. In conclusion I will present some examples of it in use and some of the problems connected with its practical application in research.

Reason analysis is an analysis of the individual reasons and motives behind the process of decision

‐making relating to various questions. It can be applied in social research generally, in public opinion research, and in market research. The method and principles of reason analysis were first expounded by Paul Felix Lazarsfeld in his article The Art of Asking “Why?” published in 1935. Over the next thirty years it was employed several times as part of the “Princeton Radio Project” and in research conducted by the Columbia Sociology School. Charles Kadushin wrote the entry on “reason analysis” in the International Encyclopedia of Social Sciences. The method then fell into obscurity for the next forty years.

The principles and approach to reason analysis: The method derives from an attempt to address the variance between a question posed generally and the individual ways in which people answer “why” questions. In the first step we ask respondents a simple question about what reason led them to make the decision they did. This decision may be a serious life decision, such as deciding to marry their chosen partner, moving into a new home, visiting a psychiatrist, or, conversely, smaller individual decisions, such as buying a new car, or a particular cosmetic product. The fact is that people usually give just one reason for their decision. Data obtained this way cannot be processed by simply categorising respondents according to what type of response they gave to this one introductory question. Every respondent naturally has all sorts of reasons for their decision. What is important is that we can hear various types of responses to the introductory question from the same mouth. The reason for buying a new Citroen C3 in light blue may be a personal preference for this brand and the person’s old car has just broken down and the cost of repairs is rising. An important motive may be that the opportunity arises to write an old car off for scrap while there is also a sale on a particular new model. The person’s partner may have heard something complimentary said about a particular model of car. The brand, model, and its accessories may be recommended by the showroom salesperson while a TV commercial aired last Sunday also encouraged a person to buy. A combination of any or all these reasons is the only correct and full answer to the original question. The “reason analysis” method looks for and proposes a concrete “tree” of questions, an “accounting scheme”. The next step is preparation of the structured interview in which variant questions are posed about the “quality of a product”, both the product replaced and the newly bought one, how the product was evaluated in advertising, by the seller, by friends and acquaintances, and the “circumstances of the situation”. In the end, it is necessary to group the responses into classes and types according to which responses are most alike and where significant differences are between them. The method did not catch on mainly owing the demands it puts on researchers. In most survey type research it was replaced by the factorial approach, which examines the effect of individual causes (influential factors) jointly for an entire sample of individuals or other units.

The question for today is whether this method has a place in current social research. The paper presents examples of research situations which directly require individualised models of data collection and whose objective is to reveal and analyse further typologies of actors making certain decisions in a given situation and under the influence of individual factors.

Access/Direct link

Conference homepage (abstract)

Year of publication2009
Bibliographic typeConferences, workshops, tutorials, presentations