The present study is founded in transportation research but is interdisciplinary in that it is drawing on research areas like consumer behaviour, cognitive science, economic psychology and marketing. The results will provide knowledge to 1) researchers, who need a deeper understanding of travel behaviour in order to develop better models, 2) policy makers, who need information on transport demand for planning the infrastructure and for evaluating environmental consequences and 3) marketing analysts, who need information on the whole complex of business and discretionary travel pattems.
Long range travel decisions are by nature more complex than decisions behind e.g. everyday commuting and is largely overlooked in travel behaviour research. The study is limited to discretionary long-range travel. A literature review was made covering behavioural research including economic psychology, consumer behaviour, decision making, cognitive science, long-range discretionary travel and transportation. As a result a conceptual model of the decision process cycle and a set of research propositions were formulated and later tested against collected panel data. An analysis of the often long decision processes require the availability of diverse and complicated data. A telephone sur vey was conducted in the form of a randornly selected panel of 230 respondents that took part in all interviews.
Different way of classifying decision processes were discussed. Especially a model of the degree of decision making effort was studied. The model was used in an attempt to categorise collected decision process data. Behavioural research describes the role of associative knowledge structures, talled schemata and scripts, in the decision process. The concepts of schema and script are used to explain the decision process. The `naturalistic decision making' approach proposes that decisions are not results of choosing among alternatives but consists of a direct selection of a `highly likely' option (based on experience). If the evalu , ation of the first option leads to uncertainty the option is modified or another option is searched for. The process of deciding is thus assumed to be sequential. Applied to the area of transportation this would mean that activity and travel alternatives may exist but they are typically not considered at the time of the first travel idea. Based on collected data a modified framework including eight phases for the decision process cycle was proposed.
The findings were compared with the set of propositions and show support for the propositions. The proposition regarding classification according to the degree of problem solving effort was supported in the differente between regular (> 40% of all trips) and non-regular trips. Frequent findings of a certain concentration of variables decided at the same time indicate support for the proposition saying that a travel idea is typically structured as a schema. The lack of reported destination alternatives support the proposition saying that a travel decisions are typically made without a choice among considered alternatives and that the process is sequential. If the decision maker is uncertain about the firstly considered option, another option is considered (may be to stay at home). The degree of decision toncentration increases with age and trave1 frequency. This supborts the proposition saying that the con centration increases with experience and routinisation of the decision process. The trip cancellation rate decreases with travel frequency. This supports the proposition saying that the cancellation rate depends on the travellers' uncertainty of travel preferences and on previous travel (influencing the routinisation of the decision process). Finally future research is discussed and a set of research topics are suggested.
Institutionen för infrastruktur och samhällsplanering , 1998. , 277 p.