Value driven decision forecast model: An application on online flight ticket purchase

Consumption fact is a structure that exists in many parts of our life and has been the subject of many different researches. There are various factors affecting consumption fact and consumers' behavior. In this sense, it is not easy to determine which option will be chosen by the consumers on a certain product or service. One of the most important reasons arises from the decision making process has different stages and products or services are compared with each other on the consumer's mind at these stages. Decision making process has a subjective and also complex structure because it can change from person to person. There are many different methods suggested to determine the preferences of consumers. But the lack of an approach deals with individual personal values -means a value-driven structure- stands out in these methods. It was thought using means-end chain approach which reveals the personal values on this purpose. Means-end chain approach which is based on cognitive theory, is a modelling approach tries to reveal how perceived alternative products or services by consumers and which purposes and values has been taking into consideration on decision making process. MEC purposes to uncover the attributes, benefits and values elements which effect people' choices as a hierarchical chain. In this study, it is discoursed on a decision forecast model which reveals consumers choices on a particular product or service. It was resarched which attributes and consequences reach to the personal values with means-end chain approach on a definite target group while buying online flight tickets. Consumer preferences was researched by evaluating obtained data through Analytic Network Process (ANP). A model which takes into account of personal values on the consumer's choice was generated by forming pairwise comparisons between the attributes -which obtained with means end chain approach- and options at the application area. It was considered that suggested modelling approach may give more efficient outcomes on the high involvement products that requires more cognitive effort. In accordance of this approach, Online flight tickets purchase behaviors were researched as product (service) in this study. The study was based on Turkish airline companies (Atlasjet, Borajet, Onur Air, Pegasus, Sun Express, THY). Means-End Chain approach is the first step of the modelling was applied on the college graduated people between 25-30 ages and 26 people participated in the study. In-depth interviews were performed with these people and the laddering technique that is the base of the MEC theory is used. It was asked to rank the study-included Turkish airline companies respecting online flight tickets purchase and to compare the first two companies in the rankings from the participants. It was aimed to reveal the values and consequences by asking repeatedly 'why' question for each attribute that uncovered after this comparison. Data obtained this way, was first encoded, then the implication matrix that shows direct and indirect relations was formed.A cut-off value was determined and data on the matrix transposed to the hierarchical value map. In the second stage of the modelling, the hierarchical network structure that was a result of the Means end chain approach was integrated to the developed ANP decision model. By doing that, it was provided the effect of values and benefits has been taken into consideration onthe consumer preferences addition to product attributes. ANP gives chance to reflect different types of relations and connections that at the same level on the model. In other words, ANP models the problems which can't be presented as hierarchically, as a network. The change of consumers preference was observed with two model used in the study. 1. Classic decision forecast model 2. Recommended value-driven decision-forecast model It was tried to show personal values should be also taken into the consideration on the consumers preference by the recommended model. In this study, also it has been tried to show up that the classic decision forecast model was insufficient about determining consumer's choice and could give deceptive results. The data obtained from the previous survey was transferred into the ANS model, and the final step was based on the pairwise comparisons between attributes and options with a new survey that was participated by 27 people. These pairwise comparisons were placed in the both of models and it was assumed that two models would give different results from each other and the recommended model would give more accurate results. The results from both models were matced and compared by applying Hotelling's T2 and the paired T tests. The found T2 value showed H0 hypothesis should be rejected, that means two models are different from each other. The result of the T test shows H0 hypothesis can be rejected for THY and Pegasus andBorajet is on the line. For the other three airline companies (Atlasjet, Onur Air ve Sun Express), the results couldn't be rejected. It is reserved that the results of T test could change if the survey would be applied on more people.