Analysis of tradesmen perceptions of foreign tourists visiting Alanya using econometric methods

Foreign currency input from foreign tourists is an important economic source of Turkey. This sector, which is characterized as a chimney free industry, can be seen as the export of the country's goods within the country. The fact that foreigners leave the country satisfied at the end of the holiday, and their revisit intention is positive will increase the tourism input in the medium and long term. Although there are many factors affecting tourist satisfaction in the literature, the shopping experiences of tourists and their relations with local tradesmen have been on the agenda recently. In this study, the effects of foreign visitors' shopping experiences and local tradesmen and guides' relations with tourists on tourists' revisit intentions were analyzed with the help of Multinomial Logit (MLOGIT) and Generalized Ordered Logit (GOLOGIT) models. The findings were compared with the results of the Poisson Regression model, which was derived from this study. According to the results of the analysis, the attitude of tourists in the face of the positive trades behavior and good product quality significantly increased the revisit intention; attitudes taken by tourists in the face of the negative behavior of local tradesmen and guides were found to significantly reduce the revisit intention. The main purpose of the study in terms of econometrics is to compare the Multinomial and Ordered logit models with each other on the parameters, conformity of the procedure, ease of application and the provision of assumptions. The dependent variables of these models are located in the same data set, one of which is nominal and the other is ordinal scale structure and measures almost the same information. Before starting regression analyses, the data were subjected to factor analysis and stable factor loads were obtained that resulted very closely together with both vertical and oblique rotation methods. During the application, the Generalized Ordered Logit model was introduced because the Ordered Logit model did not provide the 'parallel regression assumption'. Although the number of statistically significant parameters is the same in both models, GOLOGIT stands out in terms of providing detailed examination for each level of each qualitative independent variables and in terms of making fewer model predictions than MLOGIT. Parameter estimates obtained in both models are in conforming to the results of the study 'Poisson Regression Analysis with Tourism Data: Analysis of The Effects of Foreign Visitors' Local Tradership and Shopping Perceptions on The Frequency of Visit', derived from this study.