THE RELIABILITY ASSESSMENT OF TRANSPORT AND SOCIOLOGICAL RESEARCH AS A QUESTIONNAIRE SURVEY USING STATISTICAL DATA PROCESSING PACKAGES

Improving the quality of passenger services provided by mass transit is one of the most important areas in meeting the needs of society. People's standard of living depends on the quality of urban transportation services. In a market economy, mass transit users determine their satisfaction with the transportation services through various surveys. The surveys are a valuable source of information on the urban passenger transportation operational status and are usually organized in the form of questionnaires. At the same time, the questionnaire development errors can affect the study results and lead to incorrect conclusions. Consequently, the questionnaire reliability is important directly from the developing stage. The questionnaire reliability is the reproducibility of the study results and their resistance to the random factors interference. The questionnaire design inaccuracy should be eliminated by using statistical packages to assess the questionnaire content quality and calculate the Cronbach's alpha. From the coefficient values, it is possible to make a decision on the questionnaire internal consistency, its compliance with the purpose, reliability and application in practical work. Thus, the study used well-known statistical data processing packages such as MS Office Excel, Statgraphics Centurion, IBM SPSS and Statistica. A test sample of 20 respondents was used to check the questionnaire reliability and the respondents' opinion related to the passenger transportation quality components. In general, all data processing packages were suitable for calculating the alpha coefficient. From the all software considered, Statistica was the most functional due to wide range of helping tools while improving and evaluating a survey form, and MS Office Excel was the least functional since most calculations had to be performed manually. In the Statistica environment, simulating the number of questions and the value of alpha allows significant increasing the questionnaire quality. The results of the statistical analysis showed the questionnaire was quite reliable – from 0.763 to 0.771. According to passengers, the most important factors to generate the transportation quality were the travel cost, travel time, and passenger safety, and the least important are the service culture and the vehicle appearance. Further research should be aimed at developing and evaluating more comprehensive surveys on other features of passenger and freight transportation, and logistics.