A very common issue which appears while choosing a research method is the problem of different scales of measurements. This problem appears when considering a large number of variables and many levels of variables. Another very important problem of choosing the appropriate method of measurement is to create a situation that most closely reflects the actual decision-making process by the respondents. The paper presents the methods of revealed preference measurement representing decomposition, composition and mixed approaches. The article presents also a quite new method: the adaptive discrete choice method, which allows to apply a large number of attributes (variables) and their levels. This method also reflects the actual consumers' decision-making process.