Computer Assisted Telephone Interview (CATI): what one gains, what one loses?

CATI is much better than face-to-face interview with respect to organizational and financial questions. It takes less time, facilitates standardization of asking questions, causes less errors during interview, permits control of their work, and is cheaper. However one may ask, what is validity of data collected by means of CATI? In this article, which is based on empirical data I take up two two important issues: (i) to what extent CATI sample differs from the sample realized in traditional survey, and (ii) whether answers to questions asked in CATI and traditional survey are the same or differ: if so in what way