Both techniques are based on the use of telephone as an instrument of interpersonal communication and any of mass media, mostly electronic (television, radio). They have not appeared with purpose of researching public opinion, but firstly for commercial purposes. A discussion in studio the show as a whole represents a 'verbal provocation' regarding the audience from which examinees emanate, where the role of a journalist is to create atmosphere, influence the course, etc. Regardless their specifics over-combination of events and mostly non-scientific motives, they are closer to a written questionnaire than interview regarding their characteristics. A fault is that it is impossible to make a questionnaire with more than one question, 2-3 at the most, and to avoid total confusion. It is clear that the questions giving at least basic data on examinee (gender, age, education etc) will be omitted, and that is the reason why we do not know whom we have questioned. Stratification of obtained results by 'Vote System' technique is possible if only fixed telephony have been used, while it is impossible with OESMS Premium Rate' technique. The fact that examinee pays for answering the question eliminates or at least considerably reduces participation of poorer layers of citizens. On the other hand, organizers motivated firstly by earnings do not limit number of calls from the same telephone number although technical possibilities exist, which interested parties misuse (e.g. political parties) in order to directly influence in favor of wanted option through systematic and organized calling of numbers. OESMS Premium Rate' technique also has the problem that the use of mobile telephones drops with increase of age of citizens in Serbia, and regarding that short messages (SMS) in our country is possible to send only through mobile telephony and considering that Serbia is in the second place in the world regarding age of inhabitants, this reduces representation of research. Finally but the most important -these research techniques disable in advance a choice of representative sample by using statistic method, which truly and absolutely denies the value of scientific generalization of obtained data. This invalidates any intention for their using in scientific research of public opinion, but surely, it does not invalidate their using in order to create and manipulate public opinion.