The use of leading questions as part of interrogation techniques and the factors associated with the witness suggestibility are well researched in the scientific literature. Nevertheless, literature on suggestibility in intentionally or unintentionally poorly designed social opinion surveys and push polls is scarce. The objective of this paper was to test how easily the public is swayed by leading questions in poorly designed surveys, whether certain group types of respondents in terms of their susceptibility exist, and to investigate the relationship between several demographic and personality characteristics and susceptibility. For this purpose, an online self-administered survey was conducted, containing a block of neutral and several blocks of leading questions about the respondents’ opinions on minority groups. The results indicated that the negative wording in questions was reflected in a more negative opinion on minority groups, while there are differences between respondents in their level of suggestibility. Three groups of respondents exist, i.e., non-suggestible (with no shift in opinion), suggestible (who express a more negative opinion on minorities following the suggestion), and contra-suggestible respondents (who react on the contrary and give a more positive opinion after a negative suggestion). Susceptibility to suggestion is associated with respondents’ age, openness to experience, conscientiousness, and social desirability. The results suggest that contra-suggestibility might counterbalance the error made by the leading questions, at least when more controversial topics are being surveyed.