Abstract: Prevalent in everyday non-technical use, probability is highly prone to misconceptions which have been the object of much research. One way to deal with these misconceptions is through anchoring situations, as advocated by Clement (1987) and Fast (1997). This paper uses a similar framework, with two versions of a questionnaire and interviews conducted with college science majors. Misconception-prone situations were given in Version A of the questionnaire, and possible anchors in Version B. Results showed that although erroneous notions existed, anchors could be effectively used to overcome them.