In the research literature on the cultivation hypothesis, there are two types of measures: first order and second order. First-order measures require respondents to provide estimates of occurrences. Second-order measures assess respondents’ attitudes. This study seeks to determine whether this measurement distinction results in operationalizations that are independent or related to one another. It was found that the two are related to a weak degree at a general level, but stronger relationships are found on a contingent analysis. Subgroups defined by relevant situational variables and certain general variables (such as IQ and especially age) reveal significant differential patterns of relationships. Also, evidence was found of an asymmetric pattern on several topics where first-order estimates usually form a basis for second-order beliefs.