Researchers often use panel data to study change and stability of social patterns. However, repeated interviewing can affect respondents’ attitudes in a panel study by raising awareness and triggering reflection processes on surveyed topics (cognitive stimulus hypothesis). We investigated change in respondents’ attitudes about abortion and the underlying mechanisms of attitude change across six waves using data from a survey experiment administered within a probability-based and a non-probability panel in Germany. We manipulated the frequency of receiving identical attitude questions on the same issue. We estimated multiple-group and longitudinal structural equation models to differentiate change in the measurement of reported attitudes from “real” attitude change. We show that repeatedly asking about abortion increases the reliability of respondents’ reported abortion attitudes, providing support for the cognitive stimulus hypothesis. Our results also suggest that improved response behavior due to general survey experience further enhances attitude reliability when answering identical attitude questions repeatedly.