After explicit instruction on how to give possible bidirectional (two-way) causality explanations for a correlation, 240 students from eight sections of social psychology and research methods courses wrote "reverse causality" explanations on various test questions, creating a total of 882 answers. Averaging across multiple graded attempts over four semesters, only 45% of the explanations were correct. The highest average for a single question was 67% correct. A regression predicting correct answers showed that students with higher grade point averages (GPAs) were more likely to answer correctly. Controlling for GPA and topic of the correlation, later attempts were significantly more likely to be answered correctly than earlier attempts.