When it comes to "Correlation and causation", the "third variable problem" is that a correlation between two variables (e.g. "attending class" and "good grades") may be caused by a third variable (e.g. "students who are highly conscientious or hard-working might always attend every class and also study a lot for exams").
Therefore, conscientiousness independently causes attendance at class AND good grades, without attendance at class necessarily having any direct causal effect on grades.
Select one of the following: