A friend's assignment question is "Does class attendance affect GPA?"

She has 3 survey questions:

  • How many courses do you have a week?
  • How many classes did you miss a week?
  • What is your GPA?

She is now told to analyse her data: her assignment question told her to analyse by combining two of the questions, but in her case, I told her she would need to use all three.

I was thinking of doing (percentage of class missed, (class missed/class taken a w/) compared to GPA.

She thinks she can just do classes missed compared to GPA. I told her this wouldn't work because someone could attend 3/3 of their classes, while someone else can attend 3/5 of their classes, and it wouldn't be well represented.

Could someone give some insight on the best method to represent the data to graph?

  • 2
    $\begingroup$ Your being proxy here for a friend doesn't affect the fact that to us this is self-study, and I've tagged accordingly. $\endgroup$
    – Nick Cox
    Nov 28, 2013 at 18:57
  • $\begingroup$ This survey will be unable to answer the question: at best it can answer the question whether reported class attendance is related to reported GPA. There are good reasons to suppose that the reported values will not agree with reality (just as almost everybody reports themselves to be above average in driving ability, sexual prowess, and law-abidingness). It cannot establish any kind of causation, anyway, but only observe a correlation (if one exists). $\endgroup$
    – whuber
    Jun 30, 2014 at 1:55

1 Answer 1


Edit - to address @whuber's comment. If your friend's interview data is not reliable, it is possible to try retrieving this type of information with the school/university administration. Not sure how it works where you are, but in some places teachers/professors record the students' class attendance.

I believe you have a point, i.e. taking 3 classes out of 3 is not equal taking 3 classes out of 5.

% of missed classes would allow to compare students with different number of classes per week. But what other effects (hypothesis) would you expect to affect GPA?

For example, consider the maximum number of classes in a week as being 20.

Student A has 5 classes per week and missed one. Hence, he is 4/5. Student B has 20 classes per week and he did not miss any class (20/20).

Which student is expected to have a greater GPA?

How do you expect number of classes (or number of courses) to relate to GPA? Is there any hypothesis you can formulate from this?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.