Forums › Ask ACCA Tutor Forums › Ask the Tutor ACCA PM Exams › Sales forecast-time series analysis-Exam Jun'12 exercise 3

- This topic has 1 reply, 2 voices, and was last updated 7 years ago by John Moffat.

- AuthorPosts
- November 22, 2015 at 7:30 pm #284640
Good evening,

in the solutions of the exam Jun’12 Ex.3 it’s stated that “candidates could also have used linear regression”.

If I put on a graph the sales trend over time I see that the relationship is not linear, so regression could not be used.

In any case, even if it was a linear relationship and I wanted to use the math of the regression equation, I would only have the variable y=sales. I wonder what the “x” variable would be in that case. Time in terms of Q1,Q2,Q3,Q4 cannot be the “x” variable as it is not a math reprensentable variable.

If I had sales and advertising cost over time it would be easy, but with sales only, I don’t understand how to use regression.

Thanks a lot if you can shed a bit of light on it and show the solution using regression.

Best regards

LuiginaNovember 23, 2015 at 7:14 am #284668I don’t know what you mean by “not a math representative variable” because you can certainly use the numbers 1, 2, 3, 4, 5 etc as the x variable.

However as you will have seen from our Lecture Notes, the examiner has said that there will not longer be calculations involving time series or regression in the F5 exam.

You are expected to have heard of them (they are assumed knowledge from Paper F2) but you will not get calculation questions. - AuthorPosts

- You must be logged in to reply to this topic.