Time Series Question

[dream chaser]

Hi Guys,

I have got a couple of questions regarding time series questions.

Q1. As shown in the attachment 'time series', when the data is shown with the actual years(i.e 2004, 2005, etc..) not 1, 2, 3, then on CAS, we should put the x-axis values as the actual years as well right and when you form a regression line for the data, you should sub the actual year into the equation, not 1,2,3 right?

How can you actually tell when you have to use 1,2,3 as the years or the actual years?

Q2. Also, with the other attachment 'predicting', don’t we have to seasonalize the data afterwards? Or does that depend on the information given. Or will the answer of Q7 be such that that will be the seasonalized value? How can you tell when the answer you get from the regression line is the deseasonalized data or seasonalized data?

The actual questions are shown in the attachment of this post

All replies will be very much appreciated. Thanks 

[yourfriendlyneighbourhoodghost]

Hi Guys,

I have got a couple of questions regarding time series questions.

Q1. As shown in the attachment 'time series', when the data is shown with the actual years(i.e 2004, 2005, etc..) not 1, 2, 3, then on CAS, we should put the x-axis values as the actual years as well right and when you form a regression line for the data, you should sub the actual year into the equation, not 1,2,3 right?

How can you actually tell when you have to use 1,2,3 as the years or the actual years?

Q2. Also, with the other attachment 'predicting', don’t we have to seasonalize the data afterwards? Or does that depend on the information given. Or will the answer of Q7 be such that that will be the seasonalized value? How can you tell when the answer you get from the regression line is the deseasonalized data or seasonalized data?

The actual questions are shown in the attachment of this post

All replies will be very much appreciated. Thanks 

Hey,

Almost always, you input into your CAS the numbers associated with the year, not the actual year itself, or else it can take a while. Although, if the question does state you must input the actual years, then I would. But I highly doubt, they will do this.


For your second, question, I don't think you need to do anything else than what the questions asks. I would use the values they give you, and since they haven't stated to seasonlise anything, then I wouldn't.

Have you checked the answers?

[AISHAB]

Hey,

Q1) i think you add the values of 2014, 2015, 2016 and 2017, then divide that by 4
Then add the values of 2015,2016,2017 and 2018 and divide it by 4
Add the two valurs and divde the answer by 2 to get a centred mean.

Q2) The question doesnt ask us to deseasonalize it so just calculate 6.182 - 0.061 x 3
which should give 5.999

Could someone please confirm if this is right

[yourfriendlyneighbourhoodghost]

Hey,

Q1) i think you add the values of 2014, 2015, 2016 and 2017, then divide that by 4
Then add the values of 2015,2016,2017 and 2018 and divide it by 4
Add the two valurs and divde the answer by 2 to get a centred mean.

Q2) The question doesnt ask us to deseasonalize it so just calculate 6.182 - 0.061 x 3
which should give 5.999

Could someone please confirm if this is right

That is exactly what I would do (: