SPSSX Discussion

(Small) Time series: How to check if the values are significantly different?

Classic

List

Threaded

3 messages Options

Walter Funk

(Small) Time series: How to check if the values are significantly different?

Dear SPSS-fellows,

this seems to be a statistical and a SPSS-question:

I have aggregated data for 12 consecutive months. Some variables are
normally distributed on the aggregated level, others are not. I want to
know for each variable separately if the 12 values do differ
significantly from each other or not.

If I had a cross section, I knew what to do. But since I am not familiar
with time series data my question is: Which SPSS procedure do I have to
use for a little time series like this to test for differences?

Thanks for your help
Walter

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

Marks, Jim

Re: (Small) Time series: How to check if the values are significantly different?

With only 12 months of data, you don't have a true time series (you
can't tell if September is consisently different from April, for e.g)

I would start with a plot of the variables, and test mean differences by
month. This will give you a rough indication.

Your underlying data is based on repeated measurements from the same
sample, so you should use repeated measures if you need more rigorous
testing

--jim

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Walter Funk
Sent: Thursday, January 17, 2008 8:53 AM
To: [hidden email]
Subject: (Small) Time series: How to check if the values are
significantly different?

Dear SPSS-fellows,

this seems to be a statistical and a SPSS-question:

I have aggregated data for 12 consecutive months. Some variables are
normally distributed on the aggregated level, others are not. I want to
know for each variable separately if the 12 values do differ
significantly from each other or not.

If I had a cross section, I knew what to do. But since I am not familiar
with time series data my question is: Which SPSS procedure do I have to
use for a little time series like this to test for differences?

Thanks for your help
Walter

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command SIGNOFF SPSSX-L For a list
of commands to manage subscriptions, send the command INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

ViAnn Beadle

Re: (Small) Time series: How to check if the values are significantly different?

If sequence is important and the OP is looking for some sort of pattern,
then a RUNS test using the mean as a cutpoint could be used.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Marks, Jim
Sent: Thursday, January 17, 2008 8:27 AM
To: [hidden email]
Subject: Re: (Small) Time series: How to check if the values are
significantly different?

With only 12 months of data, you don't have a true time series (you
can't tell if September is consisently different from April, for e.g)

I would start with a plot of the variables, and test mean differences by
month. This will give you a rough indication.

Your underlying data is based on repeated measurements from the same
sample, so you should use repeated measures if you need more rigorous
testing

--jim

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Walter Funk
Sent: Thursday, January 17, 2008 8:53 AM
To: [hidden email]
Subject: (Small) Time series: How to check if the values are
significantly different?

Dear SPSS-fellows,

this seems to be a statistical and a SPSS-question:

I have aggregated data for 12 consecutive months. Some variables are
normally distributed on the aggregated level, others are not. I want to
know for each variable separately if the 12 values do differ
significantly from each other or not.

If I had a cross section, I knew what to do. But since I am not familiar
with time series data my question is: Which SPSS procedure do I have to
use for a little time series like this to test for differences?

Thanks for your help
Walter

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command SIGNOFF SPSSX-L For a list
of commands to manage subscriptions, send the command INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD