Date: Mon, 14 Apr 2008 11:55:35 +0200
Reply-To: Luca Meyer <lucameyer@vodafone.it>
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: Luca Meyer <lucameyer@vodafone.it>
Subject: Statistical question: How to do time series analysis?
Content-Type: text/plain; charset="us-ascii"
Hello,
This is not a strictly SPSS related question but I still need some advise.
I have a dataset that looks like:
DATA LIST FREE /MONTH V1 V2 V3 V4 VA VB VC VD VE VF.
BEGIN DATA
200701 8,27 8,44 7,97 , 0,51 0,16 0,77 0,67 0,48 0,82
200702 8,39 8,42 7,95 , 0,4 0,4 0,82 0,67 0,76 0,44
200703 8,23 8,43 7,9 8,14 0,4 0,4 0,83 0,74 0,6 0,82
200704 8,54 8,55 8,27 8,38 0,69 0,28 0,87 0,66 0,64 0,88
200705 8,48 8,43 8,3 8,3 0,74 0,28 0,83 0,64 0,61 1
200706 8,44 8,4 8,19 8,31 0,37 0,4 0,88 0,55 0,54 0,94
200707 8,27 8,18 8,13 8,03 0,6 0,2 0,9 0,58 0,54 0,7
200708 8,38 8,33 8,15 8,22 0,43 0,36 0,98 0,65 0,84 0,98
200709 8,42 8,42 8,27 8,25 0,6 0,12 0,86 0,63 0,46 0,98
200710 8,69 8,71 8,31 8,55 0,69 0,36 0,89 0,68 0,54 0,58
200711 8,78 8,8 8,44 8,73 0,51 0,48 0,96 0,74 0,57 0,96
200712 8,39 8,42 8 8,23 0,34 0,28 0,85 0,59 0,51 0,84
END DATA.
EXE.
While the V1 to V3 are answers to a questionnaire which uses different
samples from month to month - that is independent observations, the VA to VF
are indexes of quality of a service and therefore I do not think that can be
seen as independent.
I would like to correlate the variables V1, V2 and V3 to the other variables
and would like to make some statements on the quality of such a correlation.
Is there some procedure you can advise I could use? I am not familiar with
time series analysis and OLS regression analysis doesn't seem right here as
(a) observations are correlated and (b) data point are scarse.
Thank you,
Luca
Mr. Luca MEYER
Market research, data analysis & more
www.lucameyer.com <http://www.lucameyer.com/> - Tel: +39.339.495.00.21
No virus found in this outgoing message.
Checked by AVG.
Version: 7.5.519 / Virus Database: 269.22.13/1376 - Release Date: 13/04/2008
13.45
=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@LISTSERV.UGA.EDU (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
| 
