I’m sure that people have studied different schemes and made recommendations. I’m not familiar with that work. However, suppose you summed purchases and divided
each person’s purchase total by the sample total to get the fraction of the total that each person’s purchases represents. My idea is that in given set of people the high value people are the ones whose purchase totals are a higher percentage of the sample
total.
Gene Maguin
From: SPSSX(r) Discussion <[hidden email]>
On Behalf Of Mark Webb
Sent: Wednesday, January 9, 2019 1:31 AM
To: [hidden email]
Subject: Weighting respondents on respondent value - some pointers please
I can rank my sample by monthly gas / petrol purchase. Money value.
A respondent with a high monthly spend is more "valuable" than one with a lower spend.
I want to weight my data ito this value thinking. i.e. one person may be twice as 'valuable' than another.
I'm used to weighting data to correct for demographic over/under representation but am struggling to get my mind around this value weighting.
I have tried Z scores with little luck.
Is there a standard way of weighting to a respondents perceived value?
Any pointers would be useful.
Regards
Mark Webb
Line +27 (21) 786 1124
Cell +27 (72) 199 1000 [Poor Reception]
Fax to email +27 (86) 5513075
Skype tomarkwebb
Email [hidden email]
===================== 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
Locked
