Motivational and Behavioral Classes for User Segmentation

[sportstechie]

Hi everyone,

in the course of a research project with a sample size of > 1000, I intend to first group users in unique motivation/motive classes. These will be derived from qualitative user statements which are manually attributed to multiple binary motivation items- min 5 up to 10 items such as Social, Enjoyment, Achievement or Fitness; multiple occurrences are possible. As the data appears at the moment, users would often state single motivations such as „fun“ or two motives such as a combination of „fun“, „achievement“. Anyway, all combinations are possible. What statistical method is useful in that respect to associate an individual user with a specific motivation class?

As a second dimension for the segmentation analysis, I would like to derive behavioral classes based on mixed variables such as intensity of sport consumption/week, number of different sports types consumed as well as binarily coded variables such as paid subscription (yes/no). Again, what statistical method/clustering approach would be feasible for this kind of data?

In the end, I would like to derive dominant user typologies conducting a cross tabulation of motivational and behavioral user classes.

What other resources would you recommend to take into consideration for such a research project. Thanks in advance for your support and sharing any valuable information you may be able to contribute.

[Maguin, Eugene]

In terms of spss procedures, I suspect  you would be most interested 'clustering' the data. Others on the list know a lot more about clustering routines than I do. That said, it may be helpful to them if you would elaborate on frequencies and covariations of those manually attributed motivation items. Sounds like users state their motivations in their own words and you code those words into your preset dichotomous categories. But, it also sounds like many or most users are stating just one motivation which would seem to cause a failure in statistical clustering algorithms.
Gene Maguin


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of sportstechie
Sent: Sunday, May 12, 2013 7:56 AM
To: [hidden email]
Subject: Motivational and Behavioral Classes for User Segmentation

Hi everyone,

in the course of a research project with a sample size of > 1000, I intend to first group users in unique motivation/motive classes. These will be derived from qualitative user statements which are manually attributed to multiple binary motivation items- min 5 up to 10 items such as Social, Enjoyment, Achievement or Fitness; multiple occurrences are possible. As the data appears at the moment, users would often state single motivations such as „fun“ or two motives such as a combination of „fun“, „achievement“.
Anyway, all combinations are possible. What statistical method is useful in that respect to associate an individual user with a specific motivation class?

As a second dimension for the segmentation analysis, I would like to derive behavioral classes based on mixed variables such as intensity of sport consumption/week, number of different sports types consumed as well as binarily coded variables such as paid subscription (yes/no). Again, what statistical method/clustering approach would be feasible for this kind of data?

In the end, I would like to derive dominant user typologies conducting a cross tabulation of motivational and behavioral user classes.

What other resources would you recommend to take into consideration for such a research project. Thanks in advance for your support and sharing any valuable information you may be able to contribute.




--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Motivational-and-Behavioral-Classes-for-User-Segmentation-tp5720191.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
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