Need help on this !

[Abhay V]

Dear fellow users ,

   I have to develop a simulation model on consumers online fast food purchase behavior, where 1) convenience , 2) Perceived healthiness , 3) perceived Quality, 4)Shopping experience and 5) Discounts and promotions are independent variables having an impact on perceived value for money as dependent variable. 402 responses were collected on 

5-point Likert scale for structural equation modeling purpose.

How to do simulation in NETLOGO using above variables and data ?

If any other alternative way is possible please suggest , as I want to explore limitless possibility for simulating different scenarios and publish papers on them.

Your guidance and help in this regard will be highly appreciated.

Regds

Abhay 

National Institute of Technology , Bhopal, INDIA.

 

[Dale Frakes]

The hammer I like to hit everything with is the logistic function.  I might treat all those variables as being between 0-1 then you can set up a formula like...

perceived_money_value = 1 - 1/(1 + exp(B_conv * convenience + B_health * healthiness + B_shop * shopping + B_disc * discounts))

The nice thing this setup is that your value will always come back between 0 and 1, regardless of how big one of the inputs might be.

Next you can run a logistic regression on your Likert data to come up with the regression weights (B_conv, B_health, etcl) used in that formula.  Note this a bit iffy in a statistical sense because Likert data isn't really interval data, so this regression violates some assumptions.


Dale

--
You received this message because you are subscribed to the Google Groups "netlogo-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to netlogo-user...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/netlogo-users/80e1eb07-24b3-43bf-a536-8edcf5490075n%40googlegroups.com.

-- 
Dale Frakes
Adjunct Instructor, PhD Student
PSU Systems Science
dfr...@pdx.edu - http://web.pdx.edu/~dfrakes/