# SPSS Approach - Unsure

#### bhumieka

##### New Member
I've been working on this assignment with a group of people. And we all have differing approaches for this question:

Use the explanatory variables in the data set to explain the WTP answers (for a description of the variables, see Appendix 1). Start with the following specification. The dependent variable is WTP. Use as regressors six scenario dummies, whether the respondent is a member of an environmental NGO (MembEnvNGO), whether he/she recycles his/her organic waste and/or his her old paper (RecycOrgWaste and RecycPap), whether he/she donates 100 euros or more to environmental charities per year (Donat100), whether he/she lives in a rented house (HousRent), whether his/her house is connected to the municipality’s sewerage system (NoSewer), his/her age (Age), the household’s net monthly income variables (Inc1150, Inc1800, Inc2600, Inc2600h), and whether the respondent is unemployed (Unempl). Interpret the results, and try to find a specification that explains the data better.

WTP is willingness to pay. I think that a Multi Anova needs to be created. Each treatment would be one of the six scenarios. I'm just not sure how this can be done with SPSS. I'm new to using the program. This question is actually from a Environmental Economics course.

I would appreciate any feedback for this question. This is a 10 part question and this is the only part I'm having trouble with.

Thank you!!!

#### CB

##### Super Moderator
Hey there, the suitable analysis depends on what type of variable Willingness to Pay is. If it's a categorical variable a MANOVA is appropriate, but if it can reasonably be considered as a continuous interval-level variable, multiple regression (Analyze > Regression > Linear) is probably the way to go. The fact that the question refers to the predictors as 'regressors' makes me think this is what they're after.

As an extra hint, I might suggest using a stepwise entry method in the regression to identify a more parsiminious solution with only the 'best' (statistically significant) predictors.

Good luck!