# What Analysis method should I use?

#### gregjjs

##### New Member
I am taking classes in SPSS but I need help for my project. I am doing a project to see if cell phone towers affect the value of surrounding properties (over the last 20 years). I have collected the following. Locations of cell phone towers in many communities in Maine. Properties surrounding the town and the previous town valuation of the lands. I also have the latest town valuations (which is after the tower construction). Since towns do valuations The I have data on the properties between when the tower was built and the latest town valuation.

What spss analysis should I use and what should be the dependent(s) and independent(s) data sets. Thanks for your help.:shakehead

#### Karabiner

##### TS Contributor
It is slighlty confusing that you put this question in
a software (SPSS-) specific forum. I would suppose
that you can find the procedures which you want to
use in the software manual or in SPSS-specific tutorials.
Looks like an ANOVA or regression approach.

With kind regards

K.

#### bruin

##### Member
I am taking classes in SPSS but I need help for my project. I am doing a project to see if cell phone towers affect the value of surrounding properties (over the last 20 years). I have collected the following. Locations of cell phone towers in many communities in Maine. Properties surrounding the town and the previous town valuation of the lands. I also have the latest town valuations (which is after the tower construction). Since towns do valuations The I have data on the properties between when the tower was built and the latest town valuation.

What spss analysis should I use and what should be the dependent(s) and independent(s) data sets. Thanks for your help.:shakehead
First of all, unless you are randomly assigning certain communities to have more cell phone towers and then building the towers yourself, you DON'T have an independent variable, and you CAN'T conclude that cell phone towers affect the value of surrounding properties (regardless of the outcome of your statistical analysis). It is a correlational design, and you cannot conclude cause and effect.

That said, just do a simple pearson correlation: