Regression Excel Spreadsheet

[MAIorBust]

Hey guys, I hope this is not considered taboo but I am having problems with Microsoft excel 2007. I am trying to build a regression spreadsheet and it is giving me a headache. Does anyone have one the would be willing to e-mail me?

p.s. no this is not an assignment from my mentor. I am trying to teach myself how this works before I take my A.I. statistics class. Thanks.

 

[Jerry Bone Jr]

Hey guys, I hope this is not considered taboo but I am having problems with Microsoft excel 2007. I am trying to build a regression spreadsheet and it is giving me a headache. Does anyone have one the would be willing to e-mail me?

p.s. no this is not an assignment from my mentor. I am trying to teach myself how this works before I take my A.I. statistics class. Thanks.

Regression (linear) for statistics:
x=input, y=output
Exy = Sum of all (x times y)
xbar= mean of x = Sum of all x / n (n = # of x values)
ybar= mean of y = Sum of all y / n (n = # of y values)
Sx = Standard deviation of x = SqRoot of [Sum of all (x - xbar)^2 / n-1]
Sy = Standard deviation of y = SqRoot of [Sum of all (y - ybar)^2 / n-1]
r = correlation = [Exy - n(xbar)(ybar)] / [(n-1)(Sx*Sy)]
b0 = ybar - b1(xbar)
Slope = b1 = r(Sy) / Sx
yhat = predictor
Line of best fit = least-squares line = [yhat = b0 +b1(x)] = (y=ax+b)

 

[Terrel L. Shields]

http://office.microsoft.com/en-us/excel/HA011119631033.aspx

http://office.microsoft.com/en-us/marketplace/EM103681821033.aspx?CategoryID=CE011277931033

Lessons
http://www.mste.uiuc.edu/patel/amar430/intro.html
For $14 you can get a simple one that will write the formula and give you heads up on whether the analysis is valid. I use this one a lot.
http://www.business-spreadsheets.com/

Tons of it on line in shareware, etc.

 

[Metamorphic]

Hey guys, I hope this is not considered taboo but I am having problems with Microsoft excel 2007. I am trying to build a regression spreadsheet and it is giving me a headache. Does anyone have one the would be willing to e-mail me?

p.s. no this is not an assignment from my mentor. I am trying to teach myself how this works before I take my A.I. statistics class. Thanks.

There should be nothing to "build".

If you put your data in two columns, do an xy graph, and apply a trendline. You've done it. Its not complicated enough that you need to build a template for it. If you want, after youve got things tuned up your can do a user defined style and save a little time on the next time.

 

[VegasWayne]

You can do multiple regression analysis from the Data Analysis option under the tools menu. If this option is not available it can be added from the tools add-ins option and adding the Analysis ToolPak. The first column in your worksheet should be sales price and the other columns are for the factors to be analyzed (GLA, pools, fireplaces, etc.) The data must all be in numeric format (use 1 for pool, 0 for no pool). once the columns are set up run the Regression tool from the data analysis option. set the Y value to be the column with the sales price and the x values to the rest of the columns. Run the analysis and read the results.

A template is not necessary as you will use different columns for different types of data (condos don't need a pool column). Get your statistics textbook out and look up what the test results mean to determine if the analysis is meaningful. A good result will give you a formula such as:
SalesPrice = constant + A(GLA) + B(#ofPools) + C(bathrooms) + D(LotSize)

A, B, C, and D can be used to make adjustments in the sales comparison approach. a sample of 30 to 80 sales usually is needed for a good result.

 

[Joe-NC]

I liked David Braun's program. (www.braunappraisal.com) I got the trial version when it was in the beginning stages. It looks like he has improved it quite a bit. It is an Excel regression program for appraisers.

Check out this video.
http://www.youtube.com/watch?v=0c9ofa21Xnc

If you find out more about this program (price & where to get it) please let us know. If anyone knows of anything better please let us know that too.

 

[Terrel L. Shields]

That looks considerably better than Real Stat...

 

[Flygirl 152]

Joe, I have Davids 1004 MC program which I like a lot. I was considering buying his regression analysis software too, so I appreciate your feedback.