$ per Square Foot Significance

[Charlotte Dixon]

This is classic example of the old saying "a little knowledge is a dangerous thing".

Oh how true this is! Reminds me of the Relators continuing ed class I attended in April. The class was taught by a well-intended, knowledgable, respected local appraiser. By the end of the evening the Realtors were shouting out:
"WHATDAYA GIVE A GARAGE, WATDAYA GIVE A POOL, WHATDAYA GIVE A PORCH"?

With notepads and pens in hand, they feverously took notes. What a nightmare that night was! Since then, I've noticed that some Realtors do CMA's on a URAR form and they use adjustments that they've seen on appraisal reports.

Back to the subject: I can't imagine how you could ever explain an appraisal and the adjustments you noted on the report to a Realtor. You don't even have to talk with that person and owe them no explanation. Tell them to take some appraisal courses and pay for them like you did.

 

[jtrotta]

Austin

Response; more than you can afford; if 80,000 appraisers are doing it all wrong and your the only one doing it right, Cape Kennedy has a shuttle waiting for you. Your methods appear to FIT your area, where I'm located it would be difficult to locate 30 sales of similar Style etc. in the many small towns we have; the large towns it might be possible, but I don't have that kind of time to research it right now.

You state that if I can't find 30 sales, then you can't report a "most probable price" - I beg to differ with you in that, if you can use the information via the National Association of Realtors, you can defintely support a "most propable price theory"; And if 80,000+/- appraisers are creating fraudulent and misleading appraisals, the Banking Industry would have an excellent opportunity to be even wealthier than they are now- but I don't think that your statement would hold water in court :!:

Your comments on "emotion" are as limited as your thoughts on this issue, you know better and for all the knowledge you appear to have, how can you stoop to that kind of tactic.

As far as your; FNMA cannot dictate appraisal theory; your right they can't, but of all the many times I've been to court, it certainly opens the Judges eyes when you can use it for support in your demonstration.

I truly enjoy math, I am not intending to dispute your theory, it just may not apply to my area, based in part on your requirements.

8)

 

[TEL2002]

Multiple Regression Anaylsis thread for Appraisers Forum.

Sounds like a great idea. Count my partner and I in on it.

 

[Farm Gal]

Unless there is an overwhelming response, I am going to suggest that we move such discussion -particularly as specific to certain programs and their cost benefit ratios to the tech forum....
It seems the most appropriate under existing headers...

Will post there.

 

[slacker]

slacker: Stick to your voodoo methods or astrology or whatever you use.

<span style='color:blue'>Austin,

Relax and take a deep breath. I think your Chi-Squared Distribution is getting wound a little too tight.

I'm actually very interested in the whole linear approach but like everything else I think it has it's place. Any Stats professor would tell you the same thing. The first statistics class I ever had started with the professor proving mathematically with a regression model that trailer homes cause tornadoes. His point was that regression is a great tool; just beware of how the numbers can be manipulated.

I do not doubt your level of knowledge on the subject. I just know that I spend too much time as it is explaining adjustments in addendum so they can be easily understood. I can only imagine trying to explain the "Correlation Coefficients" between my data sets. I can hear the UW now. "Hey Slacker, that's real interesting. Now, can you give me two more 4 bedrooms that can support value?"

And really, Voodoo............Astrology?? You shouldn't be plagiarizing Terrel like that. At least when Terrel first accused me of using those techniques he had a point.</span>

 

[Austin]

Slacker wrote:

“The first statistics class I ever had started with the professor proving mathematically with a regression model that trailer homes cause tornadoes. His point was that regression is a great tool; just beware of how the numbers can be manipulated.”

Reply: I am aware of it slacker, that is why I don’t make time adjustments based on historical trend lines because I know extrapolating can get you into big trouble. It is also why I don’t use the so-called accepted appraisal practice of using the marketing grid programmed in to all FNMA form programs. The reasons being: The results are equivalent to saying that cost equals value; the sequence of adjustments is completely wrong; the method of making a size adjustment is not market supported; adjustments like quality of construction, design & appeal, site view, etc., are covariant variables that cannot be separated out and treated separately; and you can’t solve problems involving covariant variables with an algorithm that is not capable of solving problems of that nature. The end result is a violation of standards 1-1B & 1-1C resulting in an appraisal that has compounded errors that adds up to a fraudulent and misleading appraisal report. Wonder what your professor would say if he knew you had been doing all of the above and them some, and charging people for the thinking you did to come with the results you reported?

 

[slacker]

The end result is a violation of standards 1-1B & 1-1C resulting in an appraisal that has compounded errors that adds up to a fraudulent and misleading appraisal report.

Dude, I'm with ya! You keep coming at me like I'm disagreeing with you. I think the regression idea is a good one. Just give me one that is easy to use and more importantly easy for the und user to understand.

For now though, I'll take my 3 to 6 six comps and come up with a range of value after adjustments are applied and make my value opinion from there. I don't think I would be in violation for that?

Regression or no regression, the only method that the typical buyer is worried about is: "What did the guys house down the street sell for, and was it similar to the one I'm buying? If it's bigger, I know I'll be paying more for it."" That's market value. I don't know too many buyers that are examining the fluctuations in price per square foot as the size of the property goes up or down.

 

[Mike Garrett RAA]

Thanks for the headache, youse guys!. Always reminds me of the blind men trying to describe an elephant.

1. We all do multiple regression...some of us just don't go to the trouble of creating fancy graphs and charts when doing it.

2. You will never convince a REALTOR® that cost per square foot isn't a valid unit of measurement.

3. You will never convice most appraisers MLR is the only way to go.

4. A valid sample is a valid sample. It doesn't have to be 30, 70, or 1,000. It is a sample of YOUR market data and that could be small or large.

I like to show Marshall and Swift to REALTORS®. It leads to the conclusion there are literally hundreds of different numbers based on size, quality of construction, time, and location. Once they grasp that principle it is easier to lead them to "there is no one number".

More and more I find the principle of contribution by percentage (%) is the best in making adjustments. As an example, what is the adjustment for a two car vs three car garage? Is it the same in a $100,000 home, a $250,000 home or a $600,000 home? The same applies to square footage adjustments. We use a different number as both size and price change.

As usual, it appears that many of you feel the REALTOR® is your enemy. We MUST learn to identify our "publics" of which the real estate agent is one. Communicate, Educate, Appreciate.

 

[slacker]

Here's some real significant square foot data.

Had the area between the two uprights of the goal post been about 200 sft bigger, the Falcons would have beaten the Bears this weekend.

Although I'm sure it can be shown statistically that the Bears actually lost.

Where's the Super Bowl This Year? I need to start making plans.

BEARS!!!!!!!!!!!!

 

[Austin]

Slaker: I hate to give you guys the bad news, but the entire appraisal process including cost, sales comparison, and income approach, are collectively a crude form of non-mathematical regression analysis. Each of the three approaches is a form of regression analysis all based on intuition adjustments, not math, and using a totally incorrect algorithm. I say this because it makes me laugh every time some one makes snide remarks about my MRA methods and warns me about the possible abuses. The accepted appraisal methods are showcases of gross abuses of regression methods so this is a case of the pot calling the kettle black.

Mike Garrett in his post just gave an example of multicollinearity in action with his example of percentage adjustments. I don’t know if you know it or not Mike, but what you are doing is dealing with multicollinear variables correctly. Congratulations! Total sale price per square foot is a treasure trove of information because everything affecting value is encapsulated inside that $/sf number and as Mike just illustrated, you can’t deal with covariant variables on the FNMA programmed marketing grid because the algorithm is totally incorrect. Covariant variables have to be dealt with in the aggregate, like Mike said, by using percentage adjustments as one method. Once you have adjusted all of the sales in this manner and graph the GLA vs. the adjusted sale prices, if the trend line is perfectly linear you have done a perfect job of adjusting the data set.

It is not a matter of switching over to MRA from another system, it is a matter of doing the regression you are already doing correctly.

PS: About two years ago I did an appraisal for a bank I usually don’t do work for. It was involved and had a dwelling adjoining a farm the borrower also owned, so I valued it using a regression program. I got a call from the local banker wanting me to explain this thing I sent him. I explained what I had done. He said: "I understand now. My college professor gave us a demonstration of that once and I see where you are coming from." He said his problem was that his reviewers couldn’t understand the correct method of doing an appraisal using this method, but they could understood the incorrect methods very well, so would I mind redoing the house appraisal incorrectly for benefit of his reviewers. He was not being sarcastic either.