Editor’s Note: This is Part 1 of a new series on Bias.
Easy – Just don’t be biased. Don’t even think about it!
Alternately, we can have the government regulate and require it!
To know what ‘unbiasedness’ is, and how to achieve it, we have to be able to measure it. We cannot enforce anything, unless we can tell when it approaches zero, no bias.
To intelligently consider the options, we need to know what bias is. As we have discussed here before, there are two kinds: 1) personal/human bias; and 2) analytic bias (data or algorithm).
In my classes in econometrics and statistics is where I learned the word “unbiased.” I have not heard the word used in the world of prejudice or unacceptable discrimination.
Unbiased is clearly applicable to the practice of valuation. Whatever is the goal, or problem to be solved—the models and algorithms should lead to that result without bias. In the world of statistics, econometrics, and psychology, we can reduce this analysis further, by asking the question, how do we avoid bias? In valuation, and in other analytics purposes, this can be salami-sliced into six parts: accuracy, precision, validity, sufficiency, appropriateness and reliability.
- Reliability. This combines accuracy, precision, and validity – plus consistency across varying situations. Yet reliability itself has to be measured in human terms. Is the result ‘sufficient’ for practical use in decision-making? We need to add this into our methodological soup.
- Accuracy. A measure of how close a single measurement is to the true or accepted value. Problem here is how do you know the ‘true’ value to compare to? Hmmm?
- Precision. How close together are repeat measurements to each other? No problem here, the deviations can be quantified by measures of variation, such as standard deviation (sd).
- Validity. Does the study measure what it is intended to measure? The study model or data cannot answer that question. It takes a human.
- Sufficiency. Is the result sufficiently accurate and precise (and valid)? This is also a judgment call. If a human is making a decision, that human must evaluate all of the above.
- Appropriateness. This is the hardest of all! How do we know that the right question was asked in the first place, even if we have great validity? Does the human know what question to ask? Did we pick the right human? And is that human question-biased?
Wow! I had no idea that being right is so hard!
In valuation, analytic bias can come from two primary types of error. One is data selection bias, and the other is algorithmic selection bias.
Data selection involves both the subject property, and the comparable data. An error in subject measurement or categorizing results in a model bias of exactly the amount of the error. An error in any one ‘comparable’ is moderated (or averaged out) by the other comparables. If there are three competitive sales in the data set, a $100,000 error results in a direct bias of $33,000. If ten competitive sales are included, the expected error is only about $11,000!
And if the original $100,000 error was in one comp as a single case (because it is a ‘random’) – then the probable overall error for ten competitive sales is dramatically less, perhaps around $2000 – $3000. It is now a visible data outlier, easily identified, wrangled and fixed by the expert modeler – the appraiser.
Lesson #1 then is (analytically) more data is better than less data (usually). The science of data, and Evidence Based Valuation (EBV)© emphasize the use of the right data in the right amount.
More unbiased bias issues coming in coming blogs. Stay tuned.
Michael J Simmons
August 3, 2022 @ 7:11 am
Hi George. Isn’t the challenge here more fundamental. The charge is that the ‘same house’ in two different neighborhoods should enjoy the same value. The problem is that markets do not recognize that imperative. In fact, they specifically distinguish those neighborhood differences. What’s more, this misguided perspective (in my opinion) acts as an impediment to improving and supporting those elements that make a neighborhood healthy and desirable … the real key to wealth building. Making the right investments by government or communities to inject sustainability into neighborhoods that suffer from the lack of those elements is the only path to solving for that equation – and in the process help raise a new generation of citizens where hope and opportunity become part of those inalienable rights that our constitution promises.
This is not an appraisal problem – this rests on the shoulders of us all as a society.
J. Parsons
August 3, 2022 @ 7:23 am
All the data that is needed for an appraisal is out there. It’s just about getting it, which is time consuming. Humans have a built-in bias. All of us have this. The difference is being able to recognize that we all have a built-in bias. HGTV did a series on the color of a front door. The correct color of the front door did not increase the value (cost of money to carry a loan P&I). but it sold faster (increased marketability). The best data is the home that sold next door (not a mile away). Same floor plan. No adjustments. Proof? I have 5 land sales. I spoke with all the local builders on the cost per square feet (market price). You say support. It’s all there. Time, positive or negative, the data is available through the MLS data ports (statistics). Yes, were all in a hurry. Rush through the reports. I recently took a comparable photo. I would not have known that the location was inferior to subject, unless I had driven it. I had to go back to the report and make an adjustment, after I took the comparable photo. There is nothing like “boots” on the ground. Desktops, hybrids, there fine, with a high loan to value ratio. In the next 5+- years, who is going to get the blame for ALL the foreclosures. There will not be any appraiser’s around to blame. Credit the AMC Model.