Academic Literature for Multi-Factor Decision Making

Abstract: After writing my article Multi-Factor Decision Making Math, I found and reviewed relevant academic literature about Multi-Criterion Decision Making (MCDM). I found no criticisms of my beliefs in the literature. I found that the MCDM and epistemology literatures mutually ignore each other, to their detriment. And I determined that CF’s decision making method is a new approach that MCDM researchers are unaware of. I will summarize my review and also discuss Geographic Information Studies and normalization.


This is a companion discussion topic for the original entry at https://criticalfallibilism.com/academic-literature-for-multi-factor-decision-making/

Purpose: Comment on an aspect of the article I have some experience applying.

I use GIS in one of my businesses. I hadn’t looked for any literature on it and wasn’t aware any existed. But after reading the article I think the suitability modeling page refers to a similar or same problem as I use GIS for. As a practical matter this problem presented itself to me as one of figuring out how to use a government web site (County GIS) to figure out what I wanted to know, which is things like: Is it feasible to build at all on this property? If so, what types and sizes of houses would be feasible and where, generally, would you put them? I’m not making placement decisions down to where you would actually start digging/building - those are done by people on site. I’m just doing preliminary internet research.

The where should you build a house problem is definitely multi-factor in nature. For simplicity I’ll talk about only two of the factors: slope and flood plain. But there are others.

Slope is covered explicitly on the suitability mapping page. Like the Hugag, it’s generally preferred to build on a gentle slope or no slope at all. The cheapest building spots are perfectly flat, and as you start to add slope it gets more expensive because of the need to fill/grade and possibly reinforce the foundation. The expense changes as the slope changes but there are a couple of different breakpoints to consider with regard to the expense of houses on slope. In some regions for some values of slope it actually gets less expensive as you add slope for a while.

The first breakpoint comes when the slope results in an elevation change of more than half a story over the relevant dimension of the house. Before that breakpoint, less slope is cheaper and more slope is more expensive. After the half story breakpoint, for a while, more slope might get cheaper or at least not significantly more expensive because you’re probably going to do something like a walk-out basement (explanation) which takes advantage of the slope to add a feature to the house. But if the edge of your basement where you’re putting doors & windows is still somewhat underground, then you have to dig it out horizontally to the point where it is more than a whole story so water won’t flow back toward the basement’s door(s) when it rains. If you’ve got more than half a story of slope the cheapest spot then becomes 1 story of slope, because then you can just open a door straight out the back or build a (relatively cheap) elevated deck to walk out on. Either way you will have no water intrusion problems with no excavation other than for the house itself.

More than 1 story of slope rapidly gets way more expensive to deal with, as at that point you either have to put part of the foundation on pylons or have additional stories between the “ground level” on one side vs. the other. Pretty much no one wants to do that kind of stuff unless it’s in a high end/expensive area or they otherwise want it for aesthetic reasons.

What slope degree these breakpoints turn into depends on the dimensions of the house you’re trying to build and how it’s going to be oriented.

The flood plain problem is analog in nature (floods happen with variable frequency and amount) but the government has legislatively turned them into two binary factors: Flood plain itself (also known as Special Flood Hazard Area or SFHA) determines whether you have to buy flood insurance if you have a mortgage, which almost everyone does. Not in a flood plain is better than in a flood plain, of course, since not in means you don’t have to buy flood insurance. But then within flood plains there are different zones related to how often it floods there and how high the water goes. The zones are also binary (you’re in one zone or another) and affect what the insurance costs actually are and in some counties what mitigation measures you are required to take when building. The mitigation measures (such as raising the foundation) also add cost to the build.

The two factors of slope and flood plain are sometimes anti-correlated, as the term “plain” implies (a flat area). Sloped land is usually not in a flood plain because the water just runs off, whereas flat areas, if they’re also surrounded by higher elevations, tend to collect water and be in flood plains. Sometimes a high flat spot isn’t available, so you have to choose a spot with slope or a spot in a flood plain.

Slopes and flood plains are displayed on GIS maps as layers with different colors and indeed, if you try to display too many layers at once in a GIS it’s too confusing to use. Because I’m not making final placement decisions, for my purposes it’s good enough to work with one or two layers at a time and switch between them to figure out it a given pinned spot is a good candidate house site for a particular type of house based on binary pass/fail for each of the layer sets I’m using.

The GIS systems I use don’t let you combine factors in square tiles larger than 1 pixel on the map, nor control the colors based on breakpoints. I think it would be useful if I could create specific colors based on the breakpoints I care about for a given type of house. I could probably do that if I got the raw data & built my own GIS. But I don’t think I do enough of this type of thing to make the effort worthwhile. I don’t know what the people who make final placement decisions use, but I suspect it’s the same stuff I use plus on site considerations and intuition.