Friday, April 08, 2016

Views from the Abyss #22: Triangulation

Q. You previously alluded to "triangulation" as a means to approximating a truth while discussing ideologies. Could you elaborate on that?

A. Certainly.

Imagine you are an employer, and are currently recruiting for a position in your company. You've come across a resume for one Mr. E, which has taken your interest. It details work experience and qualifications, like you would expect, but it doesn't really tell you who Mr. E is as a person. So before you invite him in for an interview, you want to get an idea about who he really is, to see if he's likely to be the sort of person who'd fit your dynamic, fast paced, but family like corporate environment.

So you decide to contact someone who knows him, and ask them to tell you about him. The requirement is two sides of A4, absolute honesty, and all parties are happy to agree to this.

The problem with this approach should be immediately obvious. Depending on who you ask, you're going to get a very different response. Contradictory even. And as there is nobody that can give you an objective answer (as there is no objective answer to give), it's fair to conclude that whatever anybody tells you is going to be so heavily biased as to be utterly unreliable.

This is where triangulation comes in—let's use those biases to our advantage!

Instead of simply asking one person who knows him, let's ask several. And not just people at random, but people whose biases can be established in advance. One might, for example, choose to ask the following:

  • His former/current boss
  • His mother
  • A former/current co-worker that generally thinks positively of him
  • A former/current co-worker that generally thinks negatively of him

This will give us four very different perspectives. We're going to hear about him from somebody who is primarily interested in his work ethic, somebody who loves him unconditionally (or at least, should do), somebody who has developed respect for him as an adult, and somebody who has come to dislike him under the same circumstances.

All of these viewpoints will conflict and contradict each other, but this is only a problem if you take them as literal truth, and that's the last thing we intend to do. On the contrary, we know that each account is going to be heavily biased, and because we have a strong idea in what way they are biased in advance, we can easily determine which aspects of each are likely to be overstated, in order to effectively tune them out.

And with that in mind, we can begin to develop a more coherent and consistent picture of who this Mr. E is.

Q. This is all well and good, but I notice a number of flaws with this analogy. For example, what happens if they ignore your instruction and try to each give you a neutral picture? And how do you determine the coworkers that view him positively and negatively?

A. There is no such thing as a perfect analogy. When used as an explanatory aid, it's perfectly acceptable for the analogy to have its own shortfalls, providing they are not also shortfalls of the phenomena being explained.

With that said though, there are weaknesses with this approach. It is not an exact science. This is why we talk in terms of approximating a truth.

No comments:

Post a Comment