By Nathan Barton

In the first installment, I compared politicians and engineers, explaining why I believe that few engineers are involved in politics – at least in elective politics.

To help non-engineers understand engineers better, and how we view politics, consider testing.  Let me explain why so many engineering qualification exams and general tests for selecting and instructing (training) engineers are written the way they are.

Many people like true or false type questions.  So do I, but usually NOT for engineering subjects.  Why?  In the real world, while there are clearly stated and known principles of RIGHT and WRONG, for any given problem, it is seldom black and white.  A bridge seldom must be built ONLY using EXACTLY 4 each 20W520 stringers, and only very infrequently will excavating 150 cubic yards of material using an Acme 792 10-kip loader use EXACTLY 25.4 gallons of ultra-low-sulfur Diesel fuel.  So, stay away from those except for the most rudimentary of position qualifications.  It might be okay to ask the temporary summer hire in the sanitation department, “True or False:  the residential storm water fee is $4.25 per month.”  But you don’t want to ask some politician running for the US Senate, “True or False:  The First Amendment protects free speech by ANYONE.”  Or rather, you shouldn’t HAVE to – in reality, we SHOULD, and I shudder to think how many of the currently serving 100 could NOT answer that.  But on to engineering testing.

First, of course, consider “word problems.”  Much hated by generations of elementary school students, many teachers today have given them up: everything is already “set up” when the student is asked to solve the problem.  Well, the real world doesn’t work that way.  Nearly EVERY problem an engineer is called to solve is a “word problem” in which a (normally qualitative) problem description must be converted into numbers which can then be used in equations (normally NOT provided as part of the question) to identify what is needed, and THEN converted back to a “word solution.”  Politicians, of course, don’t think of REAL solutions, so they can skip all the analysis (except what is needed to keep the smoke, mirrors, and that curtain over in the corner intact).

This leads us to HOW questions in exams (not real word) can be written, and graded. In other words, let me explain why so many engineering qualification exams and general tests for instruction are organized as multiple-choice questions.

Good multiple choice questions, like good true-false questions, are very difficult to write. In many professions (and of interest to libertarians), the desire is to have a single “essay” question to test a wide range of knowledge and skill.  However, it is difficult for instructors and evaluators to review those.  There is a place in engineering training for those. (“Design a wooden bridge capable of supporting a Class 60 MBT in crossing a 200-foot river with a muddy-silty riverbed and sandy-loam banks 15 feet above normal high-water mark.”  Eight to ten hours work, and testing their knowledge AND skills AND common sense – and about 20 hours to evaluate and grade.)

But it is possible to determine the test takers’ knowledge, skills, and common sense (in a particular field of engineering) with a quiz of about 50 multiple choice questions, which will take about two hours to take (“show your work, please” is a common thing even on multiple choice questions in engineering schools) and about ten-twenty minutes to do the first grade/evaluation.

Of course, the good instructor goes back to see in WHAT the student demonstrated a lack:  especially if the questions are written well.  And it is VERY VERY hard to write them well. While in solving engineering problems there is very seldom ONE single good solution, any multiple-choice question for engineers must have just ONE correct answer.  So writing those questions and answers requires great skill.

In the next part of this essay, I’ll look more deeply into that idea of testing and demonstrating skills by multiple-choice questions.  Let me leave you with this thought:  both politics and engineering – and testing for skills and knowledge in those fields – are handicapped by the fact that NO test question can ever provide all of the information needed and the complexity faced by even a relatively simple real-world problem.

The problem is really the point that I made in the beginning: engineers SOLVE problems by breaking big problems and big issues into small ones, which can be solved one at a time.  Politicians cannot and will not do that, instead making problems bigger and bigger, while lying (to themselves and others) about how they have the knowledge and experience (and power) to solve the big problems.