Measuring success in one metric

For some things, measuring success is pretty obvious, and the metrics are nicely binary. e.g., “Did the boomer go BOOM after you pulled the trigger?”

In other things the metrics are either much harder to measure, or there are “good” reasons that the people measuring success don’t want a good metric; it would show they are failing miserably. Or, worse, they are no longer needed to “do the job.” Pick just about any political appointee and the example writes itself.

Anyway, the reason I ask is that I am, among other things, a teacher. Yeah, I know, taking one for the team here guys, leave me alone about that, will ‘ya? So in preparing for an upcoming interview, I started to think about what sorts of things I can ask them – that’s always a “fun” place for the interview to go, because it can’t be any trivia you can just read off the school web-site, but it also cannot be something that exposes glaring problems or hypocrisy in their system, because after they uncomfortably give you a non-answer you’ll not be offered a job. So it’s a balancing act.

So this question popped into my head, and I thought I’d bounce it off ya’ll to see what sort of trouble I might get myself into, but also maybe find some good follow-ups. I’ve got what I think it’s a pretty good measure of success, but it would likely open a huge can of zombie attack worms the size of anacondas, which I don’t want to deal with just yet. So, the question is:“Right now, ‘success’ in K-12 public schools is normally measured by a school with very simple things like graduation rates, graduation-on-time rates, college attendance rates, rates of earning 4-year degrees, etc. I can see why a school would do that. But throw all that thinking aside. If you could gather all the data the schools, business, the NSA, everyone, had about each cohort of students that pass through these halls, if you could comb all the data and point it down to one number 20 years from now, what would you want to measure to know how well you are succeeding? Incarceration rates? Mortality rates? Full employment rate? Percentage with median or higher income? Drug addiction rates? Welfare rates? “Happiness” rating? Number of Nobel Prizes received? Average credit score? Home ownership rates? Average debt load? Some sort of wonky hybrid thing? How would you measure success?”

I like this question because it strips bear a lot of the high-sounding but hollow “vision” and “mission” statements, and gets to the core of what they actually value, what their actual goals are. I’m sure they’ll flip it around and ask me what I think, but I think I can side-step that easily enough. (“That’s immaterial. You are the leader. You’re hiring someone to implement your vision, not take over. If you have a good metric, I’ll take the job if offered and do everything I can to help you get there.”)

If folks here are nice, I might even let people know what I think the best metric is.

So, thoughts, spitballs, observations, suggestions, or cogitations from the peanut gallery?

12 thoughts on “Measuring success in one metric

  1. You get what you measure.

    Some people stare at me confused when I say this but it’s very important. If you start measuring the wrong things you can incentivize things you don’t want.

    For example there have been some idiots who start measuring programmer performance by the number of lines of code they write. End result less readable code as they inflate the line count. Same thing often happens with regards to features and bugs. If one is used as the metric of performance while the other is ignored you will lopside to having a lack of good features for an abundance of change with poor code quality.

    It’s very hard to measure “success” because everyone has a different definition. Overall having someone turn into a productive memeber of society is the most basic version of success in my mind.

    • The “lines of code produced” metric dates back to the early days of computers. I always liked E.W. Dijkstra’s take on it: he explained that the metric should be “lines of code consumed” — since the use of a lot of lines often means there is excessive complexity, and therefore bugs.

      • I can make incomprehensible code in a few lines as well as many lines.

        Making clear, concise code is almost an art. The programmer needs to have an excellent command of the language and a through understanding of the task the code must accomplish. Then they need to distill it down to it’s essence with a minimum of entanglement with externalities.

  2. ask them if they are producing good people who can think for themselves, and don’t necessarily spoon up all the garbage they are fed by the talking heads and authorities.–

    that will assure that you will not be hired, but, it should be fun.– you know, if you just get a wild hair up your rear and think the interviewer is an absolute jerk, and you would like to pull his chain a bit. and, it does sound like fun.

    john jay

  3. This isn’t ‘the’ metric, but it might be useful. College entrance success just measures inflated GPAs and SAT scores. But measuring four year degrees actually awarded, particularly in STEM majors, might provide a decent metric for study and analytical skills imparted in K-12.

  4. I would measure earnings after graduation. Perhaps starting salary or maybe money earned in the first three years after school. After that the school influence diminishes and other factors dominate.

  5. With Barron’s point in mind (I used to work someplace that placed great faith in what we termed “furlongs per fortnight” presentations of various types of measurements, eg., very pretty in 4 colors but worthless as statistical representations of what was actually happening), here’s my suggestion.

    Since they undoubtedly believe they are successful, ask them, first, exactly how they define “success” and the top 3 criteria they use to define “success” and where those 3 align with their “roadmap” criteria, eg., the group of statistical markers indicative of consistent travel in the correct direction (“correct direction” being defined as “confluent with their definition of success.” As to whether their definition of “success” is the correct one, that’s for you to decide.). Grad rates are easy to determine, as is GPA, college acceptance rates, college grad rate, income at 10, 15 and 20 years, but….own-institution grad rates are easy to determine, GPAs are easily inflated (through various means), and everything own-post-grad is suspect, statistically, because there’s no control over measurement methodology or criteria. Who’s better off 12 years later – the computer wiz with an M.A. and Phd. working for Huge Company or the person who became a forest ranger in Montana right after high school and now knows more about biology and botany than anyone at Stanford?

    This certainly enters the minefield of asking uncomfortable questions, but with some rephrasing to make the questions more gentle, there’s an opportunity to gain information useful to determining if you want to be associated with their venture. After all, you’re interviewing them just as much as they’re interviewing you. Probably more, because the downside of hitching one’s wagon to the wrong star is potentially greater than owning the wrong star.

  6. That’s a hard one to pin down, and you have to bear in mind Barron’s comment at the top (‘you get what you measure”), unless you just don’t tell anyone they are being measured.

    But I think I would boil it down to “Total number of dependents claimed (accurately) on a tax return, measured additively per year, counting a spouse as a dependent”. That covers whether you’re married, whether you have children, whether you are providing for those children, how long you successfully provided for them.

    And I’ve give a bonus to anyone who successfully eliminates the income tax to make up for the lack of claimed dependents. 😉

    • This is on the right track. I’m thinking number of children x years married ÷ number of marriages. The only measure of societal health that matters is whether people get married, stay married, and have kids and pass their values on to them.

  7. Everything I’m seeing here reflects and/or buys into the notion that the ‘customer’ for these schools is ‘society’. The proposed metrics are more macroeconomic and utilitarian. The important metrics are ones analyzed by economists and sociologists.

    How about something really wild? Ask the parents and students. You’d want to revisit both after some number of years to see if the school served them well in the longer run. Schools get this sort of feedback from the self-selecting subset of parents that show up at school board meetings, but that’s like online polls. Isn’t it?

  8. This is an interesting thread that I have had some experience with and certainly have the correct opinion (weak smile here please). In an early part of my career, I was an Industrial Training Manager for a large company. We struggled with developing effective training and determining what was effective and what was not. What certainly was not effective was responding to requests for “X” hours of training on (name the subject) from line managers for their crews. This is not unlike schools today where students are confined for a period of years and success at unstated or unrelated goals is expected. We know how well that works. We found an effective consultant that had developed a system to develop training focused on the accomplishment(s) needed to eliminate the performance problem. (Sounds easy but is not necessarily so.) Needless to say, some training now only took a few minutes to be a measured success while other training took much longer than the “X” hours requested. We eventually got our line managers “trained” to request an analysis of their specific performance problem so that we could train people effectively. This same consultant had a pet cause trying to get this type of training into public schools with what he dubbed “The Accomplished Citizen” program.

    So, the question to ask would be something related to how would they know that their graduated student could be considered an “Accomplished Citizen”. This goes back somewhat to the early comments in this thread which I must say shows the biases of the writers. An accomplished citizen does not always have a four-year STEM degree. Perhaps the better measures to look for will be related to how well the graduates function in and contribute effectively to society. A key question to ask is: How many of your graduates are self-sufficient and in how much time? You can take that question in many directions, but do not loose sight of the fact that society requires many varied accomplishments to be done well in order to sustain the general welfare of the people.

    I once had a high school senior tell me that he planned on becoming a garbage collector. I had the presence of mind to ask why he was planning that instead of chiding him for setting such a low goal. He explained very succinctly how it would work to his benefit financially and how he would actually be better off financially than his peers who went the four year college route. Some people look at jobs as a means to fund their fulfillment elsewhere while some expect the job to be their fulfillment.

    A word of warning however. I used this type of question in a job interview, trying to elicit an idea of how my performance would be measured. I just got a deer in the headlights response and no call back. Don’t expect many education administrators or business managers for that matter to be able to think objectively beyond what they have been conditioned to expect.

  9. Since the answer, and the appropriate metric, varies by person, no one question works for a large cohort – what I suggest is to ask whether the students met their goals: ideally, students could write down their goals freshman year, possibly with some guidance as to what they were (degrees, income, marriage and kids, travel, etc.) and then set a number of years later and see how close they came to meeting them.
    If the goals are doable but a stretch, then the graduates are prepared for life if they meet them or come close – if they are still living at home and working close to minimum wage with a college degree, then they haven’t. It would be easy to develop a 10 point scale to quantify degree of meeting the goal; it could even be done as part of setting the goal.

Comments are closed.