1

I'm not talking about solipsism, or the evil demon/genius of Descartes.

I'm talking about the McNamara fallacy, which is obviously a fallacy. Can we say that everything that exists is measurable, even if we cannot measure it a some point in time given some state of the art? That would explain the McNamara fallacy.

It is also often said that what doesn't get measured cannot be managed. But what is measured, is cheated/played with. Are managers hopeless?

Philosophy is said to be very good for critical thinking, knowledge, etc. How could we know if we could not measure the effect of learning philosophy?

If we cannot measure and objectively compare, how can we know that some perception is not the placebo effect, confirmation bias, or just a delusion of any sort? That's what control groups are for.

Finally, metrics could be wrong, as they often are, but how could we know without different and better metrics? In the McNamara example, body count may be a bad metric, but we may only be able to reliably tell so after comparing with other metrics, e.g. area under control, economic cost, etc.

Improve this question
8
  • If the question isn't clear please let me know. It may look like there are several questions, but there is only one written in different ways. Metrics...
    – Trylks
    Apr 13, 2017 at 14:32
  • Have you ever heard the expression that someone "knows the price of everything and the value of nothing?" That's the fallacy of metrics. As McNamara learned.
    – user4894
    Apr 13, 2017 at 15:20
  • That's very true, but the question in that case would be: how can we know that our perception of value is correct if we cannot measure it?
    – Trylks
    Apr 13, 2017 at 15:58
  • How can a perception of value be correct or incorrect? Perception of value is subjective. Else markets could not exist.
    – user4894
    Apr 13, 2017 at 16:13
  • 1
    The perception is subjective, but the value/utility must be objective. Otherwise nobody can know the value of anything, as in the sentence you mentioned, as nothing would have any value (objectively, a quality of the object), it would be subjective (something that the subject/observer assigns arbitrarily). Logically, the perception is correct when it matches the actual value of something (minus working error range). Otherwise, McNamara would have nothing to learn.
    – Trylks
    Apr 13, 2017 at 20:03

2 Answers 2

Reset to default
2

In social science and humanities, qualitative research methods systematically collect observations, identify phenomena, and analyze them without quantification. Consider this report by the research nonprofit FrameWorks. (There are plenty of peer-reviewed examples — qualitative methods are widely used in anthropology, for example — but this was the first fully-public document to show up when I searched my research library.) The researchers gathered data by interviewing two groups of people ("experts" and "members of the public"), then analyzed these data to identify common themes (in the way people talked about the value of natural spaces), and were especially interested in picking out themes that were different between the two groups.

This study involved systematic observation and analysis, but not quantification. So I would say that it's identifying things that aren't "delusions," but without measuring anything in the McNamara sense. "Objectivity" is highly ambiguous. You might think that quantitative measurement and analysis are objective processes because they're "mechanical" or "automatic," meaning roughly that the outcome doesn't depend on the researcher — i.e., different researchers would get the same results. Qualitative data collection and analysis are not mechanical in this sense, so you might worry that they're not objective or reliable. But qualitative researchers use other approaches to ensure that their data analysis is reliable. For example, they'll continuously re-analyze their data until they reach saturation. Or they'll develop definitions of the concepts they're analyzing, have two or more researchers independently analyze the data according to those definitions, and then compare the results.

Improve this answer
2
  • Thank you. At the moment I have to check the qualitative methodology more thoroughly. I will do so with your links and more. To the best of my knowledge, qualitative is a form of binary quantitative, i.e. there are only two quantities: 0 and 1. E.g. for "qualities you wish in your co-workers" we would get sets of words, we can measure plenty of things in bags after union, intersection, etc. And it would still be qualitative. So my current doubt is whether qualitative methods are in fact measuring binary properties, and suitable methods when we cannot measure such properties either.
    – Trylks
    Apr 14, 2017 at 19:23
  • That's definitely not what "qualitative" means in a social science or humanities context. I recommend taking a look at the example I linked to, if you haven't already.
    – Dan Hicks
    Apr 15, 2017 at 21:00
2

Let me ask the flip question: even if I measure something, why should it be correct? Just because something is measurable doesn't mean that it is not a delusion. Astrology uses extensive measurement but its not really true. Eugenics used measurement but it wasn't really true knowledge. There seems to be a fundamental fallacy of presumption in this question. Perhaps it would be much more fruitful to talk about why you think measuring provides a solid ground of assessing what is true and real and what isn't?

Improve this answer
1
  • Precisely, as you mention, measurable means more easily falsifiable. The point of falsifiability is, in colloquial terms: "it is not a delusion, because if it was, we would have noticed". As it was noted with McNamara. For things that cannot be measured this is not so easy, similarly to what cannot be seen/measured, as the Emperor's new clothes, it may be the case that we cannot see/measure them, or that they do not exist at all.
    – Trylks
    Apr 14, 2017 at 21:00

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .