# Terminology in sociology & statistics

What is the word commonly used in statistics when the thing that you want to measure cannot be (or was not) measured directly, so you assume that another thing that can be (or was) measured gives an estimate of the frequency of the thing that you want to measure?

For a (silly) example: You are interested in knowing how many Roman Catholics were in the survey but that question was not asked. Number of children in family was asked, so you use that figure as a (silly, I know) basis for inferring RC status, say when x>4.

I am pretty sure there is a term of art for this, something like token, substitute, stand-in, but can’t seem to recall it.

• I'm not gonna answer since I'm not sure, but I've seen the verb "infer" used in those situations. "The actual value wasn't directly measured, but can be inferred by the rate of X statistic...". Mar 3, 2019 at 18:05
• I'm not sure this question is a great fit for us. I'm flagging it for migration to Cross Validated. Mar 3, 2019 at 18:07
• I went ahead and posted it at “cross validated.”
– GAS4
Mar 3, 2019 at 22:41
• I'd call it a "proxy", often computed or logically inferred. 2nd Definition at Link below, reads "A figure that can be used to represent the value of something in a calculation. ‘the use of a US wealth measure as a proxy for the true worldwide measure’" en.oxforddictionaries.com/definition/proxy Mar 3, 2019 at 22:54
• Ys, thank you. That’s what I was looking for.
– GAS4
Mar 4, 2019 at 3:56

This is really @Amadeus's answer (so I'll delete this if s/he posts under his/her own account), but I didn't want it to be lost in a comment:

I'd call it a "proxy", often computed or logically inferred. 2nd Definition at Link below, reads "A figure that can be used to represent the value of something in a calculation. ‘the use of a US wealth measure as a proxy for the true worldwide measure’" en.oxforddictionaries.com/definition/proxy – Amadeus Mar 3 at 22:54

• This is the term I was looking for.
– GAS4
Mar 14, 2019 at 15:37

Operationalization.

"operationalization is a process of defining the measurement of a phenomenon that is not directly measurable, though its existence is inferred by other phenomena. Operationalization thus defines a fuzzy concept so as to make it clearly distinguishable, measurable, and understandable by empirical observation."

https://en.m.wikipedia.org/wiki/Operationalization

"How" you measure a thing is how your study is operationalized. E.g. how "liking ice-cream" turns into a categorizable dataset.

Some of these other terms, like direct and indirect measurement don't really apply in the social sciences, because we're not dealing with physical objects.

https://www.socialresearchmethods.net/kb/unobtrus.php

It is simply called an indirect measure.

• @weakdna In what way does me giving the correct terminology not answer the question?
– user37204
Mar 13, 2019 at 20:58