Nine key standards, or principles, of scientific research can be gleaned from texts on research method. As I see them, these are:
7. Objectivity (Pending)
9. Representativeness.
I say something about each of the nine standards in separate posts, as shown in the links above. What I want to briefly talk about here is the measurability standard.
Measurability
Measurability means susceptibility to measurement. It means, in other words, that what is to be measured must embody quantities and/or qualities -- or have features, dimensions or characteristics -- that can be subjected to appropriate or 'true' measurement. Concerning the defining characteristics of a measurement, Bell (1999) notes that:
"A measurement tells us about a property of something. It might tell us how heavy an object is,
or how hot, or how long it is. A measurement gives a number to that property.
Measurements are always made using an instrument of some kind. Rulers, stopwatches,
weighing scales, and thermometers are all measuring instruments.
The result of a measurement is normally in two parts: a number and a unit of measurement, e.g.
‘How long is it?... 2 metres."
The
measurability rule is anchored on the above conceptions, and so requires that the variables around which the researcher intends to collect data should be measurable, or susceptible to acceptable ‘measurement’ (Leedy, 1980: 46)
.
This is easier done in the natural sciences than in the social sciences; in quantitative studies than in qualitative. Still, one
must, in the social sciences too, endeavor to quantify, measure and evaluate. Indeed, the guiding principle of the measurability rule -- its corollary, in other words -- is this:
"What can be measured must be measured." Thus, not measuring what can be measured is not an option allowed anyone.
In quantitative research, the measurability standard features prominently in all methodological procedures that revolve around, or build up to, hypothesis testing. The testability
of a hypothesis is indeed a function of the measurability of the variables that constitute
it. Such measurability is, in turn, dependent on the indicators chosen to represent them; but one is not entirely free to choose just any indicator(s).
They must be such as most peers or reviewers or supervisors or sponsors can be
persuaded to accept as valid and reliable. Still, one has some leeway in choosing the indicators one wishes to use, so long as the reviewers, supervisors or funding entities are prepared to see or accept one's findings “in the light" of the chosen indicators.
Viswanathan (2010: 285-210) has interrogated in depth the challenges encountered in "Measure development procedures," particularly in connection with quantitatively-oriented social science research. Among these challenges, which he believes are not sufficiently appreciated, is the underpinning assumption, which he faults, that "a construct can be [readily and invariably] isolated and examined." Here is his line of thinking:
"By measuring or manipulating individual constructs, relationships between constructs are studied and substantive hypotheses about these relationships are tested. The very notion that numbers can be assigned to attributes of people, objects, or events is presumed on being able to study attributes or constructs separate from other constructs. After all, measurement relates to rules for assigning numbers to attributes of people, objects and events. [But] this is not the case for many phenomena. A complex network of constructs may influence a phenomenon and may not be separable into individual constructs for purposes of measurement."
Measurability procedures implemented in qualitative -- that is, 'non-numerical' -- research attempt, not always successfully, to overcome or circumvent such intractables as Viswanathan has pointed out. These procedures include, for example,
“comparative
judgement (arranging factors in a hierarchy of importance)” or
“scaling
(correct versus incorrect responses to a given set of questions)” (Leedy, 1980: 46). Other ‘non-numerical’ strategies that have been used, according to
Worsley (1992: 113)
include:
“careful classification, organizing and combining of field-notes
[which] requires patient checking and cross-checking and the application of
systematic techniques.” But he acknowledges that
“numerical methods are sometimes applied to field-data too” (Worsley, 1992: 113). Indicators are often used
as proxies/surrogates for more fuzzy or abstract concepts, such as
status (Worsley, 1992: 96). Such indicators are inevitably numerical.
One may also, in connection with the
nominal level of measurement (the
simplest level), and even with the
ordinal
level of measurement, calculate frequencies/percentages, and
cross-classify in qualitative research. Scaling, moreover, is not just about
correct v. incorrect. Incorporating comparative judgement, one may also use
scales of: high v. low, do v. don’t, very high to very low, very good to very
bad, strongly agree to strongly disagree.
To be fair to everyone, no one is quite attempting stubbornly to square the circle in this measurability conversation. Still, the challenges of measuring require, and will surely be tempered with, constant and open-minded vigilance -- as well as an abundance of capacity to deal with critiques and contrary views, including (so be it) one's own self-criticism. Bell (1999: 1), like others before her, subsumes this conversation under the theme "uncertainty of measurement". Thus:
"Uncertainty of measurement is the doubt that exists about the result of any measurement. You
might think that well-made rulers, clocks and thermometers should be trustworthy, and give the
right answers. But for every measurement - even the most careful - there is always a margin of
doubt. In everyday speech, this might be expressed as ‘give or take’ ... e.g. a stick might be two
metres long ‘give or take a centimetre’."
