Qualitative_data_analysis

WHAT IS QUALITATIVE DATA? 29
what a ‘school’ is, we can count schools. Indeed, it may be hard to describe and
compare qualities entirely without enumerating them. To identify a boundary to
our concept of a ‘school’, we may need to consider more than one example. If for
many examples we can recognize some characteristics commonly associated with
this concept, then we may have more confidence in assigning this category. The
development of categories is rooted in repeated observations, and this entails
enumeration of what data does or does not ‘fit’. Therefore enumeration is not a
luxury extra, but integral to how we classify data.
Statistics is just another form of counting. However, statistical procedures often
require assumptions about the probability of obtaining responses which can only be
satisfied when using random samples, typically through the survey method. Where
we have satisfied these assumptions, there is no reason why we should not adopt the
appropriate procedures, whether for testing for associations between variables or
generalizing from a random sample to a larger population. Where we have not
satisfied these assumptions, we can still use statistics to examine the quantitative
aspects of our data, for example, to be more rigorous in recognizing or creating
classification schemes. Some simple statistical procedures, for analysing frequencies
and cross-tabulations, may prove useful in analysing even the most idiosyncratic and
unstructured data. This use of ‘quasi-statistics’ (Becker and Geer 1982) can enhance
the rigour and power of a qualitative analysis—providing always that we keep in
mind just what the numbers mean.
It is more useful to define qualitative data in ways which encourage partnership
rather than divorce between different research methods. In suggesting that
quantitative data deals with numbers and qualitative data deals with meanings, I do
not mean to set them in opposition. They are better thought of as mutually
dependent. Number depends on meaning, but in a sense meaning also depends on
number. Measurement at all levels embraces both a qualitative and a quantitative
aspect. However, the nature of this relationship changes as we move up the
measurement hierarchy. The more stable and fixed the meanings we can assign to
data, the more we can use with confidence the elegance and power of mathematics.
The more ambiguous and elastic our concepts, the less possible it is to quantify our
data in a meaningful way. We can use a T’ai-chi T’u diagram (Figure 2.7) to
symbolize this relationship, as this depicts a dynamic balance of apparently opposing
forces (cf. Capra 1983:119–120), in this case qualitative and quantitative.
The diagram reflects the mutual dependence of both types of data. It indicates
that meanings cannot be ignored when we are dealing with numbers, and numbers
cannot be ignored when we are dealing with meanings. Each complements the
other, though at lower levels of measurement questions of meaning are uppermost,
while at higher levels of measurement, questions of number loom largest.