You are on page 1of 13

Third Edition

Research

j: /, .

Guahtitative and

, Perspectives
. " " 'v

Ted Palys
Permission to copy obtained through ',Simon Fraser University
Calvin College Annual Academic License
with Copyright Clearance Center

THOIVISON

NELSON

Australia Canada Mexico Singapore Spain United Kingdom United States


4 CHAPTER 1 - PERSPECTIVES ON RESEARCH
...........................................................................................................................................................................................................

But beyond these principles there are few things Nor did I make up these terms: members of
on which all scientists agree, and the variation each camp actually use the term "qualitative" or
among perspectives can be immense. This seems "quantitative" to describe their data, or the
appropriate given what is perceived to be at stake- approach underlying how the data were gathered,
ultimately no less than truth itself. when they write books and journal articles. But
Not surprisingly, over millennia, much time has the knowledge that qualitative and quantitative
been devoted to considering such fundamental researchers have about each other-at least as
questions about truth and knowledge as whether revealed in polemical articles-can be very sim-
"facts" exist; how we might recognize one if we saw plistic and stereotypical. - Members of each group
one; and how to distinguish "legitimate" data and can become extremely self-righteous, believing that
"reasonable" interpretations from their opposites. If their perspective is the only "rational" choice any
you're serious about becoming a researcher, you truly thinking person would make. Of course, by
owe it to yourself and to your colleagues not only implication, this means that anyone who makes j
to enrol in research courses where these issues are the opposite choice cannot be a very thoughtful j
discussed, but also to take courses in such disci- person at all. Kidder and Fine (1987) capture the
plines as history and philosophy, where the nuances spirit well with their ;tongue~in-cheek description
of those issues are considered more fully than Lean of how quantitative researchers perceive them-
do here. selves as seekers of ','numerical precision," 'charac-
terizing qualitative researchers as "navel-gazers." In
QUANTITATIVE AND QUALITATIVE contrast, qualitative researchers refer to their data
RESEARCH as "rich in detail," writing off quantitative
..... : : . researchers as "number crurrchers." .Both "betray
This chapter embodies more modest goals,..- not onlya preference for-one but also a distrust of
to begin to talk aboutithe varied methods and the other" (57).
perspectives of rhesocial sciences by introducing -But my distinction between qualitative and
you to-two research traditions or perspectives: quantitative perspectives is. nor-simply descriptive.
. tg:n3ili.t~~ire., approaches and quantitative In abstracting the concepts qualitative and quanti-
·'~le·.''''~,,~~fY \" . tative to explain what appear to be-oonsistenrpar-
approaches.. . ,
When'f dichotomize all research into only two terns in what I observe, I am also being infirential.
traditions; I'm simultaneously on solid ground and And that's where I'm admittedly on less solid
shaky ground. There's abundant evidence to·sug~ ground because the boundary between qualitative
gest that dividing research and researchers into and quantitative can-be a hard one to demarcate
these two categories describes a state of affairs that clearly. Ir'sa bit.like dividing people according to
really 'exists in the world. Practically any social sci- whether they're "shorruor "tall": it may be a useful
ence department in: any university, and .certainly distinction to make, especially when picking
any discipline as a whole, will contain people whose players for the basketball team, but any dividing
research falls within one of these two perspectives. line between the two groups is bound to be ~tbi-
And although such persons sometimes coexist quite trary and hard to justify beyond .the specific cir-
happily, they sometimes inhabit two virtually sep- cumstances in which it's used. Despite many clear
arate worlds: they speak with those in their own differences between rherwo research. camps>, there
group more often than with those' in the other; are also areas of significant overlap; and researchers
they tend to read and publish in different journals; who .are allegedly from one side often incorporate
and often they attend different conferences, con- features from the other.'
/'_;".,/iIII"

gregating only with people who share their episte- But dichotomies can be useful pedagogical
mological assumptions. devices, and because I think that's the case here, I'll
NEL
CHAPTER 1 - PERSPECTIVES ON RESEARCH 5

use this one to give you the flavour of each tradi- while secretary Auguste Comte, to express the
tion. I'll first discuss the dichotomy in abstract ideas that the world consists of phenomena which
terms and then show how one dispute can be are real, useful, certain, precise, organic, and rel-
understood in qualitative-quantitative terms. Keep ative and that knowledge consists in and only in
in mind, though, that in distinguishing berween the description of the coexistence and succession
qualitative and quantitative research and of such phenomena. (225)
researchers, I'm actually much more interested in
forging links of communication berween the rwo Many have treated "positivism" and "quantitative
than in driving them further apart. approaches" as virtually synonymous (e.g., see
Blumer 1969; Denzin 1978; Schwartz & Jacobs
Quantitative Approaches 1979). Although equating these terms may have
been understandable in the 19th century, it's no
Quantitative approaches have traditionally been longer accurate to do so. Being a positivist does
the dominant of the rwo traditions: they represent imply a quantitative approach, but engaging in
the mainstream in a wide variety of social science quantitative research no longer means you are nee-
disciplines (e.g., sociology, psychology, economics, essarilya positivist. Those who have avoided quan-
political science, business administration). tification because .they. are dissatisfied with
positivism are simply practising guilt by associa-
A NATURAL S,CIENCE MODEL tion, and an imperfect association at that.
Although quantitative approaches have a long
philosophical lineage, ,their contemporary forms A REALIST,PERSPECTIVE
are often traced to 'the mid- to late 19th century. The attribute most strongly associated with posi-
Such individuals as Auguste Comte (in sociology tivism is its realist perspective. Filstead (1979)
and social psychology) and Wilhelm Wundc(in notes that "at the heart of the distinction berween
psychology); noting the tremendous theoretical the quantitative and qualitative paradigms lies the
and technological advances in the natural sciences, classic argument in philosophy berween the schools
argued 'that natural, science methods could' be. of of realism, and idealism" (34). MQ§O! vigorously
service to the social sciences as well. The metaphors applied in the context, of positivism, realism impJj~s
they used to describe the challenge.to social scien- adhering to the' notion that there isa reality oqt
tists were replete with natural science imagery, For there-that awaits our, discovery. .Positivists-aimqp
Comte, "societies and groups [were] organisms- uncover the/facts .andro understand. the lawsx;Q_~
analogous to biological or physical organisms- principles that account for those fact~ositiy!sts
that exist and behave-in accordance with objective maintain that we. need' only think of the "right"
and external laws" (Faulconer & Williams 1985: theoretical concepts and develop. techniques that
1181). For Wundt, physlts and chemistry-were the are, sufficiently: precise to .measure.and- test them~
models to which social scientists should aspire; he The positivist reliance on .a natural science
viewed psychology's task as one of uncovering the modeland a realist perspective influences the-types
"atoms of consciousness" (Adair 1973). of theoretical concepts and data considered legiti-
mate for.inclusion in any science of human behav-
A POsITIVIST EPISTEMOLOGY iour. References to Charles Darwin's evolutionary
Comte, Wundt,.and others embraced an-empirical theory were frequent among turn-of-the-century
tradition known as' positivism. tBhaskar (986) positivists. John B. Watson (e.g., 1913), for
notes that example; was clearly impressed by Darwin's work
the term "positivism" was first used systernati- and its impact on the biological sciences.-A. prime
cally by Saint-Simon; It was adopted by his erst- reason for this advance, said Watson, was. that
NEL
6 CHAPTER 1 - PERSPECTIVES ON RESEARCH
...........................................................................................................................................................................................................

Darwin resisted the temptation to treat the human causes and effects (or outcomes). Recall the image
species as a special entity; instead, he treated of billiard balls being knocked around arable men-
humans as just another organism, subject to the tioned earlier. We see the causes (e.g., the white cue
same scientific principles as any other. Regarding ball hits a red ball), observe the effects (e.g., the red
psychology's love affair with metaphysical notions ball moves and bounces into a pocket), and can
like "consciousness," Watson asserted that develop principles to describe that action (e.g., the
angle of incidence equals the angle of reflection, as
it is granted that the behaviour of animals can be any physicist or pool player knows) without wor-
invesrigated without appeal to consciousness ... rying about what is going on "inside" either ball.
The position is taken here that the behaviour Similarly, early positivists felt that organisms can
of man and the behaviour of animals must be be treated as "black boxes": any invisible processes
considered on the same plane; as being equally that might go on inside (such as thinking in
essential to a general understanding of behaviour. humans) are deemed irrelevant; all that really.
[The social sciences} can dispense with con- counts is what goes in (the causes) and what comes
sciousness. (176) out (the effects or outcomes). Only those causes
external to individuals were deemed "legitimate" -to
Watson's rationale rests on Occam's razor, the scrutinize, largely because· such forces,and processes
principle that simplicity is preferable where theo- ' are most amenable to observation and measure-
ries are concerned. That is, if two theories account ment. We can't see people's-thoughts 'orsmetives,
for a phenomenon equally well, then the one that but we can see what people do. Classic positivists
does so more simply (i.e., using fewer variables and argued that only external; observable forces can be
positing less complex relations) should prevail. If considered "real" and that only such "real'; vari-
we can account for human behaviour and animal ables are worth considering -in the "science" of
behaviour with the same theories, argued WatsOn; human behaviour. ..
why develop separate theories with separate con- In sociology; for example,· positivism' was ori-
cepts? ginally epitomized by .Emile Durkheim. Like
The ideal theory would be both' simple arid Watson" Durkheim argued for a.knowledge that
comprehensive, involving concepts that are adapt- was ·based on objective, observable: causes and
able to any specific situation, in the same way that effects. For him, this iinplied getting outside the
the theoretical concept of "gravity" or "gravitational individual: " ,
forces" can be used to explain the big (e.g., the
movement of the galaxies), the small (e.g., why and We must, therefore, consider social phenomena
how electrons revolve around an atom), and the in themselves as distinct .frorn .rhe. consciously
mundane (e.g., why you need to wear suspenders formed representations of them in the mind; we
when your· trousers are too big). The ideal theory 'must study them objectively' as external things,
would also limit itself to expressing relationships for it is this character that they 'present to us.
among variables and to expressing them precisely, . (Durkheirn1968 [1938]: 252) ,.
preferably in mathematical. terms, rather like
Eigstein's simple but provocative statement that Natural science perspectives on "objectivity"
E=m? were also adapted to the quantitative cause.
Dealing with "reality," while a: necessary step, was
OBSERVABLE CAUSES AND EFFECTS not in itselfsd:'n.~~ufficient to ensure scientific
This mechanistic purity was also sought with progress. Clearly, we would also have to avoid
respect to the variables that were to be included in tainting our measurement of reality with subjec-
any analysis. flhe'world was seen to be made up of riviry..Dernonstrating the reliability and validity of
NEL
CHAPTER 1 - PERSPECTIVES ON RESEARCH 7
· .

one's techniques was thus also paramount, since Hagan also quotes a case in which on'e social sci-
such demonstrations provide assurance that the entist chastised another for "romanticising crimi-
constructs being considered are more than just the nals" and losing his sense of objectivity:
fuzzy and self-serving creations of an isolated
investigator. We ought to restrain impulses, including benign
impulses, that prevent us from seeing the world
OBJECTIVITY THROUGH SOCIAL DISTANCE realistically. Just as anthropologists cannot be
Positivists suggest that the route to objectivity trusted (intellectually) when they "go native" to
requires investigators to depersonalize the research the extent that they glorifY rather than study their
situation, like the proverbial Martian who naively preliterate societies, so a criminologist who has
investigates these strange beings called humans (see gone native cannot be trusted to tell us what
Lofland & Lofland 1984). "Good" data are dis- criminals are like. (Toby 1986 [1938]: 2, quoted
passionate data, far removed from their source. The in Hagan 1989: 157)3
closer one comes to dealing with people on a one-
to-one basis, the more dangerous the situation NOMOTHETIC ANALYSIS
becomes, since one might De' tempted to resort to The ideal of detachment is also consistent with the
metaphysical concepts su~h .as thoughts, percep- quantitative preference for aggregated data, which
tions, attitudes, and values. compile responses from many persons so thatgerr-
Indeed, many quantiratively oriented research eta! trends or patterns across people are-made-vis-
textbooks suggest that the worst event that-can ible, a process called nomothetic aii:i:lysi's~'s
b~fall anyone who engages in-fieldres~aich is for opposed to idiographic analysis.t which is ori-
him or her to "go native" or overidentify=wirh ented toward the case study. This approach assumes
those being studied; -rescarchers sometimes that, across many responses, all "exceptions to the
become So attuned'ahd sensitive ·to the-culture or ,nile,"whether in a positive or negative direction,
group they're investigating that they take on the will cancel one another out-in any group as a
perspective of-that group's members, leaving their whole, making the group "average" the purist state-
supposedly more appropriate detached, analytical ment of how someone in a given situation "typi-
perspective 'behind. Hagan (l989) makes the cally" or "normally" behaves.
common: 'argtimenrrthat 'the appropriate role for
researchers lisstudied neutrality.one should nei- THEINDEPENDENT (CAUSAL) VARIABLES: SOCIAL
ther love nor hate the group one studies, and one FACTS The belief in thedesirabiliry of aggregation
should always maintain some social-distance. He can also be seen in the quantitative attachment to
explains that a problem commonly experienced in social facts, influential aspects of social life that
fieldwork-is ' ;' i~dividuals do not create and that continue to
. " ' operate no matter' how we feel about them .
the 'ttnde'izcyofobservers to oueridentijy-ioith groups: Durkheim (e.g., 1968 [1938]) argued that social
, Thereare.examples in the literature of an anthro- facts, the most appropriate causal factors for social
pologist who married' a cannibal chief arid' of scientists to investigate, exert their influence coer-
other individuals who; without being aware of it, cively, particularly when we try to resist:
have taken orr the mannerisms of'the'groups they
have studied. "Going native" is .a 'situation in Asocial fact is to be recognized by the power of
which the researcher identifies with and 'becomes external coercion which it exercisesor is capable
a member of the study group, and in' the process of exercisingover individuals, and the presenceof
abandons ;his or her' role." as an objective this-powermay; Ire recognizedin its turn either by
researcher; (156.i.57;' italics in' original) the existence of some specific sanction or by the

NEL
8 CHAPTER 1 - PERSPECTIVES ON RESEARCH

resistance offered against every individual effort circumstances which may have had a share in the
that tends to violate it. (1968 [1938): 250) production of the phenomenon are neutralized
and, consequently, do not contribute [0 its deter-
Social facts were seen as desirable research foci mination. The average, then, expresses a certain
because they represented an external reality that state of the group mind. (Durkheim 1968
was thought to influence us and exist entirely inde- [1938): 249)
pendently of our opinions of it:
A DEDUCTIVE APPROACH
The most important characteristic of a thing is For classic positivists, prediction demonstrates
the impossibility of its modification by a simple understanding: if you truly understand a phenom-
effort of the will ... Social facts have this charac- enon (e.g., hurricanes, depression, birth rates), YOLl
teristic. Far from being a product of the will, they should be able to predict its occurrence. Not sur-
determine it from without; they are like molds in prisingly, therefore, quantitative researchers prefer
which our actions are inevitably shaped. the hypothetico-deductive method (known more
(Durkheim 1968 [1938): 253) simply as deduction, or the deductive method),
which involves making predictions and assessing
Thus, for Durkheim, the important social facts their success in an ongoing process of theory devel-
of life, and hence the appropriate causal variables opment.
to study, were such social practices and institutions Chapter 2 discusses this approach in detail; here
as education, religion, the law, and the economic we need only note that it involves beginning with a
systcm.? You and I clearly did not cause them; they theory, deducing a hypothesis (prediction) from the
existed before we did. They influence us all, theory, gathering data to test the prediction (and
although the nature of the effect may vary. And hence also the theory that gave rise to it); and then
even had you and I not been born, they would still either looking for another situation in which to test
exist and still influence whoever happened to be the theory (if the prediction is borne out) or dis-
here. For example, if you were born in Canada, carding or revising the theoryfif the prediction proves
you were born into a capitalist economic system; inaccurate). In. the ideal situation, the effects of cer-
Canada would still be capitalist even if you had not tain variables can be assessed with all other influences
been born here. That system is a social fact of your held constant, making the true experiment the
life; it has affected you in ways that-differ (rom the deductivist's method of choice (see Chapter 10).
effects of being born in, say, a socialist or cornrnu-
nist co uri try. Qualitative Approaches
THE DEPENDENT (OUTCOME) VARIABLES: AGGRE·
The choices that characterize .qualirative
GATED RATE DATA To measure the effects of social
approaches have traditionally been the opposite of
facts, Durkheirn recommerided relying on official
those made by quantitative researchers on every
rate data (e.g., birth rates, divorce rates, suicide
dimension discussed above. Although qualitative
rates, crime rates). Such data deal with matters rel-
researchers' opinions about quantitative science
evant to and affected by "social facts," are outside
vary considerably (from respectful tolerance to
the influence of researchers or of the individuals the
complete rejection), Schutz (1970) is neither
data described, and describe "reality." In addition,
extreme nor atypical.
they make it easier to compare two areas or to com-
Schutz disagrees with behaviourists' and other
pare an area with itself over time.
positivists' choice .to.,investigate a mechanistic
Since each of these [rate) figures contains all the world fr;om the aloof stance of the knowledgeable
individual cases indiscriminately, the individual social scientist, but his disagreement isn't based on

NEL
CHAPTER 1 - PERSPECTIVES ON RESEARCH 9
· .

a belief that such a science would necessarily give very entity we seek to understand. The philos-
"wrong" information. Rather, he feels that in the ophy that expresses this view is known as
long run, such an approach would inevitably fall phenomenologism.
short of a comprehensive understanding of
human action. And it is important, Schutz asserts, PHENOMENOlOGISM
to adopt a method for the right reasons: not Phenomenologists maintain that any effort to
because that method is easier, not because we asso- understand human behaviour must take into
ciate it with some prestigious field of inquiry, and account that humans are cognitive beings who
not because it's expedient to adopt it in the short actively perceive and make sense of the world
term, but because, over the long haul, it's the right around them, have the capacity to abstract from
thing to do: their experience, ascribe meaning to their behav-
iour and the world around them, and are affected
The basic postulate of the methodology of social by those meanings. WI. Thomas (1928) stated that
science, therefore, must be the following: choose "perceptions are real because they are real in their
the scheme of reference adequate to the problem consequences"; that is, in many situations the influ-
you are interested in, consider its limits and pos- ence of "reality" (if indeed such a thing exists inde-
sibilities, make its terms compatible and consist- pendently of our experience of it) pales in
ent with one another, and having once accepted comparison to the influence of our perceptions of
it, stick to it! (Schutz 1970: 270) the situation-s-indeed, those-perceptions dejineour
"reality:'
A HUMAN-CENTRED APPROACH Suppose that next week you were expected to
Qualitative researchers assert that "adequacy" in give a class presentarion on some topic related to
the social sciences begins with acceptance of a more research methods. In "reality," such a situation is a
human-centred methodology, since 'social scien- fairly trivial event; among the "big" things that will
tists, in trying to understand humawbehaviour, happen in your life, it probably won't even rank in
face challenges fundamentally different from those the top· 500. Your career; your future happiness,
faced by the natural scientist: . and your ultimate impact on the human race are
unlikely tobe significantly affected by your per-
The world of nature, as explored by the' natural formance on that one day .'
scientist, does riot "mean"-anyrhingro the mole- . Yet-ifyou're-like most students, you won't treat
cules, atoms and electrons therein. The' observa- it as trivial: You'll do the appropriate preparation at
tional field of the social scientist, however, .;. has the library; you'll <read YOUr presentation a dozen
a specific-meaningand relevancestructure for· the times before the actual day you present; making
human' beings: living. '. 'acting, and '. thinking last-minute changes and trying to learn it by heart;
therein. By a series cf'common-scnse 'constructs you may not sleep well the night before you -pre-
nhey have .pre-selected and' pre-interpreted this sent; and you'll probably coine to class on "the big
world which ·thdy experience as the reality'of their day" feeling at least' a Iitrlenervous. Can we, as
daily lives: It: is' these' thought objects of theirs social scientists, adequately understand your behav-
which ddtermine their behaviour by motivating iour without recourse to the way you perceive this
it: (Schutz 1970: 272-73) situation and the meaning you ascribe to it?
To take another example, many Canadians these
Schutz thus believes that social science ought days, seem deeply concerned about violent crime.
to view humans as thinking, 'motivated' actors; Julian Roberts of the University of Ottawa reported
while acknowledging that. as both-social scien- that concern about violent crime had increased'sig-
tists and humans; social scientists are part of the nificantly during the preceding five years (e.g., see
NEt
10 CHAPTER 1 - PERSPECTIVES ON RESEARCH
...........................................................................................................................................................................................................

.Mitchell 1994<.:), an increase that in turn is related acceptable to talk to people, ask them about their
to our collective belief that the rate of violent crime beliefs and experiences, and consider "cognitive
in Canada escalated significantly over that same processes" and their associations with behaviour-
period. In response to these concerns, citizens and fundamental differences between qualitative and
their elected representatives frequently call for more quantitative researchers remain. While quantita-
punitive sentencing, more caution in the granting tive researchers may now talk to their research par-
of parole, and "special measures" that would give ticipants, the belief that one must still retain an
courts greater leeway to incarcerate particularly aloof, socially distant stance remains; "overidenti-
nasty people and habitual offenders for a long, long fication" is still seen as a problem, and the ideal of
time. Yet Roberts also documents the "reality" that, the researcher as a detached analyst is still sought
at least as measured by the rates reported by and praised. We may talk to "subjects," but we
Statistics Canada, violent crime in Canada didn't can't trust what they say. Their job is simply to
increase at all during that period. So which is more provide us with raw data for analysis; deciding what
important in accounting for Canadians' behaviour their verbal utterances mean is still our job.
regarding violent crime: the "reality" of the situa- The difference is subtle but important. For
tion or people's perceptions of it? qualitative researchers; the choice comes down to
Phenomenologists feel that positivists, in their whether it's better to ask people what thrythink is
zeal to mimic the natural sciences, did an injustice important, and incorporate their answers into our
to the very humans, they wanted 'to understand. efforts to make sense of their behaviour, or to ask
Qualitative researchers therefore argue that any sci- only what we think is important, and then try to
ence .of human behaviour is destined to be trivial infer what they must have been thinking in order
a~r incomplete unless it takes people's percep- to give such answers. But surely it is better to ask
tions into account. Any approach that defines itself respondents directly for their own reasons than for
as phenomenological makes understanding human us to try to invent them (see also Becker [1996] and
perceptions its major research focus: if perceptions Blumer [L969] on this point).
are real .in their consequences, and if they are a
major determinantof what we do, then clearly we NUMBERS CREATE DISTANCE
must understand them and their origins. The shift to phenomenologism affected many
This approach obviously clashes with the pos- other.aspects of theory and method. For example,
itivist emphasis on observables: suddenly We must a central aim of positivism was to establish func-
"get inside people's heads" and understand how tional relations among explanatory concepts,
they 'perceive their world. Yet this transition may expressed, ideally, in mathematical (quantitative)
be less revolutionary than it might at first appear. form, Marty .phenomenologists believe that
Even many positivists have come to accept that at imposing a quantitative, measurement just remove~
least some aspects of phenomenological researchers further from directly understanding
approaches have some merit and that strategies human' experience .• Instead, of trying, to come to
that ignore our ability to think, imagine, and so.on grips with, the anger and, powerlessness a woman
are destined to come up short. Perhaps as.a result, feels when assaulted, for example, the requirements
contemporary positivism (often known as post- of quantification leave us either merely counting
positivism or neopositiuisms' is far less dogmatic the frequency of such occurrences (with all the
about excluding "unobservables"-attitudes, attendant technical problems of how to count "cor-
values, beliefs, perceptions, and motives-from rectly") or asking women to describe their experi-
the research vocabulary.f ence .on a series of ten,p.oint rating scales.
Although such changes have altered what more Phenomenologists, in .conrrast, argue, that the
positivistically inclined researchers do-it's nOW closer we can come to such experiences, say, by lis-
NEL
CHAPTER 1 - PERSPECTIVES ON RESEARCH 11

tening to women explain, in their own words, the quantification in principle, but he was cautious
nature of their experiences, the better. about how useful it might be for articulating sub-
jective meaning:
UNDERSTANDING EQUALS VERSTEHEN
Once the variables of analysis are quantified, posi- Causal explanation depends On being able to
tivists argue that the language of mathematics .deterrnine that there is a probability, which in
makes our statements far more precise. the rare ideal case can be numerically stated, but, ,
"Explanation" in the positivist sense involves inves- is always in some sense calculable, that a given
tigating which "big" variables (e.g., social facts) observable event (overt or subjective) will be fo[::""
most significantly affect human existence. Being lowed or accompanied by another event .. )'
able to identify these factors and to predict their [Nonetheless], statistical uniformities constitute' "
occurrence and magnitude becomes the acid test of understandable types of action in the sense of this "
understanding. The rationale is that if you can suc- discussion, and thus constitute "sociologicalgen'.\'\'1
cessfully predict the values a variable will take on, eralizarions," only when they can be regarded asY'
you must understand the phenomenon in ques- manifestations of the understandable subjective~}'
tion; conversely, if youunderstand a phenomenon; meaning of a course ofsocial action. (Webe'rl1:,
you should be able to predict its .occurrence with 1968b [1947]: 30-31) ,,>I;'

some degree of consistency.


Qualitativeresearchersreject the idea that a-sta- , VALIDITY REQUIRES INTIMACY
tistical criterion can ever,define expltliation/6t Qualitative researchers believe tharrunderstanding
understanding. Instead, they embrace Max Webds people's' perceptions requires getting dose to
concept of Versteben, which involves the more "research participants" Of ,"infotmants", or "collab-
intimate and empathic understanding of human orators." You must-spend time with , them; get.to
action in terms of
its interpretive meaning t~'i1l:~ know them; feel 'dose to them, be able to
subject. While quantitative researchers sought .gen~ empathize wlthvtheir concerns, perhaps even be
eral principles of behaviour, 'Weber argued that, in one of thernvif you hope to truly understand. This
themselves, such' principles couldn't account for approach directly contradicts the quantitative view
action in context: ' that "objective": understanding requires aloof
.-_-;, detachment, lest the researcher "lose perspective."
An "objective" analysis of cultural .events.cwhich " Some- qualitative: researchers believe that one
proceeds according to the thesis that the ideal-of can never-iindersrand-a.group of which one is not
science is the reductio n of empirical reality to a part-that male: researchers .can never truly
"laws," is-meaningless .. .'The knowledge of sociali understand what it means to, be-a 'Woman, for
laws is not knowledgeof social reality but is rather example.-or that 'non-Aboriginal researchers can
one of thevariousaids used byour minds for never.know'what-ir: means to "grow up Indian," or
attaining this end ... Knowledgeof cultural events that someone who has never 'spent time in prison
is inconceivable except on a basis of the sif!!tifi~' will "never completely understand what it means to
cance which the concrete' constellations 'ofrreality do time}, "
have for us in individual' concrete situations;
(Weber 1968a [1949):; 91) AN INDUCTIVE CASE STUDY APPROACH
Associated with the view that closeness is desir-
Weber didn't completely dismiss quantitative able is the idea that researchers should listen to
research or the theories associated with inhejust their informants, aiming to understand categories
felt that one had to g6 beyond blanket assertions to and theoreticaledimensions- from the' p'et~p'et;t:i'()'e
account for action in Context: He wasn't averse to oftheir.informants' experience-and to ,incor~rat:e .

NEL
12 CHAPTER 1 - PERSPECTIVES ON RESEARCH

those understandings into their analysis, rather emic dishonesty." A realist approach to studying
than relying exclusively on theoretical categories cheating would affirm that there are behaviours we
imposed from the "outside." Accordingly, qualita- consensuaIly recognize as "cheating" and that some
tive researchers emphasize inductive approaches people are more or less likely to cheat than others.
(where observation in the field precedes the gener- Given these two assertions, our attention might turn
ating of theoretical concepts; see Chapter 2) and to trying to measure either "frequency of cheating"
ca~e study analysis. Instead of beginning with or how likely a given person or group of persons is
theory and assuming that there's one theory that to cheat; investigating why some people are more
willzvenrually account for everything, the qualita- likely to cheat than others; or why some situations
tive approach typically involves beginning with result in more or less cheating than others.
individual case studies in context, trying to under- A constructionist looking at cheating probably
stand each situation on its own terms, and leaving wouldn't deny the usefulness of any of these
open, for the moment, the question of whether approaches. But he or she also would encourage us
generalizable theoretical concepts can ever eventu- to take a step back and look at "cheating" as a
ally be drawn together in anything resembling a sociaIly produced construct. "Cheating" per se
grand theory. For qualitative researchers, theory doesn't exist in the world in any "real" sense;
isn't something you start with; it's something instead, our, using the word. says as much about
you build. those of us who are identifYing the "cheating" as it
does about the people doing it. A comprehensive
COIIISTRUCTIONISM anaIysis of cheating' requires U$ to ask other kinds
We saw earlier that classic positivists embraced a of questions: Why do we consider "cheating" some-
philosophical perspective known as realism. thing worth asking abouri Why do we consider
Phenomenologists, in emphasizing the role .of some behaviours "cheating" (e.g., looking over
human perception in understanding human behav- another person's shoulder, to see What answers he or
iour, adopt a contrasting perspective known as con- she puts down. in an exam) but not. others (e.g.,
structionism. As described by Schwandt (1994), studying together)? We might also want to inter-
view people who have been .idenrified as cheaters
constructivists are deeply committed to the ... about how they perceived their actions: Did they
view that what we take to be objective knowledge consider it "cheating" or did they caIl it something
and truth is [actually] the result of perspective. else? . How did they come to engage in that
. Knowledge and truth are created, not discovered behaviour?
by mind. They emphasize the. pluralistic and ' The realist, then, takes the existence of certain
plastic character of reality-pluralistic in the behaviour categories asa given, believing that there
sense that reality is expressible in a variety of are such things 'as "cheating," "aggression," and
symbol and language systems; plastic in the sense "crime," along with such supposedly "apparent"
that reality is stretched and shaped to fit pur- givens as "birth," "death," "taxes," and "murder."
poseful acts of intentional human agents. They Constructionists, on: the other hand, are at least as
endorse the claim that, "contrary to common- interested in why these categories interest us, whom
sense, there is no unique 'real world' that preex- or where we decide to sample in order.to investi-
ists and is independentofhuman mental activity gate the phenomenon firsthand, where the bound-
and human symbolic language." (125; the last. aries of the phenomenon are, what meanings the
sentence quotes Bruner 1986: 95) termshave for us, and how those boundaries and
meanings change over time. To be a constructionist
Suppose we were interested in the phenomenon is not to deny that certain phenomena.exist, but
of"chea:ting," known in some universities as "acad- just to insist that their existence cannot be corn-

NEL
CHAPTER 1 _ PERSPECTIVES ON RESEARCH 13
................................................................................................................................ .........................................................................

conceptually free. How can we observe anything


pletely understood unless one understands why, without having implicit categories-a product of
how, and to whom they are applied. This perspec- language, which in turn is a product of culture-
tive, which falls clearly within the phenomenolog- with which to begin our observation? And if we
ical traditions exemplified by Weber and Schutz, already possess those categories, how can we argue
has several implications. that the theoretical categories we derive ire inher-
For one thing, many of the research results we ently real? Perhaps our observations say as much
take at face value and perceive as enduring may be about us (the observers) and about the social con-
little more than transient relationships that reflect text in which we operate as about the behaviour we
the prevailing social order. While realists may be
observe.
content to try to assess the effects of race, poverty, One example of research guided by the con-
being gay, or taking illicit drugs, constructionists structionist perspective (although it actually incor-
argue that we can understand such matters only if we porates both qualitative and quantitative data) is
also understand something about how they're con- that by Menzies and his colleagues on predicting
strued and.about the context in which they occur. dangerousness (e.g., see Menzies 1985, 1989;
Bronfenbrenner (1977), for example, argues Menzies, Webster, & Sepejak 1985; Sepejak,
that much of the research that purports.to evaluate Webster, & Menzies 1984). The researchers exam-
the "effects" of particular familial arrangements ined assessments of dangerousness in a pretrial psy-
(e.g., single-parent families, children of divorce, chiatricclinic in a major Canadian city over several
daycare use) on developmental outcomes (e.g., years, in the context of the broader process-from
school achievement; :involvement in crime) is, in arrest to the assessment to sentencing and follow-
itself, meaningless .. From a constructionist per- up. Their methods included observation, inter-
spective, there's no inherent or "real" effect to' being views; and archival study.
a child ofdivorce or attending day care that's true They also included a historical analysis of
acr~ss:all .rime and space. We can understand the the. "dangerousness" construct that .encouraged a
effects of divorce only if we understand the'context broader perspective on their contemporary findings
inwhich:divorce occurs and people's'perceptions of (e.g., see Menzies 1985, 1989). Making danger-
it. If children of divorce are indeed more likely to ousness itself..-:.how it evolved as a cdnstrUct, who
become involved in j'uvenile delinquency, this pat- employed it, how it was employed-an object of
tern may simply reflect the social stigma often inquiry not only helped the researchers give.evalu-
associated with divorce or the social conditions that ation feedback within a contemporary ;evaluation
typically ptevailupdn separation (e;g., whetemorhers research perspective (e.g., see Sepejak et al, 1983)
often end up.with b6ththe children and nhe more but also influenced social' theory by articulating
ecOnomically' disadvantaged' situation). Similarly, changes in the construct's social meaning and use.
studies -that show' day care to· be associated with Menzies and his colleagues reveal a decision-
positive effects on later school performance may making process ·that seems to say at least as much
reflect the lack of universal day care; most children about the social beliefs and constraints of the deci-
who go to day care are from professional and better" sion-makers as it does about the behaviour of those
educated families and, hence, .are more advantaged being assessed. Gergen (1985) reports that
and likely to succeed anyway. In.sum, the "effects"
of day care or divorce are associated less with the similar kinds of critiques have been launched
statuses themselves thari.with the social context in against the taken-for-granted character of suicide,
which they exist; ... beliefs, .. ; schizophrenia, .. , altruism, .;. psy-
k For another thing, this approach argues that the chological disorder, ... childhood, .. , domestic vio-
world's structures are in no sense- given, nor is our lence, ... menopause, .. , and situational causes ... In
observation of them in any sense dispassionate or

NEL
14 CHAPTER 1 - PERSPECTIVES ON RESEARCH
..........................................................................................................................................................................................................

each case, the objective criteria for identifying such structed on an ongoing basis (the constructionist
"behaviours," "events," or "entities" are shown to be view). According to those who hold the latter view,
either highly circumscribed by culture, history, or our constructions of the world-and hence the
social context or altogether nonexistent. (267)8 world itself-are open to change.
As we've seen, positivist/quantitative researchers
As a third and perhaps broader example, con- tend to emphasize the measurement of outcomes in
sider the variety of meanings that some social insti- their research. This is consistent with the positivist
tutions have for different peoples. A realist division of the world into causes and effects and
perspective on matters of "criminal justice," for with the view that there are real, monolithic forces
example, would take "the criminal justice system" that rule our lives. But constructionists consider
and the existing Criminal Code as givens, starting the world a more ephemeral, transient place whose
any analysis with the view that "criminals" are those dynamics are more directly contingent on the
who violate the Code and that the extent to which meanings and understandings we use to negotiate
"crime" and "criminality" exist in a society is best our world. Accordingly, qualitative approaches are
reflected by such data as the officia] crime rates also characterized by greater attention to processes,
reported annually by Statistics Canada." particularly the processes by which constructions
.Such an approach may be fine for those who arise .and, by implication, the 'processes by which
subscribe to mainstream conceptions of justice and constructions can be changed.
the underlying values they represent, but, marty
argue that "the law", and those it deems "criminal" Comparing Research Perspectives
• •••••••••••••••••••• ! •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••.•••••

are best seen as socially .constructed entities.


Critical theo~ists note how the criminal law '(and Thus far, we've seen that numerous differences have
even the very division between criminal and civil traditionally characterized the approaches of qual-
law) and its agenciesappear to ,systema.tically ben- itative and quantitative researchers. Table 1.1 '(::om-
efit the upper classes more than the lower ones, pares these differences more, explicitly, .at least
men more than women, and those with Western as they've been associated with each approach
Europeaq/colonia.list roots more than'other (e.g., historically.
indige.Qous) peoples (e.g., see Faith 1993; [ackson ; Each element.in the table has been discussed in
1992; Lowman& Palys 1991; Monture-Oka.nee & the preceding pages, so you should now be able to
Turpel J992). define and explain each element and understand
According-to these views, the Criminal Code whyand how each is characteristic of one. or the
and cr] me statistics are anything but neutral other of the two perspectives discussed in this
descriptions .of. criminal behaviour and its [re- chapter and throughout this book. But be aware,
quency of occurrence. Instead, crime statistics toc.: that. the quantitative-qualitarive dichotomy
embody just one of many .possib]e .conceptions of is more than. an abstract set of principles that
justice and can't be understood outside the context philosophers .of science debate' over afternoon tea;
that makes thejTI,meaningfuL these approaches affect' how people do research
and are ,fundame~,5aL. to understanding any
EMPHASIZING'PROCESS ',' research you read. To see the quantitative-
A distinct difference in emphasis also follows from qualitative distinction in. action, Jet's look at a dis-
",ett~~~jJ1g,the' important elements of,t;h(:lJYf,Qdci pute in the literature; a-dispute that can best be
as essentially stable and awaiting; discovery .(the explained by the different (quantitative versus
,!'.c;;alkt::view~,or seeing rhe.worldas something that qualitative). perspectives that two sets of authors
is actively constructed, deconstructed, and recon- bring to the table.

NEL
CHAPTER 1 - PERSPECTIVES ON RESEARCH 15

Table 1.1
Comparing Qualitative and Quantitative Approaches

Qualitative Approach Quantitative Approach

Human-centred approach: people's ability to think Natural science model: humans are just another
and abstract requires special consideration organism

Phenomenological Positivist

Constructionist Realist

No variables ruled out; internal, perceptual variables Emphasis on observable variables that are external
expressly considered to the individual; social facts

Direct, qualitative verbal reports are preferred; Quantitative measures are preferred for their precision
qua~tifying responses is a step removed from people's and amenability to mathe~atr~al analysis
words and perceptions

Emphasis on processes: perceptions and their Emphasjs oncauses an,d effects: what goes in and
meanings and how these emerge and change how itcomes 'out;' inputs, outcomes
.' ; ": >, ',--; ,"_.

Valid data corne from .closeness and extended contact Objectivity is achieved .throuqh.social distance and a
with research participants detached, analytical stance

The criterion for' understanding 'is' Ve?stehen: The criteria for')understandin'g'are the 'ability to '
understanding behaviour in 'context in terms predict and-statistically significa.nt associafione:
meaningful to the actor between variables

Preference for idiographic, case study analysis Prefer~nc~ r6r ribmothetic analysis aggre'g~f~d over
" . it,"
m~nyi6kses '" " .
Preference for an inductive approach: ~tarting Preference for a deductive approach: starti~g with
with observation and allowing grounded theory to theory and creatinq situations in Which to test
el11erge hypotheses ' , -' ' ,

You might also like