Você está na página 1de 9

“While qualitative research methods produce authentic accounts

of human thoughts, feelings and behaviour, they are


unrepresentative and unreliable and are therefore of little use in
the study of politics and international relations.” Discuss.

In the world of social science, there is significant debate over the validity
and reliability of results produced by qualitative research methods.
Whereas some researchers value qualitative research, arguing that it
brings an extra dimension to the way we extract data from studies,
others, as seen in the wording of the question, see it as little more than a
bias-injecting hodgepodge of inconclusive studies. While its critics have
valid arguments, that we will acknowledge, we shall argue that this is not
completely true: qualitative research, albeit controversial, provides us
with a variety of different methods to gather data for particular studies. In
fact, these methods could very well be the best approach possible for
handling particular issues when contrasted with quantitative methods. We
also argue that qualitative research, if it is not to be the star of an
analytical show, can at least play a supporting role alongside other
research methods thus refuting the question’s claim that these methods
are of little use in social science studies. Thus, the essay shall be
structured like this: first, we shall define qualitative research along with
validity and reliability so that we know what qualitative research
embodies and what sort of criteria it needs to meet in order to be
considered a legitimate and useful method for social scientists to use.
Next, we shall examine the ethnographic methods used by qualitative
researchers (participant observation and interviewing in particular),
summing each one up with their benefits and why others argue against
them. We shall then finish by discussing how qualitative research can
involve case studies, including a combination with quantitative methods
better known as triangulation. This will ensure that even if we cannot
settle the disputes amongst social scientists in general over the merits of
qualitative research methods, we can at least finish the debate
demonstrating that in some cases qualitative research methods can at
least be used alongside their quantitative counterparts.

As is often the case in debates, this dispute is as a result of a split


between two main trains of thought amongst social scientists. On one
side, we have the positivist approach. Positivists believe that social
phenomena can be measured and studied in ways similar to how
scientists study the natural world. Positivists generally share a
“preference for measurement and quantification of observable events and
a search for statistical regularities that can be understood as casual laws”
(Seale, 1999:21). Thus, it makes sense that they will more likely prefer
formal quantitative research methods over qualitative. On the other hand,
we have the interpretist view. Interpretists work with the premise that
“methodological monism is no basis for the study of the social world”
(Seale, 1999:21). This train of thought considers the positivist approach
too simplistic and dismissive of the unpredictability of man – better known
as the ontological position (the belief that reality is not objective and
therefore cannot be quantified (Harrison 2001:77)). Thus, greater
emphasis is placed on deductive and intuitive methods to further our
understanding of how society works. Obviously, this implies interpretists
would rather use qualitative research methods – which we shall now
properly define. This is in itself a debateable issue, for as David Silverman
points out, qualitative research appears to be defined more for what it is
not (“non-quantitative”) than for what it is – it has a “negative epithet”, in
other words (2001:25). Thus, as a clear-cut definition of qualitative
research cannot be found, we shall base it on Silverman’s own criteria for
qualitative research. This means that qualitative research is: soft, flexible,
subjective, political, speculative, grounded and involves the use of case
studies (Silverman, 2000:2). Qualitative research thus sounds more
“loose” and intuitive compared to quantitative methods, which are more
objective and “fixed” in nature, adhering more to the methods used by
the natural sciences (Silverman, 2000:3). With qualitative research
defined, we need to address the question’s claim that it is
unrepresentative and unreliable. The issue of representation is better
known as validity. In the world of social science, validity is “the degree to
which the researcher has measured what he set out to measure” (Smith,
1975:61). Thus, a qualitative researcher must ensure that his methods are
appropriate for the project in question. As we will see, qualitative research
methods do appear to fit the bill for certain kinds of issues. We also need
to determine whether qualitative research methods are reliable – that is -
whether the tests conducted using qualitative methods will produce the
same, consistent results if other researchers were to use the same
methods for the same issue (Smith, 1975:58). Obviously, the more
identical the results, the more reliable qualitative research methods would
be in social science studies.

So far we have merely defined three words – it is time to add meat to the
bones and analyse the qualitative research methods themselves. A key
tenet of social science research is ethnography – the study of people. This
kind of approach “revolve[s] around the notions of people as meaning-
makers, around an emphasis on understanding how people interpret their
worlds, and the need to understand the particular cultural worlds in which
people live and which they both construct and utilize” (Goldbart and
Hustler, 2005:16). Thus, a qualitative researcher may choose to get
involved in his or her study either indirectly, such as being present at a
protest march to ask the participants questions, or directly – otherwise
known as participant observation. This method is when the researcher will
directly join the organisation or group for a certain period of time in order
to gather information for the study. This was recently seen when Channel
4 sent an investigator undercover as a member of the far-right British
National Party for six months to gather information their 2004
documentary Young, Nazi and Proud (channel4.com). A researcher can
enjoy a number of benefits from this method. For instance, participant
observation enables the researcher to “conduct the analysis by
abstracting everyday actions to uncover the principles governing
behaviour and by modifying theoretical generalisations to accord with
perceived behaviour” (Bastin, 1985:92-93). In other words, Bastin is
pointing out that alongside quantitative methods such as social surveys
and structured interviews, the researcher is able to experience firsthand
how societal behaviour measures up to trends found in his data. If we
were to base this method on Bastin’s hypothetical situation – that of a
researcher studying a community on the causes of vandalism (1985:93-
100) - then this qualitative method is certainly valid as it enables the
researcher to measure the issue through direct interaction with the actors.
Reliability, however, is arguably the Achilles’ heel of qualitative research,
for as we saw in the previous paragraph, it is the consistency of the
results that determine the reliability of a study. Thus, as Bastin admits,
the vandalism could have occurred for any number of reasons (1985:94).
This implies that researchers in different communities may come up with
different studies. Yet, interpretists would argue that this is a weakness of
quantitative research methods more so than qualitative. After all, it
through this informal, deductive method that such problems would be
identified in the first place – this begs the question as to how, through
sheer numbers and statistics, quantitative methods would fare any better.

Yet, critics of qualitative research will reply with two points against
participant observation. First, they may argue that the reliability of the
results will be fatally flawed, as transcripts and tape recordings of
interviews may blur the proper response that a person gave to the
questions (the stop-start method with tapes for instance may miss out
vital pauses indicating the respondent thought deeply before answering a
question (Silverman, 2000:10)). There is also criticism regarding the
validity of participant observation which Silverman labels as
“anecdotalism”: a researcher may use some of his data to present a
particular point, yet ignore contradictory transcripts that damage his
findings (2000:10). In this criticism, qualitative researchers are portrayed
as “journalists” whose methods are akin to the media’s more so than
scientists’, and subsequently, produce data full of bias (Silverman,
2001:26). Thus, positivists in particular accuse qualitative researchers of
producing biased results and would not be appropriate (valid) for use in
social science. Yet, there are simple explanations that can counter these
arguments. For instance, a commonly-cited refutation in regards to
reliability is from Marshall and Rossman, who argue that reliability is only
really a concern to positivists, as interpretivists (who are more likely to
use qualitative research methods) assume that society is in a constant
state of flux. Thus, it should be expected that results will vary from time to
time (1989:146). There is also a simple answer in response to the
question of validity. While qualitative researchers readily admit there is
the danger of bias in their research, especially in methods such as
participant observation, this is a problem apparent to all students of social
science. Quantitative research methods may too be subjected to bias such
as in structured interviews, for instance. As Silverman points out,
quantitative data was even accused of bias during the Margaret Thatcher
years when the calculation methods used to measure unemployment and
inflation were often changed to make the readings look good for the
Conservatives (2001:26). Thus, the argument is that validity (in particular
bias) is a problem for all methods of research – not just qualitative.
Therefore, when we take these points into account, we can consider
participant observation a legitimate and useful method of research if one
wishes to gain a more intimate understanding of communities and
societies.

There are other ethnographic-based qualitative research methods that we


need to discuss such as interviews. A researcher can interview people in a
variety of ways such as by using telephone, email correspondence,
internet instant messaging or with the simple face-to-face method. Each
form has their own cost and benefit, for instance, email provides quick,
back-and-forth interaction, but provides the opportunity for the participant
to provide brief answers and, subsequently, dilute the quality of the data.
Face-to-face interviews on the other hand provide large volumes of quality
data, yet is time consuming on the researcher’s study. Regardless of the
method chosen, Barbour and Schostak (2005:41) mention the potential
problems researchers face when applying this method: encounters can be
“messy; the level of “commitment” engaged in the interview can be a
factor; there is the problem of “truth”, leading back to our previous
discussion over the problem of bias; there could be hidden agendas at
play; and finally, the ethics of the interview – what tactics would the
researcher resort to in order to extract certain answers from his
participants? For instance, Ben Stein, a creationist arguing against the
biological theory of evolution, deliberately set up the scientist Richard
Dawkins to make him appear as if he believed life on earth came from
aliens. In the interview’s true context, Dawkins was merely speaking in
hypothetical terms (richarddawkins.net). It is this sort of occurrence that
gives critics of qualitative research the opportunity to argue that such
subjective techniques of research can lead to bias and ethical problems.
There is also a problem more seen in elite interviews – the issue of
reluctance of the respondents to express honest answers. This, for
instance, could be seen when researchers interviewed politicians in
Northern Ireland after the Good Friday Agreement; they were reluctant to
give answers that may go beyond party lines (McEvoy, 2006:184). While
such problems exist for the qualitative researcher, such interviews can
provide good quality data when conducted correctly – arguably better
than when using quantitative methods. Lisa Harrison (2001:75) argues
that qualitative interviews “allow people to talk freely and present their
perspective on their terms. In comparison to simply ticking boxes on a
questionnaire, in-depth interviews allow interviewees to use their own
language, so the discussion should flow more freely”. This, in theory,
should therefore lead to greater quality data as it gives the respondent
the opportunity to answer at length. Another advantage of interviewing
qualitatively is the use of focus groups. Rather than survey hundreds of
respondents, a focus group consists of a small number of people that are
interviewed at great length to produce quality data. Although critics will
still argue that such interviews can lead to biased, unreliable results,
Harrison (2001:76) points out the benefits of such a method. For instance,
focus groups, unlike surveys, ensure that the research question is
covered. The researcher also has the option of asking for clarification if he
is not satisfied with a response, and most importantly, focus groups
enable the researcher to understand how a respondent’s opinion is given
in relation to the answers and reactions of other participants in the group
– this cannot be noted with the use of surveys. Thus, to summarise in
regards to interviewing, there appears to be two main critiques of
qualitative-style interviews: the danger of bias and misrepresentation.
While qualitative researchers will fully admit this is a potential flaw in their
methods, that is not to say that under the right conditions, such methods
cannot be both reliable and valid when used correctly. In fact, it could be
argued that qualitative interviews are more reliable than quantitative as it
is through these methods that such problems can be identified and
rectified.

There is also debate over the use of case studies – another method used
more so by qualitative researchers. Case studies are, in essence, an in-
depth study of social phenomena from the “outside” rather than from the
“inside” (Stark and Torrance, 2005:33). That is, the researcher will rely on
a range of primary and secondary sources to come to a conclusion on an
event – pretty much like historians - thus making case studies more of an
approach rather than a research method (Stark and Torrance, 2005:33).
Famous examples include Graham Allison’s Essence of Decision:
Explaining the Cuban Missile Crisis and Carl Bernstein and Bob
Woodward’s All the President’s Men, an analysis on the Watergate scandal
that signalled the end of Richard Nixon’s presidency. And yet, inevitably,
there are criticisms regarding the validity of case studies. For instance,
Stark and Torrance comment on how there is the issue of generalisation:
it is impossible to generalise statistics based on a select on the general
population, thus questioning both whether it would be appropriate (valid)
to use such a research method and whether the results would be reliable
if used in other studies (2005:33-34). There is also the argument that case
studies, often based on the occurrence of one phenomenon, are merely
inferences “based on simple expectations of what the data would have
been had the X not occurred” (Smith, 1975:100). While these are certainly
valid arguments against case studies, they lead us on to the method that
researchers often use to mitigate these problems - triangulation.
Triangulation is the culmination of a variety of research methods in order
to extract more data from and improve the reliability of the results
(Silverman, 2000:98). As Silverman points out, triangulation can enable,
“as in trigonometry...the true state of affairs by examining where the
different data intersect. In this way, some qualitative researchers believe
that triangulation may improve the reliability of a single method”
(2000:98-99). Thus, as he points out through the scenario of a policeman
(who may use quantitative methods such as interviewing suspects and
recording arrests along with qualitative methods such as observing the
suspects and consulting documents), there can be a strong argument that
triangulating methods could be the best way to ensure valid and reliable
results. This has been justified in the real world when researchers
studying the content of political party resources employed both
quantitative methods (evidence from interviews) and qualitative methods
(the study of party election broadcasts (Harrison, 2001:110)). Yet,
Silverman acknowledges that triangulation may not always be the best
method for a researcher to use, thus indicating that there will still be a
divide between positivists and interpretists. Indeed, Robert Walker
(1985:16) mentions two arguments against triangulation. First,
researchers may struggle to determine which results to accept from his
different methods. Second, merging qualitative and quantitative research
methods may take the methods out of their proper theoretical context,
and therefore, make the methods “more of a liability than an asset”.
These arguments imply that there remains mistrust amongst positivists
and interpretists in regards to the others’ preferred choice of research
methods. However, when used correctly, triangulation methods can reap
many benefits, of which Clive Seale (1999:54) mentions four. First, the
multiple methods used produces “richer descriptions of phenomena”.
Second, team research approaches leads to the researchers working
together and voicing their concerns, thus weakening the effects of bias on
their research. Third, the range of methods leads to different theories that
in turn lead to different hypotheses – each can be compared to the data to
see how well they fit together. Fourth, and the most obvious, is the fact
that there are many possible combinations available to the researcher
engaging in triangulation – qualitative methods such as observation and
quantitative methods such as interviewing being the most common
combination. Thus, we can see that case studies, especially when merged
with other methods that can include quantitative research methods
(triangulation), do have their problems like all research methods but
certainly have the potential to produce valid and reliable results.

To conclude, it appears that qualitative research methods have been


harshly criticised. This is likely because the question has taken up the role
of a positivist, which would explain the emphasis on validity and reliability.
As we have seen, however, reliability through the eyes of an interpretist is
not much of an issue for when studying the fickle human race, change in
results should be expected. Thus, reliability is not too much of an issue
regarding qualitative research methods. The validity of qualitative
research methods was where we saw most of the debates focus on.
Through our examination of participant observation, interviewing, focus
groups, case studies and triangulation we saw that qualitative research
methods can produce more than just “authentic accounts” of human
behaviour; we saw that when used in the appropriate contexts, qualitative
research methods can produce richer quality data compared to or
combined with quantitative research methods, allow greater manoeuvre
for respondents to air their views and feelings (which can further benefit
the researcher) and, through observational techniques in particular, can
enable the researcher to make continual checks on his study to further
ensure the validity and even reliability of his results. Therefore, although
qualitative research methods have their flaws, as does every research
method, it would be too harsh to accuse qualitative research methods of
being completely useless as the question implies.

Word count: 3077


Bibliography

Texts

Harrison, L. (2001) Political Research: An Introduction. Routledge, London.


Marshall, C. and Rossman, G.B. (1989) Designing Qualitative Research.
Sage Publications, London.

Seale, C. (1999) The Quality of Qualitative Research. Sage Publications,


London.

Silverman, D. (2000) Doing Qualitative Research: A Practical Handbook.


Sage Publications, London.

Silverman, D. (2001) Interpreting Qualitative Data: Methods for Analysing


Talk, Text and Interaction 2nd Ed. Sage Publications, London.

Smith, H.W. (1975) Strategies of Social Research: The Methodological


Imagination. Prentice-Hall International, London.

Somekh, B. And Lewin, C. (eds) (2005) Research Methods in the Social


Sciences. Sage Publications, London.

Walker, R. (ed) (1985) Applied Qualitative Research. Gower Publishing


Company Ltd, Aldershot.

Periodicals

McEvoy, J. (2006) Elite Interviewing in a Divided Society: Lessons from


Northern Ireland. Political Studies Association, Vol 26 (3), 184-191.

Websites

Young, Nazi and Proud (2002), Channel 4 News. Available from:

<http://www.channel4.com/news/articles/uk/young+nazi+and+proud/257
213> [Accessed 9th November 2009, 19.07 or 7.07pm]

Lying for Jesus? (2008), Richard Dawkins. Available from:

<http://richarddawkins.net/articles/2394> [Accessed 20th November 2009,


17.34 or 5.34pm]

Consulted Texts

Patton, M.Q (1987) How to Use Qualitative Methods in Evaluation. Sage


Publications, London.

Smith, B.L, Johnson, K.F, Paulsen, D.W. and Shocket, F. (1976) Political
Research Methods: Foundations and Techniques. Houghton Mifflin
Company, London.

Strauss, A. and Corbin, J (1990) Basics of Qualitative Research: Grounded


Theory Procedures and Techniques. Sage Publications, London.
Travers, M. (2001) Qualitative Research Through Case Studies. Sage
Publications, London.

Você também pode gostar