 |
"Out
Standing in the Field:" Essentials of Pariticipant Observation
Yeah, yeah -- bad joke, I know! I couldn't
resist: it was the "punch line" on an adorable coffee mug given to me
by a doctoral student colleague right after my prospectus hearing!
I. A Delicate "Balancing Act:" Participant &
Observer
Although these two terms are indeed used jointly to characterize
this particular method of qualitative data collection, they carry
some subtle contradictions as well. Namely, here is the delicate balance
to be maintained between the two:
Figure 1.
Tradeoffs of Being a "Participant" and "Observer"
Start thinking about what is involved in observation by opening and reading
the link below.
On
Observation
* The "participant part" implies an immersive experience
in a real-world, "field-based" setting. As such it implies that
the researcher is holistically committing his/her feelings, thoughts,
emotions, etc. to that setting.
* On the other hand, the "observer part" is characterized
by the scientific approach to knowledge. That implies objective,
scientific, neutral, scholarly recording of these data or observations.
At first glance, it seems like a rather precarious balancing
act, doesn't it?
* Go too far in the "participant" extreme and the researcher
runs the risk of, as Miles and Huberman characterize it, "going
native." He/she can in essence become "co-opted" by the situation,
setting, key players, etc., to such a degree that scientific objectivity
is lost. Michael Quinn Patton describes such a study. The researchers
were observing and recording data on audience reactions to a crusade
being conducted by the Rev. Dr. Billy Graham. At one point, towards
the end of the crusade, two of the researchers put down their note
pads and walked down the aisle to join those audience members who
had accepted Dr. Graham's call.
* Go too far in the "observer" direction and you run the
risk of doing what I call "overly sterile qualitative research!"
By that I mean skimping on the "rich, thick, vivid description" that
we have learned is the hallmark of qualitative research.
For some of us, this danger is admittedly a "spillover" from
the old quantitative paradigm: say it concisely -- p-value of X, reject
the null hypothesis, etc.The qualitative equivalent would be to record,
"The support group meeting took place in a school," and think you
have "adequately" characterized the setting. But think of it this
way: would you and I necessarily conjure up the same mental
picture when we read the word, "school?" You might picture a large
room with desks in straight rows in a city surrounding. Someone else
might envision a laboratory-type setting with round tables that seat
at most 5 or 6 students, with available tactile learning materials.
Someone else could easily imagine a home school environment. Plus,
I hear that there are still some vestiges of the "one-room schoolhouse"
in existence! Hope this gets the point across! and the resultant need
for "rich, thick description!"
For other qualitative researchers, the danger here is in how
the field notes are recorded -- "jumping to conclusions," e.g., "the
group seemed friendly." Again, it comes down to providing detailed,
vivid narrative as to the evidence for this. How did you arrive at
that conclusion? Did they smile and make frequent eye contact with
one another? Was there ready agreement and support for one another's
comments and positions? Did they pull their chairs close together
and sit in a circle as they talked? For that matter, how did they
even greet one another as they came in? Handshakes? Hugs? Pats on
the back or arm? And so on and so forth.
Stop a minute and read this anthropologists mixed-method (multi-method)
field work study conducted in New Guinea.
Participation
Despite the danger in going to either extreme, it is possible to
strike a balance and have the best of both worlds. That's what I've
tried to show by the intersection of the 2 circles on pg. 2. I believe
that balance lies in the following:
- Capitalizing on your opportunity for a "closeup view" by being
in the field, by recording detailed observations of who/what/where/etc.
As previewed on Page 1, we'll talk more in depth shortly about "what
to record." For now, it's making the most of your immersion in the
field by extensively documenting as much as you can about what you
see.
- At the same time, following principles of "good" research design
and analysis. In other words: this one "keeps you honest as a scientist"
and thus helps to guard against losing your objectivity. This
would include, for instance, things like having a partner-observer
against whom you can bounce off your own observations -- really
a reliability and validity check. It also includes use of aids such
as tape recorders, video-tapes in some cases, photographs, sketches,
notebooks, etc., to help ensure accuracy (vs. "hearing what you
want to hear"). Finally, it can include something called a "member
check". In a nutshell, a "member check" means taking back
a summary of your findings and conclusions back to key informants
in the field for a reality check. You are in essence asking,
"Did I, as an outsider, capture your 'lived experience' accurately?"
In general, having a well thought out research plan, along the lines
of the "pieces" we've been learning here and in Research Design
(e.g., Methodology; Population and Sampling; Instrumentation Construction
and Validation; Data Collection Procedures and Data Analysis Procedures)
helps to remind you that there are such research "rules" to play
by,regardless of whether your study is quantitative, qualitative,
or both - and thus helps guard against your own (accidental) bias,
subjectivity, etc.
Open this link to look at an example of a participant-observer case study.
The Culture Of Military Organizations: A Participant-Observer Case
Study
Of Cultural Diversity
II. Relative Advantages of Field-Based Participant Observation
"Real" means "believable" and perhaps also "more generalizable."
The classic experimental paradigm would bemoan the fact that you are out
there "with no control and bundles of variables." Yet, at the same time,
what better way to accurately replicate "true reality" in your study?!
Odds are that because you are so closely observing and recording "real"
behaviors and conditions, your findings and conclusions will inherently
carry more credibility and relevance than those of the rigidly controlled,
highly unrealistic laboratory. Some research design books make the point
this way:
- Experimental designs, due to the high
degree of control of extraneous variables, manipulation of one or
more experimental conditions, etc., possess high internal validity.
You can more precisely say "when I change X, then Y will change"
and thus identify pure cause-effect relationships under such conditions
-- because most, if not all, potential contaminants have been explicitly
and rigidly controlled. In other words, we say they have high
internal validity (credibility/believability of asserted findings).
Yet, while applauding this ability to zero in
on strict cause-effect relationships, researchers and especially
practitioners have retorted,"That is all well and good, but my classroom
(clinic, counseling center) is hardly a laboratory! Things happen
that I simply can't control in the classic experimental sense --
life happens, you know?! So -- what's the practical usefulness of
your findings and results to my 'real world mess?!'" In other words,
low external validity (generalizability)!
- Field-based, "naturalistic" studies, on
the other hand, are inherently so "variable-laden-and-messy" --
anything goes! You take what you get when you throw yourself in
there in the field! -- that it is virtually impossible to carefully
tease out those rigid X - > Y causal relationships. So we would
characterize them as possessing relatively low internal validity.
At the same time, though, due to this same realism, it is more likely
that others with similar real-world conditions will be able to more
readily apply the corresponding findings and results of our naturalistic
investigation. Thus, the tradeoff here is greater external validity.
Please open this link to look at very interesting example of a "naturalistic"
study done on the web. It is a participant-observer study on the Xenaverse,
"the online spaces devoted to the cult following of the syndicated
television program "Xena, Warrior Princess.
The Ballad of the Internet Nutball: Chaining Rhetorical Visions
from the Margins of the Margins to the Mainstream in the Xenaverse
(Rensselaer Polytechnic Institute doctoral dissertation)
Perhaps the best way to summarize this advantage, then, is
to cite yet another of our qualitative gurus, Robert K. Yin. He
has reminded us that when you are doing qualitative research,
you are really studying two things:
i. The phenomenon/phenomena -- e.g., your target
variable(s) of interest, such as for Dee Dee Nevelle of Sierra
Vista: "the factors that constitute 'good climate' in a school"
ii. Embedded in context -- this is so important! Unlike
the sterile laboratory, you might think of this as "but those
variables go with the territory!" Like it or not, Dee Dee was
observing and identifying 'good climate' variables in this
school, with this staff, in this town, during this
time period, and so forth.
To sum up, the immersive and carefully balanced
scientific aspects of field-based participant observation
let us have a good, closeup view of both of the above features
which go hand in hand with any qualitative investigation.
- Such "close-up immersive experiences" also make
it easier for the researcher to do "classic" grounded-theory exploratory/descriptive
types of research. In other words, he/she knows it's OK to come
to the setting without any prior theoretical models in place. By observing
carefully and in detail, he/she can thereby allow potential important
variables or factors to 'emerge naturally' out of such an immersion.
Observation gives you more of a closeup look at these, and thus the
opportunity for such 'eurekas' of 'pop-through' variables, than would,
say, a one-shot, impersonal administration of an open-ended survey to
a group of sample subjects. While both are qualitative in nature, the
latter does not allow as much time, closeness, or opportunity to 'see
and discover' the potential explanatory variables in quite the same
way.
- The participant-observer-researcher may bring
"fresh eyes" into the situation and thus notice potentially important
aspects of the situation that subjects who are immersed in it might
miss or take for granted. This is the old "fish out of water" phenomenon.
Because of your 'newness' to the setting, you may more easily pick up
on little nuances that have become comfortable old habits - and thus
sublimated to the subconscious - to the subjects in the field. We've
all experienced this phenomenon to some degree, I think. Silly example,
but I hope it makes the point: I was in a large-group meeting once where
the temperature of the room was rather cold. I say it that way because
most of us somehow got subconsciously acclimated to it. Perhaps our
coping mechanisms included anything from the frequently refilled coffee
pot, to the "heat" generated by animated discussion regarding the issues
at hand! Anyway, we had a colleague join us midday; and what laughter
and surprise filled the room when that colleague instantly remarked,
"Why, this room is so cold you could hang meat in it!" Seriously, things
like trust levels; "underground grapevines" for communication; automatic
deference to one or more individuals in a group decision setting; etc.,
etc. may be so entrenched in the subjects' lived experience that they
don't even notice it, or would even think to tell you. Yet, with the
'fresh eyes' you bring to the situation, you may spot it outright. And
if it somehow critically relates to the phenomenon/phenomena you are
there to observe -- and I'll bet in most cases, it does -- then you
have just made an important discovery.
- Related to the preceding point, there may be
things that subjects are aware of, but yet are simply unwilling to talk
about in an interview. Perhaps you can observe some or all of these
through visual and other manifestations by immersing yourself in the
field. Also, the interview setting may not allow enough time for such
"iceberg-surface-level" yet potentially explanatory issues to emerge.
- Finally, here's a rather interesting twist on the
ol' "concern for researcher bias" that is part and parcel of classic
quantitative designs. Mind you, I'm not saying that isn't a legitimate
concern! If anything, our earlier discussion as to the dangers of "too
participant extreme" warn against sacrificing scientific scholarly objectivity.
But here's perhaps a hidden benefit to having the researcher's own observations
and yes, feelings, as a cross-check. Please keep in mind that the
perceptions of subjects, are just that - perceptions. As we know all
too well, perceptions do not necessarily match reality! Also, as
reviewed in Point 3 of this list, above, those participants may honestly
not "perceive" some aspect of a situation and thus not be aware of it,
due to their extensive immersion. And as Point 4 has told us, they may
be uncomfortable, embarrassed or otherwise unwilling to share candidly
in some cases, even if they are aware. So here again, rather than
being a 'bias,' you as an outsider may bring 'fresh eyes' and a 'clean'
perspective with which to more accurately 'capture' a wider range of
field-based variables and factors.
Once you have finished you should:
Go back to Participant and Observer
E-mail M. Dereshiwsky
at statcatmd@aol.com
Call M. Dereshiwsky
at (520) 523-1892
Copyright © 1999
Northern Arizona University
ALL RIGHTS RESERVED
|
