Modified: 2007-01-31
In the early 1970s a psychologist recruits you to participate in an experiment. You are to present yourself to a mental hospital, use a pseudonym (a false name), and state that you are hearing voices (a symptom typical of schizophrenia). If you are admitted to the hospital, you are to act completely normal and give the actual details of your life (other than your real name), and no longer report hearing voices. You agree to participate, and to your great surprise, when you go to the mental hospital you are admitted quickly. Your diagnosis is schizophrenia. Once in the hospital, you act normally. In addition, you go out of your way to be friendly and cooperative with the staff. You accept the medication they give you, but you secretly dispose of it afterward (as do many of the other patients, you discover). The only strange thing you do is to take notes about your experiences. At first, you take notes covertly, but soon you realize that you may take them openly. No one on the staff is really watching you (or the other patients) closely. Some of the patients, however, ask you why you are there. They are convinced that you don’t belong in a mental hospital. The staff, on the other hand, believes that you belong there. They see your note taking as symptom of schizophrenia.
Rosenhan (1973) conducted the study just described to find out if the staff at mental hospitals could reliably distinguish between normal and abnormal behavior. The hospitals ranged from large to small, rural to urban, research to private, and old to new. He discovered was that, regardless of type of hospital, the staff did not distinguish real patients from pseudopatients. (A pseudopatient is a person playing the role of a patient. In reality, pseudopatients were confederates of the experimenter. None had a prior history of psychopathology.) Pseudopatients were admitted and hospitalized from 7 to 52 days, with an average hospitalization of 19 days. All were discharged, but not because they were cured. They were labeled schizophrenic in remission (not cured, but asymptomatic).
Rosenhan argued that labeling is a very powerful determinant of behavior in a mental health setting. Time and time again, the pseudopatients’ normal behaviors were interpreted as symptoms of schizophrenia and not as indicators of normal behavior. Once admitted, the pseudopatients (and real patients too, presumably) could do little to convince the hospital staff to change their diagnosis. Another finding was that the pseudopatients and staff had very little interaction with each other despite their close proximity. Rosenhan’s experiment is an example of participant observation, and the effect of his research was profound. Today, procedures for admission to mental health hospitals are much more stringent and the definition of schizophrenia has been revised.
Definition of Participant Observation
The participant observation technique that Rosenhan used differs from naturalistic observation in that the observer is not only part of the observed environment but also a participant in the situation being studied. Participant observation has a greater potential for researcher bias, and that problem must be addressed in the design of the research. Typically, participant observation is used when naturalistic observation would be impractical or impossible.
Examples of Participant Observation
Barber-Parker (2002) used participant observation to study nurses’ integration of teaching with bedside care. The data were collected in three phases; the first phase was a participant observation of one nurse that served as a pilot study. When Barber-Parker was satisfied that there were no methodological problems, she observed the remaining two nurses one at a time; the nurses knew they were being observed. The researcher found that teaching took place concurrently with the nurses’ other duties such as administering medication and not at a separate time. Teaching was brief and repetitive and was directed primarily to patients and only secondarily to family members.
Participant observation can take place in unusual circumstances, including jail. Elizabeth Morgan, a psychologist and surgeon, was jailed for civil contempt in a custody battle with her ex-husband over her daughter (Morgan, 1996). She was incarcerated for more than 2 years because she refused to divulge the location of their daughter. The details of the custody battle include two Acts of Congress and were described by Groner (1992). During her time in prison, Morgan kept a diary that eventually ran to 250,000 words . She described her diary as a “naïve and observational record” (p. 90) that was not kept for research purposes. After leaving prison, Morgan searched her diary for instances of dissociation by inmates and employees. (Dissociative disorders involve the presence of multiple personalities and severe memory problems.). She found 40 such instances. She grouped these into five categories: named alters (named alters are multiple personalities with their own names that take control of a person), trances or sudden switches in personality, flashbacks, switches to childlike speech, and deliberate use of dissociation. However, she did not diagnose any of the women (neither inmates nor employees) as being psychotic. Nor did she believe that illegal drug use, which was common, was associated with the dissociation she observed. She concluded that the inmates’ dissociative behaviors contributed to prison employees’ absenteeism. Morgan’s study illustrates how participant observation can take place in unusual circumstances.
Conducting Participant Observation Research
Participants Unlike naturalistic observation, only humans can be the objects of participant observation research. (We do not regard efforts to communicate with animals using symbols as participant observation.)
Apparatus Most important in the choice of apparatus is whether or not the participants know they are under observation. If they are unaware, data collection requires subtle methods. In such situations, researchers are limited to covert data collection techniques. For example, some cell phones allow note taking and could be used surreptitiously. If participants know they are being observed, then recording equipment can be used openly.
Procedure The contexts of participant observation can vary widely. Rosenhan’s pseudopatients secured admission to different types of mental hospitals. Barber-Parker used the hospital in which she worked as her research setting. Morgan used her involuntary incarceration as an opportunity to conduct participant observation research. Entry into the group to be observed can be more problematic in participant observation than with naturalistic observation. For groups that are difficult to enter, such as motorcycle gangs, the success of the research depends on becoming a member. Once entry is secured, observations start. As in naturalistic observation, researchers conducting participant observations have to deal with reactivity if the participants know they are under observation. In those situations, researchers must ensure that their presence is not affecting the behavior of the observed. When participants are unaware that they are being observed, reactivity is less of an issue.
Results In participant observation research, the data are handled in ways similar to those in naturalistic observation. The data that are generated can be quantitative, qualitative, or both.
Ethics Ethical issues can loom large in participant observation research. For his dissertation research, Quinn (1987) joined a motorcycle gang but eventually had to remove himself because of his potential involvement in criminal activities if he remained in the gang. Researchers cannot use participant observation methods as an excuse for engaging in unethical or criminal activity. Also, participant observers may or may not have a commitment to the group they join. In the example above, Rosenhan’s pseudopatients were not committed to the groups they were observing. On the other hand, Barber-Parker did have a commitment to the nurses that she was observing. In participant observation, researchers must first decide what their relationship will be to the group under observation. Issues such as commitment and whether or not the participants will know that they are under observation should be approved by an IRB before a participant observation project begins. Now we turn our attention to research involving qualitative research, a rapidly developing field within psychology.