Journal of Nursing Education

BRIEFS 

Use of Behavioral Observation to Augment Quantitative Data When Studying Clinical Teaching

Elizabeth J Pugh, PhD, RN

Abstract

Research on teaching and learning has been conducted most frequently in the classroom setting. The formality and control maintained by the teacher in the classroom permit many variables to be manipulated by researchers. Those who study teaching and learning in non-classroom settings, on the other hand, find their work difficult in many and in unanticipated ways.

This paper presents methodological problems encountered by a researcher studying beliefs, intentions, and behavior of teachers of clinical nursing in clinical settings. The theoretical rationale for the study (Ajzen & Fishbein, 1980) and the results are available elsewhere (Pugh, 1980); a brief description is nevertheless necessary in order to provide context for a description of the researcher's experiences.

Faculty (n=50) and their respective students (n=358), randomly selected from eight midwestern undergraduate programs, completed questionnaires individually administered by the investigator. Both faculty and student questionnaires dealt with beliefs about 20 teaching behaviors which had been reported to be helpful and which are directly observable. Consents to participate were signed by all subjects.

Faculty rated their evaluation of the importance of each behavior, the probability of its use, their perception of the probability with which their faculty peers expect them to use the behavior, the probability with which their students expect them to use the behavior, their own attitude toward the behavior, and the perceived attitude of their faculty peers. Students rated the importance of each teaching behavior and the frequency with which their clinical instructor used each behavior. Each group of students was told that the data were to be used for research purposes only, not for evaluation of faculty.

Each faculty member was observed for one entire clinical day as he or she taught in a patient care setting, e.g., clinic, hospital, patient's home, nursing home. For this sample, a total of 84 different patient care settings in 35 different institutions in 14 cities and town in two midwestern states were used. (It should be clear that one problem encountered by the investigator was that of planning, scheduling, and traveling to each clinical setting!)

Observation rather than self-report or other techniques was used to document and describe patterns of teaching behavior among nursing faculty for several reasons. First, it provided an objective description of behavior and events not possible through self-report. A recent review of research (Hook & Rosenshine, 1979) disclosed that one cannot assume that teachers' reports of their classroom behavior correspond to actual practice. It is unlikely that teachers of clinical nursing would be any more accurate in reporting their own behavior.

Secondly, observation provided an opportunity to study the contextual background of teachers' behavior. Self-report would not have taken account of the content of the clinical setting in which behaviors were enacted (Aamodt, 1983).

Finally, the teacher-student-patient situation is not reproducible in academic settings. Complex and multidimensional psychosocial phenomena, such as teacherstudent interactions, are best studied through the use of observation in natural settings (Bogdan & Taylor, 1975; Weick, 1968).

The observational method utilized was that of observer as passively present, having limited interaction with the persons observed. The observer was introduced as a nurse-teacher or a nurse-researcher who was studying the types of situations encountered in clinical teaching of nursing. This explanation for the observer's presence was clearly presented to patients and personnel as well as subjects, in order to minimize their uncertainty and suspicion (Lofland, 1971; Weick, 1968).

Problems in studying clinical teaching are described as they were encountered at different points in the research process. Entry into academic settings and sample selection will be briefly addressed, followed by procurement of informed consent when there was a need for…

Research on teaching and learning has been conducted most frequently in the classroom setting. The formality and control maintained by the teacher in the classroom permit many variables to be manipulated by researchers. Those who study teaching and learning in non-classroom settings, on the other hand, find their work difficult in many and in unanticipated ways.

This paper presents methodological problems encountered by a researcher studying beliefs, intentions, and behavior of teachers of clinical nursing in clinical settings. The theoretical rationale for the study (Ajzen & Fishbein, 1980) and the results are available elsewhere (Pugh, 1980); a brief description is nevertheless necessary in order to provide context for a description of the researcher's experiences.

Faculty (n=50) and their respective students (n=358), randomly selected from eight midwestern undergraduate programs, completed questionnaires individually administered by the investigator. Both faculty and student questionnaires dealt with beliefs about 20 teaching behaviors which had been reported to be helpful and which are directly observable. Consents to participate were signed by all subjects.

Faculty rated their evaluation of the importance of each behavior, the probability of its use, their perception of the probability with which their faculty peers expect them to use the behavior, the probability with which their students expect them to use the behavior, their own attitude toward the behavior, and the perceived attitude of their faculty peers. Students rated the importance of each teaching behavior and the frequency with which their clinical instructor used each behavior. Each group of students was told that the data were to be used for research purposes only, not for evaluation of faculty.

Each faculty member was observed for one entire clinical day as he or she taught in a patient care setting, e.g., clinic, hospital, patient's home, nursing home. For this sample, a total of 84 different patient care settings in 35 different institutions in 14 cities and town in two midwestern states were used. (It should be clear that one problem encountered by the investigator was that of planning, scheduling, and traveling to each clinical setting!)

Observation rather than self-report or other techniques was used to document and describe patterns of teaching behavior among nursing faculty for several reasons. First, it provided an objective description of behavior and events not possible through self-report. A recent review of research (Hook & Rosenshine, 1979) disclosed that one cannot assume that teachers' reports of their classroom behavior correspond to actual practice. It is unlikely that teachers of clinical nursing would be any more accurate in reporting their own behavior.

Secondly, observation provided an opportunity to study the contextual background of teachers' behavior. Self-report would not have taken account of the content of the clinical setting in which behaviors were enacted (Aamodt, 1983).

Finally, the teacher-student-patient situation is not reproducible in academic settings. Complex and multidimensional psychosocial phenomena, such as teacherstudent interactions, are best studied through the use of observation in natural settings (Bogdan & Taylor, 1975; Weick, 1968).

The observational method utilized was that of observer as passively present, having limited interaction with the persons observed. The observer was introduced as a nurse-teacher or a nurse-researcher who was studying the types of situations encountered in clinical teaching of nursing. This explanation for the observer's presence was clearly presented to patients and personnel as well as subjects, in order to minimize their uncertainty and suspicion (Lofland, 1971; Weick, 1968).

Problems in studying clinical teaching are described as they were encountered at different points in the research process. Entry into academic settings and sample selection will be briefly addressed, followed by procurement of informed consent when there was a need for partial concealment of purpose, with most of the discussion focusing on problems during data collection.

Access to Subjects

Gaining access to subjects in eight academic institutions involved what often turned out to be lengthy approval processes. Variations in what was required in order to obtain such approval prevented the investigator from enclosing a reasonable and standard amount of materials with the initial letter sent to each dean or program director. Some deans granted approval themselves; others referred the request to a departmental research review committee; and still others required multiple copies of the full proposal and instruments to be submitted to their own Institutional Review Board (IRB) in spite of federal funding and the approval of the investigator's IRB. This process was being attempted during the famous Chicago winter of 1979: mails were slow, people were absent from work, and committees were not meeting regularly.

Fortunately, approval at the level of each clinical setting was not required. Since the investigator was studying teacher behavior, access to subjects and approval of research activities could be granted by the educational institution rather than the multiple clinical settings. Faculty were asked to merely inform their head nurses of the anticipated visit of the researcher and to assure them that patients were not the subjects of the research.

Once institutional access to subjects was obtained, procuring complete lists of eligible faculty for random selection of subjects by the investigator proved to be an incredibly and unexpectedly difficult task. Sometimes a list of volunteers was provided instead of all those eligible. It was necessary to use extreme tact with deans and program directors when it became evident that they did not understand random sampling and, therefore, the reason for the investigator's need for a complete list of eligible faculty

Informed Consent

Obtaining informed consent while maintaining partial concealment of purpose raised ethical questions in the investigator's mind. Partial concealment of purpose was necessary, for if faculty had known the purpose of the study - · to determine factors influencing congruence between their beliefs, behavioral intentions, and behavior - bias would certainly have been introduced and behavior would have been changed. Additionally, it would have been difficult, if not impossible, to obtain permission to ask students to rate the frequency of faculty members' use of the teaching behaviors. However, within the context of the student questionnaire, which asked for students' beliefs about the importance of the behavior, the frequency ratings of those behaviors seemed to be unnoticed by faculty. Observed behavior verified and explained quantitative data obtained by the questionnaires.

Each subject was debriefed at the end of the data collection period. Nearly all of them commented that had they known the full purpose, they would have certainly been more aware of their behavior and the observer, and would have perhaps tried to do what they thought they had indicated in the questionnaire as important.

Credibility of the Researcher

Some of the advantages and disadvantages of the nurse-teacher in a researcherobserver role in the health care setting were experienced. Use of a nurse-observer in a study of nursing education offered the advantage of familiarity with the setting. It was possible to select appropriate apparel, e.g., white uniform complete with nurse's cap, lab coat over dark clothing, scrub attire, which would contribute to becoming unobtrusive. It was also helpful to have an observer who could understand the context of the patient situations encountered by subjects.

A nurse-observer who was also an experienced educator offered further advantages in terms of sensitivity to the range of situations encountered by nursing faculty. Faculty did not need to explain what they were going to do next in great detail; they simply went about their activities with, as one subject commented, an "extra faculty member observing" their clinical teaching. That they forgot they were being observed was not only expressed verbally but also manifested behaviorally by two faculty who became angry and swore at their students. At this point the observer felt that she had achieved her goal of becoming non-threatening and unobtrusive.

Familiarity with the setting and identification as a nurse also facilitated acceptance by patients. The majority of teacherstudent interactions took place in the patient's room, usually at the bedside. A non-nurse may not have been able to tolerate some of the situations, or may not have been welcomed by the patients when the researcher was recording teacher-student interactions during nursing procedures.

Staying in Role

In spite of familiarity, comfort, and perceived acceptance within the setting, some problems were encountered. Staying in the role of researcher was perhaps the most persistent problem. The investigator was invited - from her point of view, tempted - to act as teacher, nurse, colleague, and consultant. Such invitations were attractive since she was comfortable in all of these roles.

It was difficult to stay in the role of researcher in situations where inexperienced faculty were either giving erroneous information to students or failing to recognize a student's need for support or assistance. It seemed so natural to want to respond to the students' questions, supply additional (correct.') information to that given by the teacher, or ask further questions to see if the student really understood what was taking place in a particular patient's situation.

The faculty in this sample tended to question extensively to assess understanding of physiology, pharmacological action, and rationale for selection of particular nursing interventions, but they gave relatively little positive feedback and formative evaluation of students' performances. The investigator's natural and overwhelming desire to assume the teacher role was further increased when students used eye contact and other non-verbal behavior to invite her to join the interaction with their facuity.

It was even more frustrating to be unable to respond as a nurse to the needs of patients or families when immediate intervention was indicated. Fortunately, the faculty member being observed usually responded in such situations by asking the researcher to assist in some manner. For example, a fainting patient of abundant proportions who was precariously situated between a wheelchair and her bed required all three of us to get her safely back into bed.

Debriefing

Faculty expressed a need for affirmation of the appropriateness of their teaching behavior; over half of the sample asked for evaluative feedback during the debriefing session. The investigator stated that the purpose of the observation had not been to gather data for evaluation, nor could she establish the effectiveness of the teacher's behavior, but that she could briefly describe what she had observed. The teacher could then decide if that was what she thought she was doing or had planned to do, placing her own value upon the behavior described.

Extreme tact and consideration of the subject's self-esteem was required when it was evident to the researcher that the faculty member requesting affirmation also needed considerable improvement in her clinical teaching. It was much easier to provide requested consultation during the debriefing period regarding management of specific student learning problems which had been encountered during that clinical day.

Advantages of Behavioral Observations

In this study of clinical teaching, use of self-reports, student ratings of teacher behavior, and behavioral observations were ail found to be useful in providing a picture of the clinical teaching of nursing. Each provided important data and each made a distinctive contribution. Most of the problems discussed so far are the result of the collection of behavioral measures, but there are several important benefits.

It would not have been possible, for instance, to have discovered that patterns of nursing faculty behavior exist if only self-report and student ratings had been utilized. Three patterns of faculty behavior were observed: Nurse, Teacher, and NurseTeacher. Nurses enact primarily nurse behavior, e.g., interact with patients, literally elbow the student aside and take over at the bedside to deliver nursing care. Teachers enact teacher behavior, e.g., determine if the student is prepared to perform a procedure and feeling comfortable about it, evaluate and give suggestions for improvement; and Nurse-Teachers appear comfortable enacting both roles appropriately. The majority of faculty in this sample identified themselves as "nurses who teach nursing," but this role identification did not in itself predict the observed patterns of behavior. However, when combined with role preparation and value placed on teaching or clinical practice, it was more closely related to observed behavior.

Summary

Observation of behavior to study clinical teaching provides advantages over the use of questionnaires alone. The great advantage is the richness of the data. Problems encountered in the process, e.g., gaining access to subjects in multiple institutions, selecting non-volunteer subjects, obtaining informed consent in a manner which will not bias the data, becoming unobtrusive and non-threatening to subjects, coping with the unpredictability of the clinical setting, and staying in the role of researcher in spite of temptations to enact other roles within the setting, are all foreseeable and to some degree preventable. Being aware of the potential obstacles permits the investigator to plan for them and to be less emotionally aroused as each new problem arises. Other nurse researchers are encouraged to utilize these techniques; the experience can be not only valuable in terms of realistic data obtained, but also can be an extremely rewarding experience.

References

  • Aamodt, AM. (1983). Problems in doing research: Developing a criteria for evaluating qualitative research. Western Journal of Nursing Research, 5, 398-402.
  • Ajzen, L, & Fishbein, M. (1980). Understanding Attitudes and Predicting Social Behavior Englewood Cliffs, NJ: Prentice-Hall.
  • Bogdan, R., & Taylor, S. J. fl975). Introduction to Qualitative Research Methods. New York: John Wiley & Sons.
  • Hook, CM., & Rosenshine, B.V. (1979). Accuracy of teacher reports of their classroom behavior. Review of Educational Research, 49(1), 1-12.
  • Lofland, J. (1971). Analyzing Social Settings. Belmont, CA: Wadsworth.
  • Pugh, E.J. (1980). Factors influencing congruence between beliefs, intentions, and behavior in the clinical teaching of nursing (Doctoral dissertation. Northwestern University, 1980). Dissertation Abstracts International, 41(6), 2521A-2522A. (University Microfilms No. 8026902)
  • Weick, K.E. (1968). Systematic observational methods. In G. Lindzey & E. Aronson (Eds.). The Handbook of Social Psychology (2nd ed.) (Vol. 2>. Reading, MA: Addison- Wesley.

10.3928/0148-4834-19861001-09

Sign up to receive

Journal E-contents