Journal of Nursing Education

The articles prior to January 2012 are part of the back file collection and are not available with a current paid subscription. To access the article, you may purchase it or purchase the complete back file collection here

Measuring Perceptions of the Exemplary Nurse

H A Llewellyn-Thomas, RN, PhD; N Sims-Jones, RN, MScN; H J Sutherland, RN

Abstract

ABSTRACT

Qualitative and quantitative data identifying and weighting the perceived importance of different attributes of the "exemplary" nurse were collected from nursing respondents at different points in their professional development. Statistically significant differences in the attribute weights reported by fourth-year students and graduate nurses were observed. Because their pool was qualitatively different, quantitative comparisons with first-year students were not possible. The results indicate that these methods may be useful in examining the processes whereby the members of different groups (eg, nursing students and faculty) formulate and modify their assumptions about their own and others' professional values.

Abstract

ABSTRACT

Qualitative and quantitative data identifying and weighting the perceived importance of different attributes of the "exemplary" nurse were collected from nursing respondents at different points in their professional development. Statistically significant differences in the attribute weights reported by fourth-year students and graduate nurses were observed. Because their pool was qualitatively different, quantitative comparisons with first-year students were not possible. The results indicate that these methods may be useful in examining the processes whereby the members of different groups (eg, nursing students and faculty) formulate and modify their assumptions about their own and others' professional values.

Introduction

Current methodologie work at the Ontario Cancer Institute involves developing ways to assess cancer patients' values (Llewellyn-Thomas et al, 1984a, 1984b; Llewellyn-Thomas & Sutherland, 1987; LlewellynThomas, Sutherland, Tibshirani, et al, 1982) using approaches from social judgment theory (Hammond, Stewart, Brehmer, et al, 1975). The successful extension of these techniques into the assessment of health professionals' values could not only further this methodologie contribution, but also provide valuable insights into the processes whereby these values are acquired and guide professional behavior. This article describes how these methods were applied to determine how nurses at different stages of professional development identify and weight the relative importance of the attributes of the "exemplary" nurse in clinical practice.

Studies that have attempted to determine how nurses ascribe weights to different characteristics of the ideal (or exemplary) nurse have been handicapped by different problems of conceptualization and design. Corwin's (1961) and Habenstein & Christ's (1985) early conceptualizations served as the underpinnings for a number of studies in which student (Davis & Oleson, 1964; Jones & Jones, 1977; Whelen, 1984) and graduate (Benner & Kramer, 1972; Kramer, 1970; Minehan, 1977) nurses reported their values and behaviors relative to presupposed models of the role of the nurse. The respondents in these studies did not themselves identify the salient aspects of the judgment problem. Kelly (1983) used open-ended questions and Likert scales to assess role perceptions and ideal expectations of first- and final-year nursing students, but the results generated by the two different approaches were not linked systematically. Heyman, Shaw, and Harding (1983) used an open-ended personal construct theory approach in a longitudinal study of nursing students, but the assessment was carried out in terms of the attraction of nursing relative to alternative occupations.

The method used here attempted to overcome these limitations by asking respondents to generate a pool of qualitative items that were subsequently grouped and quantitatively weighted by the same respondents. Furthermore, the study design involved different groups of respondents who were at different points in their professional development and could be expected to report either different attributes or different weights for the same attributes. Such evidence of sensitivity would contribute toward the argument that this approach provides valid results.

Method

General Plan

The general strategy involved three phases. In Phase 1, lists of the characteristics of the exemplary nurse were elicited from University of Toronto Faculty of Nursing students enrolled in the first year theory and fourth year nursing research courses. Approximately 100 students were enrolled in each course.

During Phase 2, the lists were used to identify and define attributes which in turn were used to construct ranking and rating tasks designed to quantify the relative importance of each attribute.

In Phase 3, these ranking and rating tasks were carried out by the same groups of nursing students. As well, graduate nurses attending an annual meeting of the Registered Nurses Association of Ontario were invited to respond to the tasks that had been developed for the fourthyear nursing students.

Procedures

Preliminary Steps. The nursing students were provided with a brief explanation of the framework of the study, the tasks involved, and the time commitment required. Emphasis was placed on the voluntary nature of participation. Students in each course were then invited to participate in the two data collecting sessions, which were one week apart and were to take place at the close of regularly scheduled classes. It was expected that each session would last approximately 15 minutes.

Phase 1. During the first data collecting session, students were asked to respond, in writing, to the question: What do you think are the characteristics of the exemplary nurse in clinical practice?

Phase 2. During the intervening week, the responses of each class were assessed to determine whether the characteristics tended to cluster within separate overall attributes. To determine the face validity of the resulting general attributes, two raters independently categorized each response into the postulated attributes, and interrater reliability coefficients were computed. When indicated, the attributes were re-defined and the categorization task was repeated until an inter-rater reliability coefficient of 0.90 was achieved.

For each class, seven main attributes emerged. The skillful application of a sound knowledge base was a clearly dominant attribute in both classes. The three attributes into which most of a class's remaining responses had been categorized were used to construct a set of eight multiattribute scenarios describing different hypothetical nurses. Thus, two sets of scenarios (one set per class) were prepared; each set used different attribute definitions because the fourth-year student responses had reflected greater sophistication compared with those submitted by the beginning students. In both sets, the knowledgeable/skillful attribute was presented in these paragraphs as a given.

Phase 3. During the second data collecting session, each class member ranked the seven main attributes generated by that class and then rated the accompanying set of scenarios. An alphabetical list of the seven main attributes of an exemplary nurse was presented. Each participating student was asked to rank the list in terms of importance, 1 indicating the most important attribute and continuing to 7, indicating the least important.

Table

TABLE 1FACTORIAL DESIGN IN WHICH THE THREE ATTRIBUTES ARE COMBINED

TABLE 1

FACTORIAL DESIGN IN WHICH THE THREE ATTRIBUTES ARE COMBINED

The students were then asked to rate the acceptability of each of the eight scenarios describing nurses. Before proceeding with this rating task, students were provided with three kinds of information: a statement emphasizing that the knowledgeable/skillful attribute was a basic assumption; their class's definitions of the remaining three most frequently cited attributes; and a definition of the word "exemplary."

The scenarios were constructed using a factorial design, in which acceptable and exemplary versions of each of the three attributes were combined in all possible ways (Table 1). The two extreme scenarios-exemplary (assigned a value of 100) and acceptable (assigned a value of O)- were used to define the ends of a linear analogue scale. The remaining six scenarios, randomized in sequence, were presented together with two replicates. The students were asked to rate each scenario relative to the defined exemplary and acceptable anchors by placing a mark across a 10 cm scale at the position that indicated their opinion. Raw scores were obtained by measuring the distance from O to the mark on the scale (Figure 1). Data on age and professional education were also collected.

During the poster session of the 1986 annual meeting of the Registered Nurses' Association of Ontario, graduate nurses were invited to rank the seven main attributes and rate the scenarios that had been developed for the fourthyear student nurses. Data about age, educational status, and years spent in bedside nursing were also collected.

Statistical Analyses. Statistical tests were carried out using Minitab Data Analysis Software (Minitab Inc., University Park, PA). The data obtained from the ranking task were examined using Kruskal-Wallis and Mann-Whitney tests. The internal consistency of the rating task was determined by computing Pearson product moment correlation coefficients for the raw scores assigned to the replicated scenarios.

FIGURE 1USING THE LINEAR ANALOGUE SCALE TO RATE SCENARIOS

FIGURE 1

USING THE LINEAR ANALOGUE SCALE TO RATE SCENARIOS

The raw scores for each scenario were used in ordinary least squares regressions to estimate the weights implicitly assigned to each of the three attributes by each of the respondents. The least squares estimators were based on the assumption that the scores for the extremes of the scales were fixed at 100 and O for the exemplary and acceptable extremes, that all of the other scenario scores, including the replicates, were subject to random error with the same variability, and that the weights for the three attributes were additive and summed to unity. Group differences in attribute weights were analyzed using pooled i-tests, when appropriate.

Results

Subjects

Table 2 describes the sample sizes, ages, and educational status of the three groups of nurses. The average number of years spent carrying out bedside nursing by the graduate nurses is included. There is a high proportion of universityeducated nurses in the graduate sample.

The Ranking Task

The first-year students generated an average of four characteristics, whereas the fourth-year class contributed an average of six. For each class, these characteristics appeared to fall into seven main attributes: collaborative ability; empathy with patients; knowledgeable/skillful caregiving; leadership ability; ability to act as an advocate for the patient and for the profession; and personal integrity.

Table 3 indicates the median ranks for each attribute according to group. There were significant differences in the median rank order of attributes among the three groups. First- and fourth-year students differed in their ranking of personal integrity and profession advocate; both were more important for fourth-year students. Note that although the medians were the same for both groups for the profession advocate attribute, the distributions of data were significantly different.

Table

TABLE 2AGE, EDUCATIONAL STATUS, AND CLINICAL EXPERIENCE OF STUDY PARTICIPANTS

TABLE 2

AGE, EDUCATIONAL STATUS, AND CLINICAL EXPERIENCE OF STUDY PARTICIPANTS

Four attributes were ranked in a different order by the fourth-year students when compared with the graduate nurses. As a group, fourth-year students assigned relatively more importance to the knowledgeable/skillful and patient advocate attributes and less to the personal integrity and profession advocate attributes.

Finally, significant differences were seen between firstyear students and graduate nurses in their rank ordering of the following attributes: knowledgeable/skillful and patient advocate were considered more important by the first-year student group, whereas graduate nurses assigned more importance to the personal integriy attribute.

In summary, with increasing nursing education and experience, the emphasis on being knowledgeable/skillful and acting as a patient advocate seemed to lessen while the demonstration of personal integrity and a commitment to the profession appeared to gain in relative importance.

Table

TABLE 3MEDIAN RANKS FOR SEVEN MAIN ATTRIBUTES FOR EACH RESPONDENT GROUP

TABLE 3

MEDIAN RANKS FOR SEVEN MAIN ATTRIBUTES FOR EACH RESPONDENT GROUP

The Rating Task

As noted earlier, both the first- and fourth-year student groups most frequently cited characteristics associated with the skillful application of a sound knowledge base. For both groups, the remaining characteristics tended to cluster in the same three general attributional areas of patient advocate, profession advocate, and collaborative ability. However, the terms used by the two classes indicated that the first-year students defined these attributes much more narrowly than the fourth-year students. The qualitatively different definitions of these attributes are presented in Figure 2; these were used to construct the two sets of rating task scenarios.

The means and standard deviations for the raw scores ascribed to each scenario are shown in Table 4; note that fourth-year students and graduate nurses rated the same set of scenarios. Table 4 also contains the Pearson correlation coefficients indicating the reliability of the replicated scenarios. Internal consistency appeared comparable and reasonably acceptable for the three groups.

The mean modeled weights assigned to each attribute by each group are presented in Table 5. Each of the respondent groups, on the average, assigned most weight to the patient advocate attribute and comparatively equal weight to the attributes of profession advocate and collaborative ability. Overall, patient advocate received the highest weight in 75% of the responses, profession advocate in 13%, and collaborative ability in 12%. The i-test for between-group differences was possible only in comparisons of the fourthyear students' weights with those of the graduate nurses because the attributes' operational definitions were different for first-year students. The mean weight assigned by the fourth-year students to patient advocate was statistically significantly higher than the graduate nurses' mean weight for this attribute.

Table

FIGURE 2FIRST- AND FOURTH-YEAR NURSING STUDENTS' DERNITIONS OF THREE ATTRIBUTES OF THE "EXEMPLARY" NURSE

FIGURE 2

FIRST- AND FOURTH-YEAR NURSING STUDENTS' DERNITIONS OF THREE ATTRIBUTES OF THE "EXEMPLARY" NURSE

Table

TABLE 4RATING TASK SCORES

TABLE 4

RATING TASK SCORES

Discussion

The motivation for this work was primarily methodologic. The authors were interested in applying methods developed for assessing cancer patients' values to the problem of determining the relative importance ascribed to different attributes of the ideal nurse. Ib avoid presuppositions on the part of the investigators, the study design included an initial phase in which the attributes to be quantitatively weighted were first qualitatively identified by the respondents themselves. Different groups of respondents were studied to allow the emergence and comparison of different patterns in both qualitative and quantitative data. Differing patterns were observed; some of these may be actual differences, whereas others may be artifactual consequences of the sequence of data collection and tool construction.

Table

TABLE 5MODELED ATTRIBUTE WEIGHTS: MEANS AND STANDARD DEVIATIONS BY RESPONDENT GROUP

TABLE 5

MODELED ATTRIBUTE WEIGHTS: MEANS AND STANDARD DEVIATIONS BY RESPONDENT GROUP

In both classes of students, the qualitative characteristics cited tended to fall into the same major attributional areas. Patient advocate, profession advocate, and collaborative ability were the most frequently cited attributes and, therefore, were used in scenario construction. When the less frequently cited attribute of personal integrity was presented in the ranking task, however, it tended to be ranked higher than either profession advocacy or collaborative ability. This was not anticipated when the ranking and rating tasks were being designed. If the data collected with the ranking task had been examined before proceeding with developing the rating task, the attributes used in scenario construction would have been patient advocate, collaborative ability, and personal integrity, and potentially greater differences between the fourth-year and graduate nurses would have had an opportunity to emerge. This illustrates the difficulties that can arise when judgments about the salience of qualitative data are based on frequency counts alone.

Although both classes of students identified the same seven general attributes in Phase I, the lists of characteristics generated by the fourth-year students reflected, as expected, greater sophistication compared with those submitted by beginning students. This precluded direct comparisons of the scenario ratings elicited from these two groups, since the same attributional definitions could not be used. On the other hand, the internal consistency of the scenario rating task appeared reasonably acceptable, as was noted in previous studies (Llewellyn-Thomas et al, 1984a; Langley, Sutherland, Wong, et al, 1987), and the method was able to detect between-group differences (fourth-year students versus graduate nurses) when the same operational definitions could be used.

Therefore, the qualitative-quantitative data collection sequence is feasible and reveals interesting within-group quantitative patterns, but normative comparisons across disparate groups are inappropriate. Similarly questionable would be the application of exclusively quantitative methods in longitudinal designs to study, for example, the evolution across time of student perceptions of the ideal nurse.

However, these methods could be useful when the study purpose is not to take static snapshots of essentially nongeneralizable value profiles, but to promote value clarification in dynamic applications in social psychology, as has been argued elsewhere (Llewellyn-Thomas, 1982). Reports of using value assessment techniques to help in making politically (Edwards, 1977; Hammond & Adelman, 1976) or personally (Beach, Townes, Campbell, et al, 1976) difficult decisions have indicated that the judgment processes demanded by these methods have helped people to articulate more clearly not only their own opinions but also their assumptions about others' preferences. Similarly, elicitation and feedback about the different perceptions of the "exemplary" nurse that are held by students and nursing faculty, by nurse practitioners and other health-care team members, and by nursing staff and administrators could be a valuable research strategy in studies of student socialization, interdisciplinary conflict, job stress, and competency assessment (Benner, 1984; Cohen, 1981; Gordon, 1986).

References

  • Beach, L.R., Townes B.D., Campbell F.L., & Keating, G. W. (1976). Developing and testing a decision aid for birth planning decisions. Organization Behavior and Human Performance, 15, 99-116.
  • Benner, P. (1984). From novice to expert: Excellence and power in clinical nursing practice. Menlo Park, CA: Addison-Wesley.
  • Benner, P., & Kramer M. (1972). Role conceptions and integrative role behavior of nurses in special care and regular hospital nursing units. Nurs Res, 21, 20-29.
  • Cohen, H.A. (1981). The nursed guest for a professional identity. Menlo Park, CA: Addison-Wesley.
  • Corwin, R.G. (1961). The professional employee: A study of conflict in nursing roles. American Journal of Sociology, 66, 604-615.
  • Davis, F, & Oleson, U.L. (1964). Baccalaureate students' images of nursing: A study of change, consensus, and consonance in the first year. Nurs Res, 13, 8-15.
  • Edwards, W. (1977). How to use multi-attribute utility measurement for social decision making. IEEE Transactions on Systems, Man, and Cybernetics, 7, 326-340.
  • Gordon, D.R. (1986). Models of clinical expertise in American nursing practice. Soc Sci Med, 22, 953-961.
  • Habenstein, R. W., & Christ, EA. (1985). Professionalizer, Tradìtionalizer, and Utilizer. Columbia, MO: University of Missouri Press.
  • Hammond, K.R., & Adelman, L. (1976). Science, values, and human judgment- Science, 194, 389-396.
  • Hammond, K.R., Stewart, TR., Brehmer, B., & Steinman, D.O. (1975). Social judgment theory. In M.F. Kaplan & S. Schwartz (Eds.), Human judgment and decision processes (pp. 271-312). New York: Academic Press.
  • Heyman, R., Shaw, M.P., & Hording, J. (1983). A personal construct theory approach to the socialization of two nursing trainees in two British general hospitals. Journal of Advances in Nursing, 8, 59-67.
  • Jones, S.L., & Jones, P.K. (1977). Nursing student definitions of the "real" nurse. J Nurs Educ, 16, 15-20.
  • Kelly, J.G. Ü983). A comparison of first and third year nursing students role perception and expectations. Australian Nurses Journal, 12, 41-43.
  • Kramer, M. (1970). Role conceptions of baccalaureate nurses and success in hospital nursing. Nurs Res, 19, 428-439.
  • Langley, G.R., Sutherland, K.J., Wong, S-, Minkin, S., LlewellynThomas, H.A., & Till, J.E. (1987). Why are (or are not) patients given the option to enter clinical trials? Controlled Clin Trials, 8, 49-59.
  • Llewellyn-Thomas, H.A. (1982). Application of decision theory in long term care. In K. King (Ed.), Recent advances in nursing: Long term care (pp. 33-50). Edinburgh: Churchill Livingstone.
  • Llewellyn-Thomas, H.A., & Sutherland, H.J. (1987). Procedures for value assessment. In M. Cahoon (Ed.), Recent advances in nursing: Research methodology (pp. 169-185). Edinburgh: Churchill Livingstone.
  • Llewellyn-Thomas, H.A., Sutherland, H.J., Ciampi, A,, EtezadiAmoli, Boyd, N.F., & Till, J.E. (1984a). The assessment of values in laryngeal cancer: Reliability of measurement methods. Journal of Chronic Diseases, 37, 283-291.
  • Llewellyn-Thomas, H.A., Sutherland, H.J., Tibshirani, R., Ciampi, A., Till, J.E., & Boyd, N.F. (1982). The measurement of patients' values in medicine. Med Decís Mak, 2, 449-462.
  • Llewellyn-Thomas, H.A., Sutherland, H.J. Tibshirani, R., Ciampi, A., Till, J.E., & Boyd, N.F. (1984b). Describing health states: Methodologie issues in obtaining values for health states. Med Care, 22, 543-552.
  • Minehan, P.L. (1977). Nurse role conception. Nurs Res, 26, 374-379.
  • Whelan, E.G. (1984). Role-oriented change among RNs in an upper-division level baccalaureate program. J Nurs Educ, 23, 151-155.

TABLE 1

FACTORIAL DESIGN IN WHICH THE THREE ATTRIBUTES ARE COMBINED

TABLE 2

AGE, EDUCATIONAL STATUS, AND CLINICAL EXPERIENCE OF STUDY PARTICIPANTS

TABLE 3

MEDIAN RANKS FOR SEVEN MAIN ATTRIBUTES FOR EACH RESPONDENT GROUP

FIGURE 2

FIRST- AND FOURTH-YEAR NURSING STUDENTS' DERNITIONS OF THREE ATTRIBUTES OF THE "EXEMPLARY" NURSE

TABLE 4

RATING TASK SCORES

TABLE 5

MODELED ATTRIBUTE WEIGHTS: MEANS AND STANDARD DEVIATIONS BY RESPONDENT GROUP

10.3928/0148-4834-19891001-08

Sign up to receive

Journal E-contents