Journal of Nursing Education

Designing Clinical Examples to Promote Pattern Recognition: Nursing EducationBased Research and Practical Applications

Dorette Sugg Welk, PhD, RN



Despite carefully planned classroom sessions and richly designed clinical experiences, nurse educators know that, realistically, students will not encounter all of the life-threatening patient situations that require recognition and intervention during these sessions and experiences. However, on graduation, these students will be legally accountable for recognition of these situations and the immediate interventions associated with them. Therefore, the challenge for nurse educators is how to structure didactic interactions that promote pattern recognition in anticipation of such yet-to-be-experienced events.



Despite carefully planned classroom sessions and richly designed clinical experiences, nurse educators know that, realistically, students will not encounter all of the life-threatening patient situations that require recognition and intervention during these sessions and experiences. However, on graduation, these students will be legally accountable for recognition of these situations and the immediate interventions associated with them. Therefore, the challenge for nurse educators is how to structure didactic interactions that promote pattern recognition in anticipation of such yet-to-be-experienced events.

One of the major reasons nurses are sued is failure to recognize significant information, such that patients are harmed (Brent, 1997; Creighton, 1986; Fiesta, 1988; Guido, 1997). The author's original interest in this subject was stimulated by Cushing's (1988) report of a new graduate who failed to recognize a heart attack and, thus, did not intervene nor refer the problem to others. The patient died, and the nurse was successfully sued for malpractice. From a legal perspective, it was expected that, despite never having seen a heart attack occur, the nurse should have recognized the emerging pattern of signs and symptoms that (minimally) constituted the emergency situation. This article reviews pattern recognition literature in a variety of disciplines, including nursing; reports original research involving baccalaureate students and heart attack situations; and suggests practical applications of this information.


DeBono (1976), in his classic book leaching Thinking, devoted considerable time to the importance of patterns:

"Patterns" is probably the most important word we have. Ever since Plato, the grand-daddy of Western philosophers, concerned himself with form, all philosophers have been obsessed with the importance of pattern, form or idea. And so they should be. Patterns cover the areas of meaning, recognition and relationship (p. 84).

Considering numerous possible definitions of a pattern, DeBono (1976) selected one applicable to pattern recognition in clinical nursing practice (i.e., a pattern is present "when the movement from the present state to the succeeding state occurs with a probability above chance" [p. 87]).

Gagne (1985) defined a pattern as a set of attributes (e.g., cues, signs, symptoms) that typically exist together, the observation of which results in a gestalt-type recognition. Furthermore, she contended that patterns tend to be learned through presentation of typical examples or those that include the minimal essential features by which an individual recognizes that pattern.

Essential and Nonessential Features

A decade before, Homa and Chambliss (1975) indicated that if common features were to be extracted from typical examples, learners needed exposure to six to nine such examples that consistently presented the essential features (e.g., in the case of a heart attack, chest pain) and vary the nonessential features (e.g., the person's gender or age). A nonessential feature may contribute to an understanding of an individual patient's case (e.g., the development of crackling lung sounds or shortness of breath), but the health care provider is not "waiting" for this sign or symptom to occur to recognize the emergency situation. In turn, a nontypical example contains some of the essential features and some nonessential features, without particular attention to any frequency.

Even without referring to what is essential and nonessential in examples, learners use rapid thinking activities to extract this information in real-life situations (Gagne, 1985). If the examples are constructed haphazardly or inconsistently, learners still try to make sense of what is important and tend to either extract a prototype or "textbook picture" for future use in recognition tasks or store all of the examples individually in long-term memory for retrieval and comparison (Knapp & Anderson, 1984). The importance of controlling these examples for retrieval in memory is supported by Cullen (1998):

... at least some of the time, sound generalizations from experience are made, those being true and law-like generalizations versus merely true but accidental generalizations. Despite this capability, inductive inferences from experiences are, in fact, often not sound (p. 134).

Regarding example construction, Das-Smaal and de Swart (1986) showed that when irrelevant or nonessential features (e.g., gender, age) frequently are paired with essential features (e.g., chest pain), learners attribute importance to the nonessential features. For example, this may occur if a teacher tends to talk about the "man older than age 40" who is having a heart attack and learners link age and gender to the health situation as essential to recognition of a heart attack when, in fact, they are not. In pattern recognition theory, "irrelevant or nonessential" is used in the strict sense to describe features that need not be present. In nursing, a risk factor of age or gender may influence nurses' observations, but if nurses use this information as an essential criterion for recognizing a heart attack, a heart attack in a young woman, for example, may not be recognized as such.

Human Pattern Recognition Development

Pattern recognition research has continued in both the human and computer-oriented domains. Erickson and Kruschke (1998) described human categorization experiments in which they studied the use of rules and typical examples in learning and classification performance. They found that participants were better able to identify a block pattern when the frequency of essential pattern identifiers increased. This training also resulted in better recognition of when a set of stimuli was not the pattern under consideration. They also found that elevated frequency presentation of essential features resulted in more robust generalization or the ability to identify the pattern in the presence of infrequently seen stimuli (Erickson & Kruschke, 1998). Nosofsky and Palmieri (1997) also noted that if one stimulus is provided more frequently than another, recognition will be enhanced, and the pattern will be more likely to be judged typical of its category.

Hurwitz (1994) directly tested the presentation of whole versus fragmented patterns on human learning and recognition. He found that the ability to recognize the whole test pattern was not enhanced when a sample of undergraduate students were taught fragmented patterns and were expected to cognitively build the whole pattern for future recognition. This finding supports the need for educators to create well-designed whole patterns with essential features to use to help their students develop the skill of pattern recognition.

Computer Pattern Recognition Development

Pattern recognition has been considered an important component toward development of expert computer systems. Cullen (1998) reported the use of statistical pattern recognition for classifying individuals at risk, a highspeed method for clinical decision making in the ambulatory care setting. Sung and Poggio (1998) applied computers and pattern recognition theory to the task of detecting blurred human faces in images, a computerlearned task for identifying a set of distanced parameters from examples of face and non-face window patterns. This technique may be helpful in identifying individuals in a less distinct group photograph or surveillance videotape.

Hong (1997) described the use of computers to classify features of poisonous and nonpoisonous mushrooms to identify key patterns of recognition. The computer analyzed 20 iterations of 4,208 examples of edible mushrooms and 3,916 examples of poisonous mushrooms to identify five rules. Although nursing students will not encounter the number of examples or exact situations portrayed in such computer applications, these same principles may be useful in teaching pattern recognition to students in basic nursing education programs.

Pattern Recognition in Nursing

Pattern recognition theory recently has received much attention in nursing. Pattern recognition is viewed as a critical step or strategy in the process of thinking and in the diagnostic reasoning process in nursing. Fonteyn (1998) indicated that recognizing a pattern means "identifying characteristic pieces of data that fit together" (p. 20) and cited "case type," such as the medical diagnosis of heart attack, as one type of pattern to be recognized. The educational value of establishing succinct reliable patterns for novice nurses is reinforced as a way to help them discriminate lack of fit of data with a typical pattern. Some help in differentiating characteristic pieces or essential features in an array of relevant data when patterns first are introduced may support future pattern recognition.

Carnevali and Thomas (1993) suggested the need for cognitive structuring of knowledge from the outset of nursing education, "Diagnostic and treatment decisions can be made more efficient by systematically storing knowledge so that it can be retrieved when needed" (p. 31). In preclinical and early clinical stages of memory development in nurses, they recommended using typical cases to form early diagnostic reasoning pathways as memory encoding first occurs.

Miller and Babcock (1996), in their application of critical thinking to nursing, acknowledged that such discrimination requires "current knowledge and the best available evidence" (p. 16). Novice nurses lack experience with clinical situations, such as heart attacks, and may benefit from expert distillation of essential evidence or pattern features from the many possible relevant ones. Norman, Brooks, Regehr, Marriott, and Shah (1996) carried this point further and recommended that students have visual experiences to accompany lists of signs and symptoms learned in the classroom. They found that medical students, despite being able to provide memorized sign and symptom data about diseases, could not recognize these features when shown photographs depicting them. Neistadt and Smith (1997), occupational therapy educators, similarly found that students could define terminology but could not use it to describe and name observations when viewing assessment videotapes. They recommended that "curricula also need to help students direct their attention to relevant cues and offer interpretations about what occupational therapy problems those cues indicate" (p. 376).

Murphy and Friedman (1996) demonstrated differences in diagnostic reasoning dependent on knowledge representations between senior medical students and experienced physicians. Experienced physicians could differentiate ischemia from similar diagnoses by considering fewer findings than medical students could. In the absence of multiple experiences, which may have reinforced and condensed patterns for retrieval in long-term memory, the medical students displayed more iteration in their repertory grids before consideration of ischemia was evident to them. Not discounting the value of personal experience, some condensation of information, such as providing typical clinical examples, may facilitate more rapid or accurate recognition of clinical situations.

O'Neill and Dluhy (1997) described a framework for fostering critical thinking and diagnostic reasoning in nursing that highlights knowledge development from undergraduate students to beginning clinicians to experienced clinicians. Undergraduate students were reported to develop knowledge as fragmented networks and by using rule-based reasoning. Despite efforts to provide students with clinical experiences that provide more structure to these networks, this is not always possible in an actual clinical setting. Planning educational strategies, such as deliberate typical example construction, may bridge or enhance such knowledge network development.

In a study of 55 junior and senior baccalaureate students, Aquilino (1997) demonstrated links between knowledge development and diagnostic abilities, in part regarding their ability to use situational cues about childbearing families to formulate nursing diagnoses. Knowledge test scores were positively related to diagnostic ability scores. Aquilino (1997) reported that "the content knowledge a nurse brings to the diagnostic reasoning task determines how the problem will be interpreted and which items of clinical information will be attended to" (p. 115). Furthermore, Szaflarski (1997), in a description of the diagnostic reasoning process in acute and critical care, emphasized that "explicit knowledge is essential to the diagnostic reasoning process because it provides the foundation on which reasoning takes place" (pp. 293-294). In their description of the evolution of numerous nursing process models, Frauman and Skelly (1999) cited cue recognition as a key factor in diagnostic reasoning. All of these authors support the notion that initial structuring and presentation of key knowledge about health or diagnostic problems, such as heart attacks, may facilitate diagnostic ability.

In a study comparing experienced Australian nurses and baccalaureate students, Taylor (1997) found that the students did not interact during change-of-shift or "handover" reports because "they recognized very few cues due to limited experiential learning. Although they attended handovers, most of the information was meaningless" (p. 332). In turn, when actually caring for clients, Taylor (1997) reported that novice nurses noted few cues and did not seek additional data because "a lack of knowledge about clinical disorders meant that the novice was unable to recognize signs and symptoms in any meaningful way" (p. 333).

Similarly, Tabak, Bar-Tal, and Cohen-Mansfield (1996), in their study of clinical decision making in experienced and novice nurses, reported that novice nurses express more certainty and less difficulty than experts in making a diagnosis when information presented was inconsistent with the pattern of a known diagnosis. They concluded that novice nurses may be content to resort to a simplistic cognitive schema they hold about the problem rather than analyze it further. These behaviors suggest potential risks to clients when cues are linked directly to recognition of serious health problems.

The initial construction of knowledge provides a basis for critical thinking, diagnostic reasoning, and clinical decision making. In the absence of personal experiences, novice nurses require help to develop cognitive schema that they may retrieve when presented with a variety of cues in future clinical situations. Pattern recognition research has merit for application to nursing education because nursing faculty have the knowledge and experience of patterns to be recognized in nursing practice and can employ teaching strategies to facilitate such recognition. One possible educational strategy may be the use of well-developed typical examples that distill the wider body of knowledge to essential or "characteristic pieces" most likely to be present in a particular health problem or pattern.


The purpose of this study was to determine whether the type of example design used in instruction results in differences in recognizing essential and nonessential information about heart attacks, in recognizing a heart attack situation when other health situations are presented, or both. If design makes a difference, nurse educators could adapt their current strategies to better promote the recognition outcome. Five hypotheses were tested:

* There is no difference in sophomore nursing students' recognition of essential information about heart attacks based on reading typical versus nontypical examples.

* There is no difference in sophomore nursing students' recognition of nonessential information about heart attacks based on reading typical versus nontypical examples.

* There is no difference in sophomore nursing students' recognition of heart attack situations based on reading typical versus nontypical examples.

* There is no difference in sophomore nursing students' recognition of non-heart attack situations based on reading typical versus nontypical examples.

* There is no difference in sophomore nursing students' recognition of nonessential features, such as gender and age, when the frequency of pairing these features with a heart attack situation is increased.

Essential and nonessential signs and symptoms of heart attack situations were selected by a simple frequency count of signs and symptoms presented in nursing and medical textbooks at the time of tool development (Cardiopulmonary Emergencies, 1990; Isselbacher et al., 1994; Kinney, 1991; Underbill, Woods, Froelicher, & Halpenny, 1989; Wingate, 1991). All of the signs and symptoms listed in both types of example designs could occur in the presentation of a heart attack situation. The five essential signs and symptoms selected were:

* Restlessness.

* Chest pain.

* Diaphoresis.

* Nausea or vomiting.

* Anxiety.

The 14 nonessential signs, symptoms, risk factors, or characteristics selected were:

* Pain radiating down the left arm.

* Pallor.

* Male gender.

* Weakness.

* Female gender.

* Palpitations.

* Shortness of breath.

* Pain radiating to the back, neck, or jaw.

* Age older than 40.

* Dizziness.

* Sense of impending doom.

* Age younger than 40.

* Racial background.

* Crackling lung sounds.

Although the presence of the nonessential risk factors increase the likelihood that a heart attack may be occurring, they are not considered essential features of the actual clinical picture of an evolving attack, which was the focus of this study. These two lists do not represent an actual lesson plan for teaching about heart attacks. However, they serve the research purpose of differentiating the broad array of signs, symptoms, risk factors, and characteristics applicable to heart attacks consistent with typical example design, which could be applied to any health or nursing specialty.


A quasi-experimental pretest/posttest design with random assignment to two groups was used. The Department of Nursing Human Subjects Committee approved the research as exempted or posing no more than minimal risk for subjects.


The subjects were 162 sophomore nursing students from five baccalaureate nursing programs drawn randomly from the list of programs accredited by the National League for Nursing. Students had taken human anatomy and physiology courses but had no formal program coursework about heart attacks. Program directors provided access to the students and an onsite individual agreed to distribute and collect the research materials.


Two types of examples were developed, and each contained approximately 100 words. Six examples were considered typical because each contained the five essential signs and symptoms (Table 1). Six examples were considered nontypical because each contained a variety of signs and symptoms that could occur in a heart attack situation but with no discernible pattern and because of the deliberate inclusion of demographic characteristics (e.g., specific gender or age identifier) (Table 2).

In the six typical examples, there were three women and three men with ages ranging over 6 decades (i.e., ages 18, 25, 36, 45, 60, 72). Care was taken to avoid any link between age, gender, and heart attack through the demographic variables. On the other hand, in the six nontypical examples, four were men, three of whom were age 40 or older (i.e., ages 18, 40, 45, 55), and two were women (i.e., ages 36 and 50), to test whether frequent pairing of features would affect their recognition as essential heart attack features. The examples and other research materials demonstrated content validity, as verified by three nursing faculty members with experience in cardiopulmonary nursing situations.

Each participant randomly received a packet that contained:

* A pretest consisting of 19 items that participants designated by a check mark as essential or nonessential to recognition of a heart attack situation.

* Six examples (i.e., either six typical or six nontypical examples). Participants did not know which packet they had, but they knew the posttest tasks involved identifying essential and nonessential features of heart attack situations and differentiating heart/non-heart attack situations.

* A posttest identical to the pretest, which designated the same 19 items as essential or nonessential to a heart attack situation.

* A second posttest consisting of 10 brief case situations of people experiencing emergency situations. Five people experienced heart attack situations that contained only the five essential features embedded in a context. The other five people experienced non-heart attack or other types of emergency situations related to cardiopulmonary circumstances, such as pulmonary edema or fluid in the lungs, and pneumothorax or lung collapse. Participants answered yes or no regarding whether the situation was a heart attack.

Directions were provided requesting participants to complete each part in immediate succession and return each part to the packet, without reviewing previous parts. Data collection took approximately 20 minutes. Personal integrity and a faculty monitor to help with enforcement as necessary were important components of the research design. Participants also were offered the opportunity to review the anticipated responses after completion of the materials, with the cautionary note that the research materials were not meant to negate what they may have learned taking cardiopulmonary resuscitation (CPR) classes or would learn in future classes or clinical situations.


Cronbach's alpha was performed on the total number of items for each of the instruments, rather than on individual subscales. The 19-item posttest had a reliability coefficient of .78, and the 10 -item heart attack/non-heart attack posttest had a reliability coefficient of .56.

All five hypotheses were written as null hypotheses in this study, which means there was no predicted relationship between the variables. A null hypothesis was accepted when there was no difference between the groups on the variable of interest as stated in the hypothesis, and it was rejected when the prediction of no difference was not the case (i.e., there was a difference between the two groups on the variable of interest).

In the analysis, there were no significant differences between the two groups on any of the individual 19 items in the pretest nor in the overall pretest scores (t = .75, df = 160, ? = .46). No pretest differences occurred in recognition of either the essential features (t = 1.12, df= 160, ? = .52) or the nonessential features (t = 1.72, df= 160, ? = .29). There was a significant difference in the 19-item posttest scores (t = 9.23, df= 160, ? < .001) in favor of the typical examples group.

The first hypothesis was rejected because there was a significant difference in the recognition of five essential features of a heart situation based on reading the two different designs of the heart attack examples in favor of typical examples (p = .001).

The second hypothesis also was rejected because there was a significant difference in the recognition of the 14 nonessential features of a heart attack situation, dependent on the type of example design in favor of typical examples (p < .001).

The third hypothesis was accepted because there was no significant difference in recognition of the five heart attack situations (t = 1.37, df= 160, ? = .172). Both groups correctly identified an average of four of the five heart attack situations.

However, the fourth hypothesis was rejected because there was a significant difference in the recognition that the presented non-heart attack situation was not consistent with a heart attack situation. As indicated above, the five heart attack situations contained only the five essential features, embedded in a realistic context, to be considered a heart attack.

As noted above, there were no significant differences in any of the 19 items on the pretest. However, significant differences between the two example design groups did occur in 13 of the 19 posttest items. Three significant differences occurred for essential features (i.e., restlessness [p < .05], diaphoresis [p < .01], nausea/vomiting [p < .001]) and 10 occurred for nonessential features (i.e., pain down left arm, pallor, weakness, palpitations, racial background, pain to neck, back, or jaw, and dizziness [p < .001], and sense of doom, male gender, and age older than 40 [p < .05]).

The fifth hypothesis was rejected because the participants were significantly more apt to select "male gender" or "age older than 40" as essential to recognition of a heart attack (p < .05). Two of the nonessential features, age and gender, were treated differently in the typical and nontypical examples groups. In both situations where the features of age and gender were linked purposely to a heart attack in the nontypical examples groups, participants selected them as essential to recognition of a heart attack.

Prior to reading the different example formats, the two groups did not differ significantly in their knowledge of essential and nonessential features of heart attack situations. Knowing that the posttests would require them to identify essential and nonessential features and to differentiate between heart attack and non-heart attack situations, the participants were primed for the type of thinking activity expected of them in the study.

The typical examples group had read six examples in which the five essential features were presented consistently, although synonyms were used for common features (e.g., "nausea or vomiting" was described as nauseous, vomits up coffee, severe indigestion, sick to her stomach, stomach feels queasy).

Because the participants themselves had to glean the essential features from the examples, Gagne's (1985) belief that a rapid thinking activity occurs in pattern recognition appears to be supported. The six examples seemed sufficient to identify a pattern for these five signs and symptoms, which is consistent with the recommendation for six to nine examples suggested by Homa and Chambliss (1975). This rapid thinking activity may be in effect because the typical examples group differed significantly in the posttest selection of items that were nonessential to recognizing a heart attack situation.

The lack of a significant difference in recognizing a heart attack situation may be explained by the fact that both groups read only examples about people experiencing heart attacks. Therefore, because the five heart attack situations included at least the five essential features consistently and some of the nonessential features, both groups could identify these situations.

It is the significant difference between the groups in recognizing the non-heart attack situations that suggests the group that read the typical examples looked specifically for the essential features. Therefore, they were more successful in deciding that a non-heart attack situation was occurring. Although the nontypical examples group also read six heart attack situations, there were more symptoms presented in their examples that were common to other cardiopulmonary non-heart attack situations, such as the "crackling lung sounds" and "shortness of breath" of pulmonary edema. The typical examples group had not seen these signs or symptoms in their examples and appropriately may have rejected these situations as non-heart attack situations when they did not find the essential features present.

The deliberate pairing of "age older than 40" and "male gender" in the nontypical examples resulted in significant differences between the two groups on these items. In just six examples, it appeared that the nontypical examples group saw this link and concluded that these two features were essential to recognizing a heart attack situation, which supports researchers who found that learners linked items based on the frequency with which they occurred together (Das Smaal & de Swart, 1986; Erickson & Kruschke, 1998; Hong, 1997; Hurwitz, 1994; Nosofsky & Palmieri, 1997).

It also is interesting to note the percentage of "wrong" selections of nonessential signs and symptoms by the typical examples group after reading the examples. As noted above, none of the 10 features appeared in any of the examples they were given; yet, they retained them as essential features in the posttest. For example, on the posttest, "pain radiating down the left arm" erroneously was cited as essential (i.e., a "wrong" response) by 33% of the typical examples group (pretest "wrong" response was 90%) and by 84% of the nontypical examples group (pretest "wrong" response rate was 94%). Isselbacher et al. (1994), regarding chest pain in a heart attack, stated that "typically the pain involves the central portion of the chest and/or epigastrium and in about 30% of the cases it radiates to the arms" (p. 1066). Therefore, it does not occur in the majority of cases. Although it is not known where participants obtained their information about heart attacks, the tendency to retain one's beliefs despite contrary educational information poses yet another challenge for educators.

Presentation of case types (Fonteyn, 1998) in the form of typical examples is one form of structuring knowledge development (Carnevali & Thomas, 1993) and "the best available evidence" (Miller & Babcock, 1996, p. 16). Such a distillation of key factors or features presented in a classroom or clinical situation may reduce the mental iterations required for recognition (Murphy & Friedman, 1996), connect fragmented networks (O'Neill & Dluhy, 1997), and better link knowledge to diagnostic ability (Aquilino, 1997). Perhaps if faculty show novice nurses how key signs and symptoms link to recognition of a health pattern (Neistadt & Smith, 1997) through example development, students may gain more in clinical situations, such as change-of-shift reports and client interactions (Taylor, 1997).


Participant integrity in receiving all of the research materials at once and following the sequence directions may have been compromised at times. However, this procedure was designed to make the onsite data collection a reasonable activity for a colleague to manage so a national sample could be obtained. Although a pilot study was conducted and content validity was established, the reliability coefficient of .56 for the sample was low, which suggests lack of consistency or stability of the instrument for this group. Therefore, the construction of the 10 heart attack/non-heart attack situations could be improved to ensure more reliable diagnostic distinctions.

In actual teaching situations, nurse educators also would be cautioned not to view the five essential features selected for this study as definitive for heart attack recognition. In addition, it was not possible to control other avenues by which participants learned about heart attack situations, such as personal experience or completion of a CPR course. However, because the groups did not differ significantly on the pretest, it can be assumed these factors operated somewhat similarly for both groups.


This study represents an example of the rapidity with which patterns emerge in a learning situation and the tendencies of learners to form a prototype or "textbook picture" based on what is presented to them and to not abandon ideas even when presented with contrary evidence. Because of the inability to control the clinical situations of a given nursing program in terms of actual observations of nursing and medical situations, the control of the didactic presentations becomes paramount to establishing pattern recognition for future situations. The author has used pattern recognition theory and this research to approach classroom and clinical teaching and offers the following suggestions for practical applications:

* In preparation for presentation of a life-threatening clinical situation, preplan several examples that include the major signs and symptoms. In designing presentations, transparencies, or handouts, use clear designations to differentiate essential and nonessential features of that diagnosis or situation. Avoid a lengthy undifferentiated list that suggests the expected presence of all items. Share this differentiation verbally with students so they become more aware of this technique for the benefit of their learning.

* Provide students with six to nine brief typical examples and, as in discovery learning, ask them to identify the essential features of the situation. By using synonyms for signs and symptoms, students also are better prepared for these different but common patient descriptions in their collection of subjective and objective data. This activity can be a take-home assignment and serve as a springboard to begin the next class on a new topic. Note that this activity is not the same as a richly detailed case study for which novice nurses may not yet be ready. Consider the research of Norman et al. (1996) and Neistadt and Smith (1997) regarding whether students actually can identify a sign or symptom when they see it versus the ability to just name or define it.

* To the extent possible, preplan nontypical clinical examples and let students discover how and why these examples are not typical. Students benefit from experienced faculty's diverse, detailed practice situations, and they need to know that things do not always happen like they do in textbooks. However, students should see the example as somewhat unusual so their recall of an example to match the current situation does not retrieve only this nontypical example for comparison. Students need to have a template against which to measure their own clinical experiences as typical or nontypical.

* Take advantage of a nursing diagnosis reference format that presents use of major and minor defining characteristics as consistent with teaching about lifethreatening and other clinical patterns (e.g., Carpenito, 1999). Reinforcement of this approach in classroom and clinical situations is a great opportunity to apply pattern recognition theory on a weekly basis. All signs and symptoms do not bear equal weight above probability of chance (DeBono, 1976), and some practice with this notion may be helpful.

* Invite students in classroom and postconference situations to share an example of a clinical situation in hopes of acquiring six to nine examples for group comparison. How are these the same, and how are they different? What is common that may need to be present to recognize this clinical pattern in the future? This critical thinking activity can provide the structure of a rapid thinking activity in pattern learning that Gagne (1985) suggested.


Several conclusions were drawn from this study:

* The deliberate design of examples to consistently include the essential signs and symptoms of a heart attack situation resulted in significant differences in sophomore nursing students' recognition of these essential features and a designation of other signs, symptoms, and characteristics that were not essential.

* Whether examples of heart attack situations included only essential features or both essential and nonessential features, sophomore nursing students were able to identify other heart attack situations presented to them.

* Having been presented with essential features only, sophomore nursing students were better able to recognize when a situation was not a heart attack.

* When nonessential features, such as age and gender, were presented purposely in specific patterns in nontypical examples, a group of sophomore nursing students were more likely to select these items as essential to recognizing a heart attack situation than those who read examples where no such pattern was presented.

In an approximate 20-minute exposure to these research materials and only six heart attack examples, sophomore nursing students, whose results did not differ significantly on a pretest, demonstrated changes in their recognition of essential and nonessential heart attack features and of non-heart attack situations, dependent on the design of the examples, at posttest.

Pattern recognition theory and research, the demonstrated experiences of sophomore nursing students in this research, and practical applications to everyday classroom and clinical situations were described in this article. Together they may inform and provide meaningful strategies to promote pattern recognition of life-threatening situations and the avoidance of patient harm.


  • Aquilino, M.L. (1997). Cognitive development, clinical knowledge, and clinical experience related to diagnostic ability. Nursing Diagnosis, 8, 110-119.
  • Brent, N.J. (1997). Nurses and the law: A guide to principles and applications. Philadelphia: Saunders.
  • Cardiopulmonary emergencies. (1990). Springhouse, PA: Springhouse Corporation.
  • Carnevali, D.L., & Thomas, M.D. (1993). Diagnostic reasoning and treatment decision making in nursing. Philadelphia: Lippincott.
  • Carpenito, L. (1999). Handbook of nursing diagnosis (8th ed.). Philadelphia: Lippincott.
  • Creighton, H. (1986). Law every nurse should know (5th ed.). Philadelphia: Saunders.
  • Cullen, P. (1998). Gleaning patterns from practice: Intelligent systems in ambulatory care. Nursing Economics, 16, 133-135.
  • Cushing, M. (1988). The legal side. American Journal of Nursing, 88(7), 10.
  • Daa-Smaal, EA., & de Swart, J.H. (1986). Effects of contrasting category, conjoint frequency and typicality on categorization. Acta Psychologica, 65, 15-40.
  • DeBono, E. (1976). Teaching thinking. New York: Penguin Books.
  • Erickson, M. A., & Kruschke, J.K. (1998). Rules and exemplars in category learning. Journal of Experimental Psychology: General, 127, 107-131.
  • Fiesta, J. (1988). The law and liability: A guide for nurses (2nd ed.). New York: Wiley.
  • Fonteyn, M.E. (1998). Thinking strategies for nursing practice. Philadelphia: Lippincott.
  • Frauman, A.C., & Skelly, A.H. (1999). Evolution of the nursing process. Clinical Excellence for Nurse Practitioners, 3, 238244.
  • Gagne, E. (1985). The cognitive psychology of school learning. Boston: Little, Brown and Company.
  • Guido, G.W. (1997). Legal issues in nursing. Stamford, CT Appleton.
  • Homa, D., & Chambliss, D. (1975). The relative contribution of common and distinctive information on the abstraction of illdefined categories. Journal of Experimental Psychology: Human Learning and Memory, 1, 351-359.
  • Hong, S. J. (1997). R-MDNI: An iterative approach for generating minimal rules from examples. IEEE Transactions on Knowledge and Data Engineering, 9, 709-717.
  • Hurwitz, J.B. (1994). Retrieval of exemplar and feature information in category learning. Journal of Experimental Psychology: Learning, Memory and Cognition, 20, 887-899.
  • Isselbacher, K.J., Braunwald, E., Wilson, J.D., Martin, J.B., Fauci, A.S., & Kasper, D.L. (1994). Harrisons principles of internal medicine (13* ed.). New York: McGraw-Hill.
  • Kinney, M.R. (Ed.). (1991? Comprehensive cardiac care. St. Louis: Mosby.
  • Knapp, AG., & Anderson, JA. (1984). Theory of categorization based upon distributed memory storage. Journal of Experimental Psychology: Learning, Memory and Cognition, 10, 616-637.
  • Miller, MA, & Babcock, D.E. (1996). Critical thinking applied to nursing. St. Louis: Mosby.
  • Murphy, G.C., & Friedman, CP. (1996). Differences in knowledge representations of experienced and inexperienced clinicians as captured by repertory grids. Academic Medicine, 71(1 Suppl.), S16-S18.
  • Neistadt, M.E., & Smith, RE. (1997). Teaching diagnostic reasoning: Using a classroom-as-clinic methodology with videotapes. The American Journal of Occupational Therapy, 51, 360-368.
  • Norman, G.R., Brooks, L.R., Regehr, G, Marriott, J., & Shah, V. (1996). Impact of feature interpretation on medical student diagnostic performance. Academic Medicine, 71(Suppl. 1), S108-S109.
  • Nosofsky, R.M., & Palmieri, T.J. (1997). An exemplar-based random walk model of speeded classification. Psychological Review, 104, 266-300.
  • O'Neill, E.S., & Dluhy, N.M. (1997). A longitudinal framework for fostering critical thinking and diagnostic reasoning. Journal of Advanced Nursing, 26, 825-832.
  • Sung, K, & Poggio, T. (1998). Example-based learning for viewbased human face detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, 20, 39-51.
  • Szaflarski, J.L. (1997). Diagnostic reasoning in acute and critical care. AACN Clinical Issues, 8, 291-302.
  • Tabak, N., Bar-Tal, Y, & Cohen-Mansfield, J. (1996). Clinical decision-making of experienced and novice nurses. Western Journal of Nursing Research, 18, 534-547.
  • Taylor, C. (1997). Problem solving in clinical nursing practice. Journal of Advanced Nursing, 26, 329-336.
  • Underhill, S.L., Woods, S.L., Froelicher, E.S., & Halpenny, CJ. (1989). Cardiac nursing (2nd ed.). Philadelphia: Lippincott.
  • Wingate, S. (Ed.). (1991). Cardiac nursing: A clinical management and patient care resource. Gaithersburg, MD: Aspen.


Sign up to receive

Journal E-contents