Using an educational data mining approach, first-year academic achievement of undergraduate nursing students, which included two compulsory courses in introductory human anatomy and physiology, was compared with achievement in a final semester course that transitioned students into the workplace. We hypothesized that students could be grouped according to their first-year academic achievement using a two-step cluster analysis method and that grades achieved in the human anatomy and physiology courses would be strong predictors of overall achievement. One cohort that graduated in 2014 (n = 105) and one that graduated in 2015 (n = 94) were analyzed separately, and for both cohorts, two groups were identified, these being “high achievers” (HIGH) and “low achievers” (LOW). Consistently, the anatomy and physiology courses were the strongest predictors of group assignment, such that a good grade in these was much more likely to put a student into a high-achieving group. Students in the HIGH groups also scored higher in the Transition to Nursing course when compared with students in the LOW groups. The higher predictor importance of the anatomy and physiology courses suggested that if a first-year grade-point average was calculated for students, an increased weighting should be attributed to these courses. Identifying high-achieving students based on first-year academic scores may be a useful method to predict future academic performance.
- cluster analysis
- anatomy and physiology
a common semester at the outset of an undergraduate nursing curriculum has the potential to equip students with the academic skills necessary for successful navigation of their undergraduate journey. Key skills introduced to students at the start of an undergraduate program of study include (but are not limited to) an awareness of appropriate academic writing standards (plagiarism and referencing), independence in knowledge acquisition (study skills), and the ability to constructively engage with others to achieve a common objective (group work). These key skills are also requirements for effective interprofessional healthcare practice (21), and therefore, they are keenly sought by employers within the health sector. Also, introductory courses in human anatomy and physiology are common inclusions in an undergraduate nursing curriculum (10, 22), although prerequisite scientific knowledge that may underpin concepts in physiology is not always necessary (27).
The common language used by many different health professionals has origins within physiology, where a mutual understanding of clinical terminology and medical definitions may underpin appropriate strategies to treat patients (26, 29). However, the teaching of anatomy and physiology has presented a major hurdle in the undergraduate nursing curriculum (6, 7, 13, 14, 22). For example, it has been shown that nursing students appeared to learn physiology content best when it was related to experiences in their workplace (7), although recently, qualified nurses were more likely to indicate that physiology in their preregistration curriculum was rarely linked to clinical practice (5). These students also reported that they were generally disappointed with the lack of integration of the principles of physiology with nursing practice (7).
Assessment of student learning is a fundamental function of higher education and is the means by which we assure and express academic standards. Assessment is often used as the basis for academic progression (18), and therefore, appropriate assessments may guide student learning by influencing both the approach to learning and the confirmation of the attainment of a learning outcome (2, 19, 20). Academic achievement in introductory anatomy and physiology courses may also predict success in later years of study, and this may inform university policy regarding student progression (9). For example, competition for places on a clinical program may be decided by academic achievement in an introductory physiology course(s). Currently, there is a paucity of empirical evidence to support this argument despite a growing interest in education data mining (1, 3, 4). Educational data mining is concerned with developing methods for exploring the unique data that come from educational settings and using those methods to better understand students and the progression of students through the settings in which they learn (15). Therefore, in the current study, we analyzed academic achievement of nursing students in their first year of an undergraduate health science degree (which included two compulsory first-year courses in introductory human anatomy and physiology) and compared this with performance in a pivotal final semester course that prepared nursing graduates for the workplace. We hypothesized that first-year undergraduates could be grouped according to their academic achievement in a common suite of courses and that the grades achieved in compulsory introductory courses in human anatomy and physiology would be strong predictors of overall achievement.
This research was carried out at the Auckland University of Technology, a large publicly funded university in New Zealand. The undergraduate nursing program typically attracts students with a diverse range of preuniversity educational experiences, including both school-leavers and those reentering formal education following a period of either work or unemployment. The gender balance in the nursing program was typically 9:1 female to male.
The academic grades of students graduating with a Bachelor of Health Science (Nursing) Degree [BHSc(NURS)] in 2014 (n = 105) and 2015 (n = 94) were accessed through the university’s data management system (ARION). Throughout the analysis, deidentified, aggregated data were used, thus presenting no student privacy issues. This study did not require a full submission to the University Ethics Committee, although appropriate advice was sought from both the university research advisor and the university privacy officer; a condition was that no individuals could be identified by the researchers, nor could a student identify their own data from the analysis. An ethics committee requirement was that no student personal details were included in the data made available for analysis, thus restricting data to deidentified untraceable numerical format. All analyses were carried out using IBM SPSS Statistics 22.
The BHSc(NURS) consisted of six semesters of higher education organized into 3 yr of full-time academic study with two 13-wk teaching semesters in each academic year. At the end of each semester there was a 1-wk study break, followed by a 2-wk examination period. All students were required to attain 60 credits from relevant courses offered in each semester, with some courses predetermined as compulsory and others as elective. All first-year courses were each worth 15 credits (as were most courses throughout the 3 yr); however, a compulsory third-year course in semester 6 was worth 45 credits. This course was a compulsory requirement for all nursing undergraduates; it was intended to demonstrate and assess the application of classroom knowledge into a clinical setting. A student was assessed using a variety of common assessment procedures (e.g., assignment, presentation, examination) in each course to determine the achievement of named learning outcomes prescribed at the start of each course. A student would be awarded a grade for the course depending on their academic achievement of each learning outcome. The BHSc(NURS) degree required successful attainment of 360 credits.
The common first-year courses were 1) Human Anatomy and Physiology 1; 2) Human Anatomy and Physiology 2; 3) Health and the Environment; 4) Lifespan Development and Communication; and 5) Knowledge, Communication, and Enquiry. The compulsory final semester (semester 6) course was titled Transition to Nursing. Possible alphabetic grades for all these courses were A+, A, A−, B+, B, B−, C+, C, and C−, and these were given numerical scores of 9, 8, 7, 6, 5, 4, 3, 2, and 1, respectively.
Both courses in human anatomy and physiology were delivered as a weekly 3-h lecture (recorded at the time of initial delivery and made available to all students for the remainder of the course) and a weekly 2-h tutorial over a continuous 13-wk period.
Human Anatomy and Physiology 1.
All lecture slides could be prepurchased by students, and additional work sheets were used to support learning outcomes in the tutorial sessions. Two 1-h laboratory sessions were also part of the course, these being a bone and joint dissection (bovine) and a heart and lung dissection (lamb). Students were strongly encouraged to purchase an introductory anatomy and physiology textbook, for example, Marieb’s Essentials of Human Anatomy and Physiology, 11th Ed. (19a), and although not compulsory, attendance at both lectures and tutorials was strongly encouraged. Learning outcomes for Human Anatomy and Physiology 1 were as follows:
1. Demonstrate fluency in terminology relevant to human anatomy and physiology.
2. Classify and describe the structure of human tissue and describe gross anatomy and organization of major human organ systems.
3. Describe physiological processes of major human organ systems and the role of homeostasis in maintaining their integrity.
Human Anatomy and Physiology 2.
All lecture slides were made available for students to print at least 1 wk in advance of the lecture, copies of which remained in an online, student-accessible repository for the remainder of the course. There were two laboratory sessions, one being a blood-typing and hematocrit practical session and the other a microbiology practical that used a throat swab to grow bacteria on an agar medium and microscope slides to observe types of bacteria. The recommended text for this course was Marieb and Hoehn’s Human Anatomy and Physiology, 10th Ed. (19b). Students were expected to attend both a lecture and weekly 2-h tutorial. Learning outcomes for Human Anatomy and Physiology 2 were as follows:
1. Relate knowledge of human anatomy and physiology to the maintenance of normal health and function.
2. Apply introductory knowledge of pharmacology to physiological processes.
3. Examine the role of microorganisms in health and disease.
Transition to Nursing.
The course descriptor for the Transition to Nursing course was: “Facilitate the transition from student to registered nurse within a practice environment, exploring professional issues and integrating knowledge, research, and practice.” The course used a combination of classroom teaching and clinical practice placements that could include but were not restricted to mental health (addiction and recovery), pediatrics, maternal and child health, rehabilitation and disability, and acute medical and surgical settings. Learning outcomes for Transition to Nursing, which focused on application of knowledge, were as follows:
1. Apply concepts of theoretical knowledge (with associated skills) to the management of client care.
2. Articulate and adapt practices supported by nursing knowledge and evidence-based research of contextual health factors and professional issues to meet the client’s needs.
3. Analyze and integrate research and knowledge for effective clinical reasoning and nursing practice.
The data on students graduating in 2014 and in 2015 were analyzed separately. A two-step cluster analysis was performed on data from all students obtaining a minimum passing grade (C− or above) for all of the first-year courses. This clustering analysis, which is a form of data mining, identified clusters embedded in data, where a cluster is a group of data objects that are similar to one another (8, 16, 28). The similarity of objects in a cluster may not be obvious and may be based on a collection of measures (for example, the academic scores in a suite of courses) rather than a single criterion. Although the number of variables used to define a group are not predetermined, a sufficient number is needed to ensure adequate discrimination between items placed in a group.
The two-step cluster analysis procedure (1, 11) is an exploratory tool designed to reveal natural groupings within a data set that would otherwise not be apparent. The algorithm employed by this procedure has several desirable features that differentiate it from other clustering techniques; for example, it assumes variables to be independent such that a joint multinomial-normal distribution can be placed on both categorical and continuous variables. Also, there is an automatic selection of the number of clusters; by comparing the values of a model/choice criterion across different clustering solutions, the procedure can automatically determine the optimal number of clusters (11). In the current study, the scores for five compulsory first-year courses were used as input variables for the analysis and standardized such that the highest score for any input was 1 and the lowest score for any input was 0. These inputs were used to group similar students, with the total number of groups not determined beforehand. The log-likelihood distance measure and Schwarz’s Bayesian Criterion were used in the two-step clustering analysis (11). When groups were identified, the difference between inputs in each clusters was compared with an unpaired Student’s t-test, where a significant difference was indicated by P < 0.05.
The academic achievements in the Transition to Nursing course (the compulsory semester 6, 45-credit course) of the two groups identified in the cluster analysis were compared using Student’s t-test, where a significant difference was indicated by P < 0.05.
For graduating students in 2014 (see Fig. 1, left), two clusters were identified, and these were described as LOW (cluster 2) and HIGH (cluster 1). Students in the LOW cluster average scores were lower in first-year courses compared with students in the HIGH cluster. In the cohort that graduated in 2014, there were 57 students in the HIGH group and 48 students in the LOW group, and all inputs (i.e., the scores attained in each first-year course) were significantly different between clusters (P < 0.05 Student’s t-test; see Table 1). For graduating students in 2015 (see Fig. 1, right) two clusters were identified, and these were described as LOW (cluster 1) and HIGH (cluster 2). In this cohort, there were 43 students in HIGH and 51 students in LOW, and all inputs were significantly different (P < 0.05; see Table 1) between clusters. This analysis provided the means to distinguish two separate groups of students for each graduating year, where one group can be considered as higher-achieving students in their first year and another group can be considered as lower-achieving students in their first year; however, all were students that passed.
The two-step cluster analysis has used all scores from the five first-year courses for each student to allocate students to a cluster. Additionally, the analysis calculated the predictive importance of each score; this is the ability of each input to determine cluster assignment. For both cluster analyses, the two anatomy and physiology courses scored the highest for predictor importance (as shown in Fig. 2); this indicated that a high score in these two courses was more likely to put a student into the HIGH cluster than a high score in any other course.
For the students graduating in 2014, the mean score for the semester 6 Transition to Nursing course for the HIGH cluster was higher than the mean score for the LOW cluster (P < 0.05; see Table 2). Also, for the students graduating in 2015, the mean score for the Transition to Nursing course was higher for the HIGH cluster compared with the LOW cluster.
This study uniquely used an educational data mining approach to study the importance of academic achievement in courses in introductory human anatomy and physiology within a BHSc(NURS) degree. A two-step cluster analysis identified high- and low-achieving groups in the first year of study, and we uniquely report that the two first-year anatomy and physiology courses had the highest predictor importance. A further novel finding of this study was that students grouped as high achievers based on their academic achievement in their compulsory courses in the first year of study scored higher in a semester 6 capstone course that transitioned undergraduate nursing students into the workplace when compared with those grouped as low achievers.
A low-achieving cluster may indicate that some undergraduates are still in need of extra instruction, for example, in their anatomy and physiology courses. Identifying lower-achieving students is the first step in targeting resources to increase the academic attainment of these students; we speculate that a cluster analysis of students’ grades awarded in a suite of compulsory introductory courses is a suitable tool to achieve this (3). Additionally, since the anatomy and physiology courses were shown to have a higher predictor importance than other first-year courses, we suggest that it is these courses where it may be most prudent to allocate additional resources. There are a number of ways to improve the academic performance of nursing undergraduates in introductory courses; for example, in this institution we have introduced seminars led by successful higher-level nursing students (e.g., year 3), whereby they explain how their first-year academic experiences informed their decisions in a clinical setting. Also, with regard specifically to the anatomy and physiology courses, we have recruited nurse practitioners to emphasize the importance of anatomy and physiology in modern nursing. The university has also implemented a team approach to teaching introductory anatomy and physiology whereby each staff member can draw on different prior experiences; the team included an experienced physiologist with laboratory research specialties, a cardiac nurse specialist, and an M.D. with pediatric training. This team-teaching approach has shown students that physiology is a discipline that underpins a number of career pathway choices, all of which share a common knowledge, therefore reinforcing the interprofessional nature of a modern healthcare workforce (21).
Courses that build the academic competency of students are often included at the start of an undergraduate curriculum and are designed to prepare a student for the rigors of academic study at a university. In this study, such courses were Knowledge, Enquiry, and Communication, Lifespan Development and Communication, and Health and the Environment, and these were all delivered as a common first semester to all health science undergraduates. We speculate that although courses designed to deliver generic skills showed lower predictor importance compared with the two anatomy and physiology courses, they remain an important inclusion in the curriculum and are required for successful completion of the BHSc(NURS) program. A program that requires an understanding of the human body and its physiology, for example, nursing, often shares common introductory courses in anatomy and physiology with other health disciplines (23). These courses may be based around core principles with their origins in biology (24, 25); however, there is no firm consensus about the content within an introductory course in physiology (27). Novel data presented in this study support our hypothesis that academic achievement in these compulsory human anatomy and physiology courses would be a strong predictor of overall academic achievement.
The higher predictor importance for the human anatomy and physiology courses may be due to a wider range of scores given to students on these courses when compared with the other first-year courses. This may appear counterintuitive, as the two-step clustering procedure standardizes scores for all inputs such that a minimum grade (C−) is scored as 0 and a maximum grade (A+) is scored as 1. However, if the majority of students are awarded the same passing grade (for example, a B grade) in a course, with only a few scores distributed above and below this grade, it is unlikely that such a distribution of grades is capable of predicting high achievers from low achievers when numerous variables are considered together. Also, the higher predictor importance for the courses in human anatomy and physiology may suggest that if a grade point average for the first year is calculated based on all courses taken, an increased weighting should be attributed to the anatomy and physiology courses. This would allow a more accurate tool for discriminating students based on first-year academic achievement, particularly if there was competitive entry into a clinical pathway based on first-year performance. However, data presented in this study also suggested that although first-year academic achievement could clearly differentiate between high- and low-achieving students, high achievers may only achieve on average a slightly better alphabetic grade in a semester 6 capstone course compared with low achievers. Although speculative, we suggest that despite this small (but statistically significant) difference in academic achievement in the Transition to Nursing course, it would be unlikely to differentiate performance in the workplace.
Some fidelity maybe lost in our analysis due to all courses reporting final grades with the nine-point grading system (A+ to C−). However, this was a consistent grading system used throughout the program for all courses. Depending on the course, the final grade represented the combined academic achievement in all learning outcomes for that course. A course educator could choose to use alternative grading systems (e.g., %score, point marking, marks out of 10) to grade an assignment but was also responsible for converting that score into a letter grade on the nine-point scale. All passing marks released to students and all marks recorded on the university data management system used this nine-point scale. In the BHSc(NURS) at this university, summative assessment used to grade students at the conclusion of a course usually contained no feedback on performance, yet these grades are often emphasized by both educators and students, as performance in these assessments was the decisive factor in a student’s progression (4).
A limitation of the current study is that the findings may not be extrapolated to other undergraduate nursing programs; however, the techniques used in the current study endorse an education data mining approach consistent with methods described elsewhere (1). We suggest that these techniques should be applied to other institutions’ data where possible to identify how grades achieved at introductory level courses influence grades achieved at higher levels of study. We speculate that in undergraduate nursing programs, the influence of academic achievement in introductory science subjects (for example, human anatomy and physiology courses) may consistently provide a good indication of academic achievement at higher levels.
In the current study, we uniquely report empirical evidence that supports the notion that obtaining a good pass in introductory anatomy and physiology courses can be used to identify high-achieving students and that this may also translate into a higher grade achieved much later in the program. Courses in anatomy and physiology are consistent inclusions in an undergraduate nursing curriculum (22) and were compulsory requirements in the program reported in the current study. However, although physiology was perceived as essential for a comprehensive undergraduate nurse education (7), knowledge of the fundamental sciences (for example, chemistry) that underpin undergraduate physiology is often limited. Unfortunately, in the current study, we were unable to identify whether students in the HIGH clusters had studied sciences (for example, high school chemistry and biology) before enrollment. It has been reported (12, 14, 17) that knowledge of human physiology was perceived by qualified nurses as important but was limited in its undergraduate delivery. Thus, although physiology is accepted as an essential requirement of undergraduate nursing programs (12), the need to understand the scientific principles that underpin physiological concepts may often be overlooked.
No conflicts of interest, financial or otherwise, are declared by the authors.
S.J.B., S.W., and N.P. performed experiments; S.J.B. analyzed data; S.J.B. and S.W. interpreted results of experiments; S.J.B. prepared figures; S.J.B. drafted manuscript; S.J.B., S.W., and N.P. approved final version of manuscript.
- Copyright © 2017 the American Physiological Society