RESEARCH REPORT
Remote Histology Learning From Static Versus Dynamic Microscopic Images Sylvia Mione,1* Martin Valcke,2 Maria Cornelissen1 Department of Basic Medical Sciences, Faculty of Medicine and Health Sciences, Ghent University, Ghent, Belgium 2 Department of Educational Studies, Faculty of Psychology and Educational Sciences, Ghent University, Ghent, Belgium 1
Histology is the study of microscopic structures in normal tissue sections. Curriculum redesign in medicine has led to a decrease in the use of optical microscopes during practical classes. Other imaging solutions have been implemented to facilitate remote learning. With advancements in imaging technologies, learning material can now be digitized. Digitized microscopy images can be presented in either a static or dynamic format. This study of remote histology education identifies whether dynamic pictures are superior to static images for the acquisition of histological knowledge. Test results of two cohorts of second-year Bachelor in Medicine students at Ghent University were analyzed in two consecutive academic years: Cohort 1 (n 5 190) and Cohort 2 (n 5 174). Students in Cohort 1 worked with static images whereas students in Cohort 2 were presented with dynamic images. ANCOVA was applied to study differences in microscopy performance scores between the two cohorts, taking into account any possible initial differences in prior knowledge. The results show that practical histology scores are significantly higher with dynamic images as compared to static images (F (1,361) 5 15.14, P < 0.01), regardless of student’s gender and performance level. Several reasons for this finding can be explained in accordance with cognitivist learning theory. Since the findings suggest that knowledge construction with dynamic pictures is stronger as compared to static images, dynamic images should be introduced in a remote setting for microscopy education. Further implementation within a larger electronic learning management system needs to be C 2015 American Association of explored in future research. Anat Sci Educ 00: 000–000. V Anatomists.
Key words: histology education; medical education; microscopic anatomy education; multimedia learning; static and dynamic imaging; virtual microscopy; computers in histology education; basic sciences in medical curriculum
INTRODUCTION Histology is a standard morphological division of anatomy within (bio)medical life sciences curricula. In medical education, practical histology training is based on the microscopic study of normal tissue sections. Until now, in several institu*Correspondence to: Dr. Sylvia Mione, Department of Basic Medical Sciences, Ghent University, UZG-6B3, De Pintelaan 185, B-9000 Gent, Belgium. E-mail:
[email protected] Received 12 January 2015; Revised 8 September 2015; Accepted 12 September 2015. Published online 00 Month 2015 in Wiley (wileyonlinelibrary.com). DOI 10.1002/ase.1572 C 2015 American Association of Anatomists V
Anatomical Sciences Education
MONTH 2015
Online
Library
tions classical optical microscopy sessions are still being organized for students with microscopic anatomy in their curriculum (Drake et al., 2014). However, these time-consuming, on-campus sessions are under increased pressure for several reasons. First, universities are faced with ever increasing numbers of students along with decreasing budgets and limited financing of infrastructure (Hamilton et al., 2012; Barbeau et al., 2013; Diaz-Perez et al., 2014; Gatumu et al., 2014). Second, over the last decades, learning content has increased in all scientific disciplines as a result of expanding research (Bergman et al., 2008; Diaz-Perez et al., 2014; Drake, 2014). Moreover, in medical education there is additional emphasis on acquiring social and clinical skills in view of the future profession as a physician, for example, the “Five-star doctor” WHO report (Boelen, 1992, 1996; De Maeseneer, 2003; Anat Sci Educ 00:00–00 (2015)
Frenk et al., 2010; Weatherall, 2011). In accordance with these international developments, many medical schools have redesigned their curricula to develop teaching strategies that are less faculty-intensive (Bergman et al., 2008; Drake et al., 2009). A gradual shift from a discipline- or subject-based and teacher-oriented curriculum to an integrated systems-based and more student-oriented curriculum has been observed in United States and European medical schools (Bergman et al., 2008; Drake, 2014; Gatumu et al., 2014). With the implementation of varied and demanding integrated medical curricula (IMC), reduction in teaching hours in basic medical sciences instruction would likely occur (Dee, 2009; Drake et al., 2009; Gatumu et al., 2014). Particularly laboratory contact hours in histology were affected during which students were traditionally instructed to identify and label microscopic characteristics of human tissues and organs. Practical histology sessions have been decreased in hours invested and/or changed in terms of the instructional media used in various institutions worldwide (Dee, 2009). Especially in remote settings, digital technologies offer additional opportunities for histology instructional approaches, for example, to promote self-directed microscopy learning. However, changes in instructional strategies could affect the cognitive processing of histology knowledge. Consequently, concerns were raised associated to the level of knowledge attained by students graduating from such innovative programs (Bergman et al., 2008).
CURRICULAR REFORM AT GHENT UNIVERSITY: THEORETICAL BASE In the Faculty of Medicine and Health Sciences, an IMC was introduced in 1999–2000 (Van der Veken, 2008). The curriculum redesign was partly founded on cognitivist and constructivist learning theory. Cognitivism deals with learning through internal information processing in memory. Constructivism is an educational viewpoint extending cognitivism with a social dimension and an emphasis on authentic learning (Valcke, 2010). Both theories assume that new information has to be linked to prior knowledge through elaboration. According to cognitivist theory, elaboration of information strengthens the mental model (schema) particularly when multiple external representations (MER) are presented (Ainsworth, 2006). Following the multimedia principle of cognitive theory of multimedia learning (CTML) (Mayer, 2001, 2003), combining imaging and textual materials has proven to be beneficial for knowledge construction (Mayer, 1989, 2001). This is particularly true in highly visual domains such as histology, where images play a major role as nonlinguistic representations (NLR) (Ainsworth, 2006). Furthermore, the active learning principle in constructivism emphasizes that knowledge construction is based on active involvement of learners (Jonassen, 1999). This implies hands-on manipulation of authentic learning materials. According to an IMC, the interdisciplinary organization of learning content resulted in a schedule with several large units (modules) discussing the core principles of different individual courses (histology, anatomy, physiology) pertaining to a central theme (Nazian and Stevenson, 2013) or organ system. As a consequence, histology was no longer instructed as a stand-alone course. 2
Also the instructional strategy was modified. Following the move to self-directed learning in medical curricula, remote histology was introduced as a formal substitute for the decrease in contact hours. Time restrictions led to the replacement of traditional classroom microscopy sessions with self-directed study activities outside the laboratory (offclassroom). State-of-the-art digitization led to the adoption of two different imaging solutions suitable for remote microscopy learning. The digitized microscopy images can be presented to the user in either a static or dynamic way. A static representation implies that learners are presented with fixed photorealistic snapshots of microscopic structures to recognize and name. Another approach to practical histology is the use of total slide scanning with a virtual microscope. The learners are actively engaged while viewing dynamic virtual slides to search and compare different structures for similarities and differences across the entire length of the slide. Building on the former theoretical base, the following hypothesis is put forward: remote histology learning on the basis of dynamic microscopic images will result in significantly higher learning performance as compared to learning on the basis of static microscopic images.
HISTOLOGY AT GHENT UNIVERSITY: LEARNING ENVIRONMENT Histology at Ghent University is taught during the preclinical years of Bachelor in Medicine. According to a systems-based IMC, the subject matter of the Organ Histology course is embedded within modules along all years of Bachelor in Medicine: I. Locomotor system and integument; II. Nervous system and special sense organs; III. Gastrointestinal and endocrine system; IV. Circulatory, urinary and respiratory system; V. Reproductive system, and VI. Immune system. The sum of contact hours devoted to theory in histology in all modules is 50 hours. Students are presented with printed course material, as well as with color slide projections as PowerPoint presentations (Microsoft Corp., Redmond, WA). Theory lessons are exclusively taught through plenary ex cathedra lectures by a single teacher. Interactive teaching modalities such as team-based learning, problem-based learning, and case-based learning are currently not applied. In laboratory instruction, traditional microscopy sessions are replaced with two imaging solutions, both suitable for remote self-directed learning: static versus dynamic microscopy images. Practical histology lectures were previously organized with viewing of glass slides under light microscopes divided as 2-hour sessions per organ system. Now, a flipped classroom concept is followed where preparatory extramural activities are encouraged: students are required to study the images during their private time, yet with the aid of attention-directing cues. Afterward, a plenary response lecture is organized in relation to each module. Issues associated with studying the images are discussed during this 75-minute lecture. To evaluate both theoretical and practical histology knowledge, a summative multiple choice examination is set up for each module. Within the examination, part of the multiple choice questions (MCQ) assesses practical histology knowledge on the basis of several printed images of microscopic structures. Mione et al.
MATERIALS AND METHODS Research Design A retrospective study was set up to compare the differential impact of static versus dynamic microscopic pictures on learning performance, involving two cohorts of second-year Bachelor in Medicine students. In the second year Bachelor in Medicine, histology is an integrated part in three modules: Module I: Locomotor system and integument; Module II: Nervous system and special sense organs, and Module III: Gastrointestinal and endocrine system. For the purpose of this study, all examination questions in these modules pertaining to practical histology were used as a measure of microscopy learning performance. No preceding selection of examination questions was applied. Given the naturalistic character of this study, students’ answers were collected from the standard online data base. All data were anonymized prior to the analysis. Both instructional and test environment were formal parts of the histology course, hence no individual informed consent form was required. However, permission to investigate student test results was obtained from the Ghent University Ethics Committee (code EC UZG 2007/141).
Participants
Static format delivery of histology images. This image shows a reprint of the static image in the histology workbook.
All second-year Bachelor in Medicine students belonging to the two consecutive cohorts were involved in the study: students in Cohort 1 (n 5 253) were involved in the static learning condition during the academic year 2007–2008 and students in Cohort 2 (n 5 228) were involved in the dynamic learning condition during the academic year 2008–2009. Due to the naturalistic design of the study, differences in number of respondents were observed between both conditions. The two cohorts were considered equivalent with respect to entry qualifications: they both passed the same entry examination before the first Bachelor year and were presented with the same first year Bachelor in Medicine curriculum. Proof of equal prior knowledge is provided by a quantitative criterion as well. Given the fact that no prior records pertaining to histology are available for these students to serve as a comparative parameter, a proxy knowledge base (achieved score in a preceding cytology course) was selected to study the comparability of both cohorts involved in the study. This measure was also used as a baseline in the analysis of covariance to investigate the differential impact of studying in either the static or dynamic microscopy format. The students in both cohorts participated in the same summative examinations related to the three modules. The students who could not participate in all three examinations were excluded from the analysis. Also the results of those students who studied histology in their previous curriculum or who repeated the examinations were excluded from the dataset. Incomplete participation in the three examinations was related to individual circumstances (e.g., illness) and did not introduce bias. Moreover, larger flexibility in planning student trajectories was allowed in the new curriculum, which resulted in varying numbers of students opting for a particular follow-up course during and over the years. Of the original 481 enrolled students, 364 complete cases remained after applying the described exclusion criteria, 190 in Cohort 1 and 174 in Cohort 2. Anatomical Sciences Education
Figure 1.
MONTH 2015
Microscopy Learning Content: Static Versus Dynamic Material During both academic years 2007–2008 and 2008–2009, an identical instruction program was adopted, consisting of a theoretical and a practical part. The same histology theory was taught to the second-year Bachelor in Medicine students, and self-directed learning was used for practical histology in both cohorts. However, the key difference between both cohorts was in the format of the practical histology images made available: static versus dynamic digitized microscopy images. Figure 1 shows static condition and Figure 2 dynamic condition for image display. In both cohorts, the images were accompanied with attention-directing cues. Students in the static learning condition (Cohort 1) worked with static digitized pictures of microscopy structures. These pictures were made from traditional glass slides using a digital color camera, an optical microscope, and the software program Photoshop CS3 (Adobe Systems Incorp., San Jose, CA) for picture editing. In this way, printed pictures were studied that covered the complete histology of the human body. No marks or annotations were made on pictures by means of arrows or other signs. Attention-directing cues were provided next to the printed histology images in the student workbook. The cues consisted of short questions to guide the viewing of the pictures. These informative guiding questions drew attention to image features. The images were available at any time thanks to the personal purchase of the workbook. Students in the dynamic learning condition (Cohort 2) were presented with digital dynamic images about the same organ systems as covered in the workbook. However, these images were studied on a computer screen with OlyVIA virtual microscopy software program, version 2.1 (Olympus Corp., Tokyo, Japan). The virtual slides as well as the viewer 3
Figure 2. Dynamic format delivery of histology images. A, computer set-up in the histology laboratory; B, screen shot of the computer displaying the same image as in Figure 1 in a dynamic format.
program were accessible at all time through the Ghent University online application platform “Athena”. The OlyVIA microscopy software allowed for microscopic modalities as panning (navigating) and zooming (magnifying). These active manipulations in a dynamic format were possible as the digital images were created by scanning a complete tissue section. Consequently a more extended viewing and exploration area was offered as compared to the static printed images that center on a limited area of a tissue section. No marks or annotations were added to the digital images on screen. Attention-directing cues were presented in a concise guide and consisted of short slide descriptions with key-words, drawing attention to image features present on the slides.
Data Collection and Research Instruments As explained above, a proxy knowledge base was selected to study the comparability of both cohorts involved in the study. Each student’s score in a preceding cytology course taught in the first year Bachelor in Medicine was used as a prior knowledge measure, as well to control for individual differences between students prior to following the histology course. Assessment was organized after each module. Answers to the image-based histology questions from both cohorts were used as a learning performance measure. These questions were part of a larger module examination also focusing on anatomy and physiology. Image-based histology questions were the same in both cohorts. Consecutively, learning performance was determined after studying Module I (T1), Module II (T2), and Module III (T3). Contamination over time between cohorts was unlikely due to the large in-between time span. The image-based histology questions were presented as multiple-choice questions and were linked to static color images of microscopy structures. Module I examination included two image-based questions (out of 7 histology questions), Module II one image-based question (out of 9 histology questions), and Module III contained four questions with 4
images (out of 10 histology questions). As a result, about 25% of the histology questions were image-based. All the above-mentioned questions were used as a performance measure in this study. The number of questions varied depending on the importance of histology in the three modules. The link between the specific questions and the respective modules contributed to the content validity of the performance test. Each of the images represented a microscopic detail that was to be recognized by the student. Naming the microscopic detail as requested was carried out by selecting one of the four possible answer choices. Individual student answers were manually assessed on the basis of a nonconventional scoring method that captures information about a student’s mastery level as reflected in their responses to a question. Therefore, a gradual scoring scale was used, varying from 0 (no answer) to 2.5, 5, 7.5, or 10 points, depending on the choice for a less, more or completely correct alternative. No negative marks were given. The use of such partial-credit scoring method is a known approach for assessing complex skills in medical higher education (Lesage et al., 2013). Assigning a higher score to a more complete answer allows for differentiation in competence assessment. Partial scores provide a more balanced insight in student performance. Figure 3 exemplifies the scoring approach in relation to the microscopy image of a peripheral nerve from the Module II “Nervous system and special sense organs” examination. For each module examination, a standardized sum-score (maximum 10) was calculated. Subsequently, the standardized scores of the three module examinations were added (T1 1 T2 1 T3), yielding a total image score per student (maximum 30). This reflected the overall practical histology performance score, based on all seven image scores.
Statistical Analysis All statistical analyses were performed using SPSS Statistics, version 21 for Windows (IBM Corp., Armonk, NY). First, an Mione et al.
split up considering the research conditions and student background variables, such as gender and level of performance. The results show that both cohorts hardly differ in mean age. As to prior knowledge scores (achieved scores in a preceding cytology course), the results of an independent two-samples t-test reflect no significant differences in prior knowledge between both student cohorts (t 5 1.02, P 5 0.31). Hence, both cohorts can be considered as being equivalent in prior knowledge at the start of this study. The mean total image score (T1 1 T2 1 T3) in both cohorts suggests a differential impact of studying either static or dynamic images during self-directed learning.
The Differential Impact of Remote Learning with Static Versus Dynamic Microscopy Images
Figure 3. Microscopic image of a peripheral nerve as questioned in Module II “Nervous system and special sense organs” examination. Between brackets are reported the incremental scores used in the partial-credit scoring method to assess student’s mastery level.
independent two-samples t-test was used to study the comparability of both student cohorts. Second, analysis of covariance (ANCOVA) was applied to study differences in mean practical histology performance scores and to consider differences in prior knowledge at the start of the histology course. In an additional analysis of variance, possible differences resulting from student characteristics such as gender and performance level were considered by entering these variables in the ANOVA model and by checking for interaction effects. High and low performing students were distinguished on the basis of the median in the prior knowledge score (low when score 19.5 and high when score > 19.5). Analysis assumptions for ANOVA were checked. A significance level of P < 0.01 was put forward. Cohen’s d was calculated as an indicator of the effect size.
RESULTS Descriptive Results Descriptive results have been summarized in Table 1. Mean and standard deviation of research variables are included and Anatomical Sciences Education
MONTH 2015
Analysis of covariance was applied to compare overall practical histology scores (T1 1 T2 1 T3), taking into account differences in prior knowledge (T0). A beginner’s equality for acquiring new knowledge in the given subject area was assured by the input of the prior knowledge score (T0) as a covariate. The setting based on static versus dynamic learning material was entered as the factor in the analysis. The results of the analysis of covariance show significant differences in total image score (T1 1 T2 1 T3) between both learning conditions: F (1,361) 5 15.14, P < 0.01. With a Cohen’s d value of 0.42 (>0.3), it can be concluded that the significant difference represents a small to a medium effect size. In an additional analysis of variance, two concomitant student characteristics that could interact with the research conditions were taken into account. To fine-tune the analysis, both gender and student performance levels were added to the ANOVA model while studying their (1) main effect and (2) interaction effects with the learning condition. The results show a significant main effect of learning condition (F (1,356) 5 13.17, P < 0.01), a nonsignificant main effect of gender (F (1,356) 5 0.85, P 5 0.36) and a significant main effect of student performance level (F (1,356) 5 16.95, P < 0.01). However, the interaction effect between learning condition and gender is not significant (F (1,356) 5 0.54, P 5 0.46), neither was the interaction effect between learning condition and student performance level (F (1,356) 5 0.79, P 5 0.37) nor the interaction effect between learning condition, gender, and level (F (2.356) 5 1.08, P 5 0.34). Practical histology scores were significantly higher with dynamic microscopy images, regardless of student’s gender and performance level. Main and interaction effects are summarized in Table 2.
DISCUSSION In this observational study of students’ achievement scores, the dynamic approach is compared to the static method of self-directed microscopy learning. Although the use of static microscopy images is not a common practice in on-campus histology education, it is considered as a historic control in an autonomous learning set-up. Static microscopy images are frequently used in those settings where individual optical microscopes and glass slides are unavailable not only during home-study (Merk et al., 2010) but also in particular environments, curricula, or institutions with financial and/or 5
Table 1. Descriptive Characteristics of Cohort 1 and Cohort 2 Static condition (Cohort 1) Characteristics Number of participants (n)
Dynamic condition (Cohort 2)
Total
Male
Female
High Level
Low Level
Total
Male
Female
High Level
Low Level
190
77
113
86
104
174
67
107
84
90
Age mean (6SD)
19.14 (60.51)
19.13 (60.64)
Prior knowledge score (T0)a, mean (6SD)
19.73 (63.35)
19.42 (63.12)
19.94 (63.50)
22.87 (61.75)
17.13 (61.76)
20.11 (63.80)
19.84 (64.02)
20.29 (63.66)
23.36 (62.35)
17.08 (61.90)
Total image score (T1 1 T2 1 T3)a, mean (6SD)
20.12 (65.26)
19.86 (65.29)
20.29 (65.26)
21.62 (64.92)
18.88 (65.24)
22.42 (65.63)
21.87 (65.72)
22.76 (65.57)
23.48 (65.85)
21.43 (65.26)
a
Maximum score 5 30 points.
organizational constraints (Tian et al., 2014). Moreover, static microscopy images are still omnipresent in commercial histology books and atlases as well as in printed course materials (Scoville and Buskirk, 2007). On the other hand, dynamic virtual slides although very popular (Dee, 2009; Helle et al., 2013) have still not been implemented everywhere in histology education (Drake et al., 2009, 2014). The study at hand makes an evidence-based contribution to the static practice of imagery actually used when hands-on microscopy education is beyond the reach. The results of this study show that self-directed learning performance on the basis of dynamic images—supported with the virtual microscopy software—is significantly higher as compared to learning from static photographic images. This means that students acquire microscopy recognition competencies better when studying in a virtual microscopy environment. These findings are in line with the educational research studies in the domain of multimedia learning and assumptions derived from cognitivist learning theory. First, students in both conditions are presented with multiple external representations (MER) (Ainsworth, 2006), in which textual infor-
mation in the syllabus is supplemented with visual illustrations, being microscopy images. This strengthens the organization of information in memory, since these nonlinguistic representations (NLR) support schema construction (Mayer, 1989, 2001; Ainsworth, 2006). Also Vorstenbosch et al. (2013, 2014) argue that the use of images is fundamental to learning and assessing anatomical knowledge as visual cognition is complementary to verbal cognition in information processing. More specifically in the domain of microscopy learning, authors consider the combined use of visual (image) and conceptual (text) material as supportive for a student’s cognitive process (Braun and Kearns, 2008; Nivala et al., 2012). However, static images demand a rather basic active involvement of the learner in knowledge construction. Following Bloom’s taxonomy (Krathwohl, 2002), the learning activity of the student remains restricted to the first behavioral level: Remembering because of the emphasis on “recognizing” and “naming.” By contrast, in the dynamic learning condition students are not presented with preselected NLR but with representations they have to explore themselves (virtual slides). According to cognitivism, this should
Table 2. Statistical Results of the Independent Two-Samples t-Test and the Analysis of Variance Between Both Cohorts Factor Prior knowledge score (T0) Total image score (T1 1 T2 1 T3)
t
F
1.02
P-value
Cohen’s d
0.31
Main effect: Condition
F (1,356); 13.17
Interaction effectsa Condition–gender Condition–level Condition–gender–level
F(1,356); 0.54 F(1,356); 0.79 F(2,356); 1.08
