A validated instrument measuring students' perceptions on plastinated and three-dimensional printed anatomy tools

INTRODUCTION

Cadaveric dissection is often considered a defining feature of anatomy teaching and learning (Aziz et al., 2002; Sugand et al., 2010; Ghosh, 2017a). However, there has been a growing interest in various new anatomy tools such as three-dimensional printed (3DP) models (Lim et al., 2016; Ye et al., 2020; Chytas et al., 2020b), Augmented Reality (AR; Blum et al., 2012; Chytas et al., 2020a), Virtual Reality (VR; Nieder et al., 2000; Nicholson et al., 2006; Kockro et al., 2015; Kurul et al., 2020), and mobile-based applications (Mogali et al., 2019) to support the effective learning of anatomy (Sugand et al., 2010). This is partly due to the lack of sustainable body donation programs, cultural, religious barriers, and financial constraints to maintain the cadaveric laboratory facilities (Lim et al., 2016; Ye et al., 2020). Consequently, modern medical schools are replacing cadaveric dissection with modern tools such as plastinated specimens, 3DP, and virtual three-dimensional (3D) anatomy models to support gross anatomy teaching and learning. There are valid tools that measure students' perceptions of traditional anatomy education (Smith & Mathias, 2010; Hadie et al., 2017, 2021), but inventories that apply to modern methods using emerging anatomy tools are unavailable. Therefore, a new and valid instrument is needed to evaluate their utility and students' perceptions and attitudes toward the novel anatomy tools. This information is critical for decision-makers and educators to understand the educational benefits of modern tools and how these resources influence students' learning and improve anatomy education in a non-cadaveric dissection setting.

Traditional cadaveric dissection and prosected specimens are established tools for mastering anatomical structures and appreciating their spatial relationships (Aziz et al., 2002; Sugand et al., 2010; Ghosh, 2017a). Therefore, it is critical for modern tools (plastinated, 3D printed and others) to be produced in high fidelity with precise anatomical details and their relationships. Contemporary anatomy tools utilize innovative technologies and are usually based on the scans of actual patients (Nicholson et al., 2006), donated cadavers (McMenamin et al., 2014), plastinated specimens (Radzi et al., 2020), or computer-based 3D modeling (Brazina et al., 2014). These novel tools defer from the traditional cadavers in terms of the representation, look, and feel (Lim et al., 2016; Mogali et al., 2018), as well as their physical and mechanical properties (Jones, 1997; Sugand et al., 2010; Fruhstorfer et al., 2011). Consequently, they provide the students with different visualizations, interactions, and diverse learning experiences compared to traditional cadaveric resources (Levinson et al., 2007; Sugand et al., 2010). Students' views on new teaching and learning modalities are necessary as they help educators understand and evaluate the effectiveness of teaching and learning processes (Eagleton, 2015). Hence, an instrument is needed to evaluate the novel anatomy tools and methods from students' perspectives.

Three-dimensional printed models are lauded in terms of their quality, low cost production, ability to customize in multi-colored and multi-material structures and provided positive learning satisfaction in students (Lim et al., 2016; Mogali et al., 2018; Chytas et al., 2020b; Chen et al., 2020; Radzi et al., 2020). Plastinated specimens, although more expensive than 3DP models, are found to be an effective alternative learning tool, as they are associated with higher student performance in anatomy examinations compared to the use of cadavers (Latorre et al., 2007; Baker et al., 2013; Nguyen et al., 2019). This may be due to the organization of layered anatomy structures in plastinated specimens that allows the students to analyze and explore the deep anatomy structures (Haque et al., 2017). Moreover, the plastination technique will enable specimens to be easily distinguished from the wet cadaveric materials as they are dry and odorless, thus allowing easy access, storage, and longevity (von Hagens et al., 1987; O’Sullivan & Mitchell, 1995; Bhandari et al., 2016). Both 3DP and plastinated specimens eliminate the students from exposure to formalin odors, a feature of the traditional cadaveric dissection (Ravi & Bhat, 2011; Nguyen et al., 2019). Overall, students are satisfied and agreed on the educational value and the use of plastinated specimens to aid anatomy teaching (Latorre et al., 2007; Haque et al., 2017), yet criticized the inadequate tactile and emotional learning experience that was believed to come from cadaveric dissection (Fruhstorfer et al., 2011; Azu et al., 2012).

Aside from plastinated and 3DP models, other studies reported students' experiences with innovative virtual anatomy tools such as AR and VR compared to traditional materials (Nicholson et al., 2006; Codd & Choudhury, 2011; Ferrer-Torregrosa et al., 2015). These tools provided an enjoyable learning experience, increased immersion, and engagement (Jamali et al., 2015; Küçük et al., 2016; Moro et al., 2017; Chytas et al., 2020a; Duarte et al., 2020). At the same time, these tools found to be equally effective or better than learning from a textbook (Stepan et al., 2017; Zinchenko et al., 2020), and they can be studied from multiple views (Wismer et al., 2018). VR interventions were shown to have a positive impact on learning and promoted long term knowledge retention, albeit higher cognitive loading (Andersen et al., 2016; Ekstrand et al., 2018; Checa & Bustillo, 2020; Gloy et al., 2022), whereas AR helped to decrease the cognitive load compared to the cross-sections (Henssen et al., 2020). Training under a higher cognitive load seemed advantageous as it helped enhance the adaptation of tasks into a real scenario (Sankaranarayanan et al., 2020). The AR and VR tools allowed learners to manipulate the anatomical structures in a virtual world, which was a beneficial interaction, particularly to low spatial ability students (Jang et al., 2017). Although these virtual learning resources provide new learning experiences for the students, they are less preferred in real educational settings due to technological complexities and lack of tactile sensation than the physical models (Maresky et al., 2019; Birbara et al., 2020; Skandalakis et al., 2022). More specifically, physical models provide a distinct advantage for novice learners. They found it easier to orient anatomical structures based on the touch and feel than in virtual space (Garg et al., 2002; de Faria et al., 2016; Birbara & Pather, 2021). In summary, based on the studies mentioned above, students perceptions of the modern anatomy tools are obtained utilizing the non-validated surveys instead of a psychometrically valid and reliable instrument. Hence, this study is proposed to build a reliable instrument that may help educators and students improve anatomy teaching and learning.

In addition to modern anatomy tools, medical schools have been implementing various unique pedagogical methods such as problem-based and team-based learning (TBL; Nieder et al., 2005; Dolmans et al., 2015; Rajalingam et al., 2018). These teaching methods can be supplemented with other techniques such as note-taking guides and case studies (Kaufman, 2003). Together, they promote the development of students' knowledge, skills, and attitudes (Kaufman, 2003). However, more information is needed to gain insights into the students' learning process utilizing contemporary anatomy tools. Therefore, through a detailed literature review, the factors that affect and shape how information is received, understood, and influenced the learners' responses to the subject they learned were identified. The three prevalent factors ultimately identified perceived learning satisfaction (Chang & Chang, 2012; Topală, 2014; Eagleton, 2015; Wu et al., 2015), self-efficacy (Bandura, 1986, 1997; Burgoon et al., 2012), and humanistic values (Rizzolo, 2002; Guo et al., 2020; Souza et al., 2020).

Learning satisfaction refers to the feeling of joy when learning, cohesive levels between a person's expectations and actual experiences (Chang & Chang, 2012) and/or behavior motivated by a learner's ‘needs and likings’ (Deci et al., 1996). Wu's research team suggested five crucial indicators of learning satisfaction: course contents, administrative service encounters, teaching modalities, learning convenience, and environment (Wu et al., 2015). These factors are shown to have a positive influence on continuing learning intention and participation. The factors that promote learning satisfaction were the interactions between the perceived learnability, social learning, and effectiveness (Hu et al., 2007; Eagleton, 2015). Students' interest can influence the learning in the subject, their perceptions of its use, and relevance to their future jobs (Eagleton, 2015). Besides these, social interactions (peer-to-peer; student-to-teacher) contribute to learning satisfaction (Wu et al., 2010; Eom & Ashill, 2016). Feedback and active discussion with peers and instructors during the practical or lectures enhanced the learning and helped them develop deeper meaning and understanding of the subject (Topala & Tomozii, 2014; Eagleton, 2015; Dolmans et al., 2015; Rajalingam et al., 2018). As anatomy is a more visual subject, novel learning tools can add to learning satisfaction as they offer structures in different colors, textures, representations, and visualizations (McMenamin et al., 2014; Lim et al., 2016; Mogali et al., 2018). These aspects contribute to creative learning opportunities that may motivate students to engage with learning materials, meet their learning needs, and attain their goals. Several studies showed associations between the students' learning satisfaction and academic achievement. Positive perceptions are associated with an increased academic achievement (Meltzer et al., 2004; Al-Qahtani, 2015; Ahmed et al., 2018), whereas negative perceptions are linked to poor performance (Ferreira & Santoso, 2008; Saputra et al., 2020; Oducado & Estoque, 2021). Therefore, factors that positively influence learning satisfaction need to be in place to enhance students' learning achievement.

Self-efficacy is defined as the individuals' judgment on their abilities to effectively execute their behavior required to attain their desired results (Bandura, 1986, 1997; Farey et al., 2018). From the perspective of anatomy education, it is related to individual beliefs on dissection of cadavers, learning anatomy terminologies, structures, and applying knowledge to clinical situations (Schunk, 1991; Burgoon et al., 2012). Self-efficacy is considered the key source of the learners' motivation and influences the selection and adaptation of study methods (Tembo & Ngwira, 2016). In general, medical students recognize the importance and relevance of anatomy for future clinical training and medical practice (Ebomoyi & Agoreyo, 2007; Alam, 2011; Gupta et al., 2014). Students perceive anatomy as a difficult subject compared to the other basic sciences due to its content-heavy curriculum (Alam, 2011; Gupta et al., 2014). It involves mastering new terminologies and numerous complex relationships (Ebomoyi & Agoreyo, 2007; Alam, 2011). These, in turn, affect their approach toward the subject. Those with higher perceived self-efficacy are highly motivated, use deep study approaches, and undertake difficult tasks to achieve their goals (Simons et al., 2004; Bråten & Olaussen, 2005). On the other hand, those with lower perceived self-efficacy select more manageable tasks and tended to use superficial study approaches (Burgoon et al., 2012; Tembo & Ngwira, 2016). Abdel Meguid et al. (2020) measured self-efficacy in the anatomy education of medical, dental, and chiropractic students, and showed an increased self-efficacy associated with significantly higher performance in examinations. Evidence suggests mixed results regarding anatomical self-efficacy between males and females (Burgoon, 2008; Tembo & Ngwira, 2016). Studies found that male students possess significantly higher anatomical self-efficacy than females (Burgoon, 2008; Anderton et al., 2016; Farey et al., 2018). In contrast, Tembo and Ngwira (2016) found no sex differences in medical students' anatomical self-efficacy perceptions (Tembo & Ngwira, 2016). Therefore, it has become essential to research students' self-efficacy, and its relation to their performance as medical students must master the knowledge and be more self-confident to perform clinical procedures in real-life situations.

Humanistic values include respecting and caring for anatomy models and cadaver specimens, reflecting on the life and death of the donor, and considering ethical obligations that may influence future doctor-patient relationships (Weeks et al., 1995; Ghosh, 2017b; Guo et al., 2020). Anatomy laboratory engenders a natural point to educate humanistic values, where students meet silent mentors (donated cadavers) to learn formal anatomy curriculum (Rizzolo, 2002; Ghosh, 2017a; Guo et al., 2020). As a result, a growing number of modules for humanistic education through the anatomy curriculum has been developed (Rizzolo, 2002; Guo et al., 2020; Souza et al., 2020). Instilling respect and compassion in students toward the donors, best practices in dissection, and proper approaches to cadaveric materials became critical in enforcing humanistic values for the training of future healthcare professionals who need to be empathetic and caring (Weeks et al., 1995; Ghosh, 2017b; Guo et al., 2020). Unfortunately, there is an increasing difficulty in accessing donors/cadavers in many parts of the world (McMenamin et al., 2014; Ye et al., 2020). Hence, this creates the need for additional instruments to measure this humanistic aspect. At the same time, humanistic values are rarely validated, and limited reports document how students felt toward non-conventional tools derived from actual patients or cadaveric materials.

In medical education, several instruments such as Dundee Ready Education Environment Measure (DREEM), Approaches to Studying Inventory for Students (ASSIST), and Anatomy Education Environment Measurement Inventory (AEEMI) have been developed. These instruments measure aspects of the learning environment that encompasses students' satisfaction with learning resources, quality of activities, teaching, and engagement (Roff et al., 1997; Al-Hazimi et al., 2004; Smith & Mathias, 2010; Hadie et al., 2017, 2021; Ahmed et al., 2018). In addition, AEEMI included students' perceptions on various domains such as anatomy instructors, the significance of the anatomy, anatomy subject, learning resources, and efforts to learn anatomy and histology resources (Hadie et al., 2017, 2021). However, although the instruments mentioned above apply to the clinical environment (DREEM) or traditional anatomy education (ASSIST, AEEMI), they do not cover students' perceptions of non-conventional tools (plastinated specimens, digital tools, and 3D printed models). This implies that important information pertinent to such tools would not be captured if these instruments were applied.

Hence, this study aimed to create an instrument to measure the students' perceptions of anatomy learning satisfaction, self-efficacy, and humanistic values toward modern tools, particularly plastinated specimens and 3DP models. It employed multivariate statistical analysis, exploratory and confirmatory factor analysis (CFA) to generate structural equation models (SEM), in order to demonstrate the best-fit model with good internal reliability applicable to plastinated and 3D printed anatomy models.

METHODS

RESULTS

Sixty-three out of 163 (38.7%) first-year undergraduate medical students participated in the cardiac session. Thirty-two of the same students participated in the neck learning session. One new student participated in the neck session, which added to a total of 33 participants. In the cardiac session, 32 (16 males and 16 females) and 31 participants (17 males and 14 females) were randomly assigned to plastinated and 3DP groups, respectively. In the neck session, 18 (9 males and 9 females) and 15 (9 males and 6 females) participants were in plastinated and 3DP groups. This translated to a total of 96 responses (50 responses from the plastinated group, 46 responses from the 3DP group), which consisted of 51 males (53.1%) and 45 females (46.9%). All participants reported that they did not have any formal training in anatomy before this study. The participants' mean age was 19 years old at the time of the investigation.

Confirmatory factor analysis

Three different multi-dimensional factor structure models (three, four, and five factors) were computed from the exploratory factor analysis. The three and five-factor models did not give the acceptable goodness of fit indices (Figures S2 and S4). The four-factor model containing 19 items (Table 1) was chosen as the final assessment after the removal of 13 out of 32 items (extracted from EFA, Table S2). This was based on high modification indices (MI > 15) and low factor loadings (loadings ≤ 0.5) (Awang, 2015). This model measured the students' perceptions and attitudes on learning satisfaction (8 items), self-efficacy (5 items), humanistic values (4 items) and a new factor, limitations of learning tools (2 items) using plastinated and 3DP specimens. The mean ratings of the five-point Likert scale results (1 = strongly disagree to 5 = strongly agree) from students' perceptions of individual survey items were shown in Figure 1. The mean ratings and standard error of the mean for each of the individual factors are as follows: learning satisfaction—4.10 (±0.02); self-efficacy—3.59 (±0.03); humanistic values—3.95 (±0.08) and limitations of learning tools −1.81 (±0.17).

TABLE 1. Standardized factor loading and Cronbach's alpha for the four factors in 19 items model

Potential variables or factors	Survey items	Standardized factor loading	Cronbach alpha
Learning satisfaction	Q7. I am able to visualize the three-dimensional structures more readily using this learning method	0.72	0.952
	Q12. This learning method made the anatomy subject more interesting	0.78
	Q13. This learning method motivated me to learn more	0.77
	Q16. My learning outcomes are enhanced using this learning method	0.84
	Q17. I enjoy using this learning method	0.93
	Q18. My learning anatomy experience is enhanced using this learning method	0.90
	Q19. Overall, I am satisfied learning anatomy using this learning method	0.92
	Q28. This learning method is worthwhile for learning anatomy	0.89
Humanistic values	Q33. I treated given anatomy tools with great respect as for with the real cadaveric specimens	0.82	0.847
	Q36. I always treated the given models with great care to protect the structures in the models	0.83
	Q37. After learning anatomy from the given models, my sense of empathy is developed/enhanced	0.72
	Q44. Overall, my respect toward anatomy tools has increased after this session	0.69
Self-efficacy	Q22. I am comfortable identifying structures of the model by name and function	0.68	0.821
	Q24. I feel more confident to apply anatomical knowledge in clinical situations related to the model	0.86
	Q25. I feel better prepared tackling clinical anatomy problems related to the model	0.81
	Q26. This learning method made me more confident about my ability to self-learn	0.53
	Q31. I am able to recall anatomy knowledge during the assessment	0.57
Limitations of learning tools	Q40. I did not have much interaction with the given models due to fear of breaking the structures	0.97	0.743
	Q41. I feel my learning is impacted due to fear of breaking the structures on the models	0.62

The four-factor model was generated with the following fit indexes (Figure S3)—Chi square indicating the overall fit $x$² (147) = 211.568, root mean square error of approximation (RMSEA) that signified the closeness of fit or error per degrees of freedom, (RMSEA = 0.068 at 90% confidence interval), Incremental fitness such as Bentler comparative fit index, CFI = 0.946 and Tucker Lewis index, TLI = 0.937, root mean square residual (RMR) that represented the difference between the residuals of the covariance matrix and the model (RMR = 0.064) (Kline, 2015; Xia & Yang, 2019). The chi square ratio of $x$²/df was less than 2.0 and the correlation coefficients values between the constructs were less than 0.8, thus demonstrated adequate validity for the proposed four-factor model (Ullman, 2019).

Reliability analysis demonstrated that the Cronbach alpha for each of the individual factors was above 0.7, indicating a good level of internal reliability of the items (Table 1).

DISCUSSION

This study has developed a valid and reliable 19-item, four-factor construct to assess the medical students' perceptions and attitudes toward the 3DP models and plastinated specimens. This was needed to understand better the utility of these tools and how well they supported the students in learning anatomy. To the best of the authors' knowledge, this was the first instrument created to measure the factors such as learning satisfaction, self-efficacy, humanistic values, and limitations of learning tools using plastinated and 3DP models. The validity and reliability of the factors were confirmed by a robust mathematical approach using factor analysis.

In terms of validity, all the items' factor loadings were varied between 0.53 and 0.97, which were considered good (Field, 2009). This supported the internal structure validity of the construct, such that higher factor loadings reflected a stronger correlation with survey items represented by the factor. This was similar to other anatomy inventories that measured learning environments (Hadie et al., 2017, 2021) and teachers' attitudes toward using 3D printing in education (Gürer et al., 2019). In addition, the current metric obtained a good discriminant validity. In other words, the measurement model was free from redundant or unwanted items, thus supported the existence of four independent factors. The correlation between the exogenous constructs did not exceed 0.85, indicating no multicollinearity (Awang, 2015). The good internal structure and discriminant validity of this construct were attributable to the careful selection and development of the survey statements. In summary, the developed instrument consisted of a good pool of survey items.

In addition to validity, the four-factor construct achieved good internal consistency (α values = 0.74 to 0.95). In other words, the survey items and their factors were homogeneous within the construct. This ensured the reliability of the factors and its subsidiary survey items, thus representing a reliable instrument. The α values of this study were similar to other anatomy learning inventories (Smith & Mathias, 2010; Burgoon et al., 2012; Hadie et al., 2017, 2021). Therefore, the developed instrument demonstrated the evidence of a psychometrically valid and reliable tool for measuring students' perceptions of the use and educational advantages of the plastinated and 3D printed tools.

The integration and effective use of modern tools in anatomy education depended on students' perceptions of new teaching and learning methods, and the perceived usability and acceptability. This inventory could identify the strengths and weaknesses of anatomy education, particularly in places where anatomy instruction highly depended on plastinated and 3D printed tools. In other words, this survey could be implemented at different time points in the anatomy curriculum, such as at the end of the module, semester, or course of study. It helps the students to reflect and provide feedback on their own learning experiences with non-traditional teaching methods. This information would be helpful for educators to integrate the novel learning tools better, fill learning gaps, build strengths, and improve the teaching and learning of anatomy. Although the instrument was created based on the students' responses on the plastinated and 3DP specimens, it might be extended to other modern tools such as VR or AR. However, it would be required to repeat the factor analysis to confirm the construct validity, particularly when used in non-plastinated and 3D printing contexts.

The first factor assessed the elements that promoted students' learning satisfaction. Students would be satisfied with learning when their needs and expectations were met or exceeded to achieve self-desires and values (Checa & Bustillo, 2020). Learning satisfaction was perceived as a complex and multifactorial process. Nonetheless, factors that determined the learning satisfaction were associated with the quality of instruction, student-instructor relationship, student-student interaction, motivational environment, quality of learning resources, and activities (Hu et al., 2007; Chang & Chang, 2012; Eagleton, 2015; Wu et al., 2015; Eom & Ashill, 2016). The key aspect of practical anatomy learning involved the identification of various organs and structures, differentiation of tissue types, and their spatial relationships. To accomplish this, students needed to perceive the importance and role of anatomy tools in understanding the human body structure and important relationships relevant to their profession. Overall, students agreed with the items under learning satisfaction. The possible explanation could be related to students' curiosity, interest, and motivation in what they learn, their expectations and the learning needs were likely met by the learning environment. This session was conducted before the formal anatomy course, which might have further contributed to interest and stimulation among students to learn anatomy, thus leading to enhanced learning satisfaction. Earlier studies found that students' positive perceptions of the learning environment were associated with selecting learning approaches (Smith & Mathias, 2010; Vaughan et al., 2014; Ahmed et al., 2018). Improved learning satisfaction promoted a productive learning climate and increased students' enrollment, retention, academic performance, and success rates (Smith & Mathias, 2010; Ahmed et al., 2018).

Hadie's inventory had items that captured students' perceptions of anatomy teachers (Hadie et al., 2021). Although this study's original inventory comprised of an item related to instructor teaching, it did not acquire satisfactory factor loading. Hence, the final inventory did not contain the items related to anatomy teachers. It may be due to the self-learning nature of the activity using plastinated or 3DP models with the help of guided handouts and teamwork. Hence, there was minimal student-instructor dialogue, which encouraged self-discovery and collaborative learning to complete the given tasks. Therefore, this was consistent with the collaborative learning approach in which the learners constructed new knowledge through the interaction and sharing of information among peers (Rutherford, 2014). Sharing of information requires better structuring, logical sequencing, and cognitive processing, which may promote knowledge building, interest and motivation to master anatomical concepts. Items in factor one were related to the learning satisfaction toward 3DP and plastinated tools through self-learning and team activities, rather than teacher-centered instruction.

The second factor, self-efficacy in anatomy, measured the students' judgments on handling situations and completing tasks. Self-efficacy can influence peoples' behavior in selecting the tasks based on their perceived capabilities (Bandura, 1986). Typically, individuals choose assignments for which they possess higher self-efficacy, whereas those with lower self-efficacy tend to avoid them (Schunk, 1991, 2020). In the cadaveric dissection context, learners with higher self-efficacy tend to participate and complete the activity. In contrast, those with lower self-efficacy tend to observe and read rather than participate (Burgoon et al., 2012). This may negatively impact their learning owing to insufficient exposure to the educational activities and materials, consequently affecting their practical and written examinations (Smith & Mathias, 2010; Burgoon et al., 2012). However, non-dissection anatomy courses often utilize a combination of various novel tools (if not all) such as plastinated specimens, plastic, 3DP, and AR, VR virtual models (McLachlan et al., 2004; Mogali et al., 2018; Chytas et al., 2019). The novelty of the models may likely improve students' hands-on experience and influence their self-efficacy related to the ability to learn anatomy concepts, take examinations, and transfer information to clinical situations. However, the pooled mean score of plastinated and 3DP groups for self-efficacy items clustered between the ‘neutral’ and ‘agree’ opinions (mean, 3.59). This might be due to inadequate experience and exposure in anatomy that resulted in perceived lower confidence in the ability to identify the structures and transfer knowledge to different (clinical) situations. However, the item related to recall of anatomy knowledge during the assessment was ranked toward the marginal agree side (mean, 3.90). This might imply that novice learners lack confidence in their abilities and skills due to insufficient exposure and mastery in anatomy knowledge. It was shown in a previous study that self-efficacy of anatomy has increased toward the end of the course compared to the commencement (Burgoon, 2008). Unlike Burgoon's self-efficacy items that suited conventional cadaveric dissections and prosections (Burgoon et al., 2012), the present construct was relevant to non-cadaveric settings, particularly 3DP and plastinated specimens. The Cronbach alpha value for self-efficacy (0.82) was similar to Burgoon's (0.9), reflecting a high degree of internal consistency (Burgoon et al., 2012). Future studies may investigate anatomical self-efficacy between different anatomy tools and assess the predictive validity of this construct through analyzing students' performance in anatomy assessments.

The third-factor humanistic values assessed the students' attitudes such as care, respect, and empathy toward anatomy tools. In most medical schools, anatomy education was mainly supported by cadaveric and bony materials prepared from donated bodies. Previous reports advocated teaching humanistic values in medical education and connected science with humanity (Santibañez et al., 2016; Douglas-Jones, 2017; Ghosh, 2017a). Students deeply respected the silent mentors for their generosity and gift for the advancement of science, handled the specimens carefully, and treated them with dignity during their laboratory sessions (Lai et al., 2019; Souza et al., 2020). The dissection and student-cadaver relationship developed during the laboratory was likely to influence the learners' attitudes toward the humanistic values and future doctor-patient relationships (Rizzolo, 2002; Ghosh, 2017a; Guo et al., 2020). The limited access to cadaveric materials and the curricular reforms created a need in anatomy education to rely on non-cadaveric alternative resources. Although they were effective learning tools, students felt they missed the emotional connection and experience with non-dissection modalities (Fruhstorfer et al., 2011). An earlier preliminary study by the research team showed that the students respected the 3DP models as much as plastinated specimens as they were reconstructed from an actual person's imaging data (Mogali et al., 2018). The survey results showed that students agreed with items related to respect and care toward the anatomical tools. This is consistent with previous studies which involved actual cadaveric specimens (Lai et al., 2019; Souza et al., 2020). This postulate that students appreciated plastinated specimens or 3DP models were prepared from the donated cadavers or their imaging data, who are once the real-life persons. However, pooled mean score related to the empathy item attained a slightly higher than neutral opinion (mean, 3.35). These results were based on two learning sessions (approximately 2.5 hours). Therefore, this instrument may be utilized at regular intervals in anatomy courses to investigate the students' perceived humanistic values in the non-traditional cadaveric education setting. This would assist educators in instilling and enhancing medical humanities amongst medical students before their clinical rotations.

Interestingly, the factor analysis supported the existence of a new factor which was the limitations of learning tools. Although this factor was not the objective of the study, it contained relatively high factor loadings. Through careful examination of the nature of these items (related to fragility and breaking of structures), it was revealed that the items formed a coherent fourth factor. The authors believed that having an additional (and more counter-balanced) subscale added to the overall validity of the measure. This factor had only two items, but they may be reason enough to measure the straightforward psychometric variables. In this case, the fear of breaking structures impacted the perceived interaction with the tools and learning. Effective learning required careful examination, close inspection, and manipulation of the structures in the anatomy tools. This was necessary for the students to recognize how the different organs and structures interact with each other. Previous studies have highlighted the breakage of structures in the plastinated specimens or cross-sections (Fruhstorfer et al., 2011; Riederer, 2014). The risk of breakage or damage of structures in 3DP models may be less than the plastinated specimen as they were primarily produced in rigid materials. However, there is a high risk of losing the details or breakage of structures during the post-processing of the models. In terms of fragility and longevity, future research on 3DP models that utilize different materials (rigid or soft) is needed. In this study, students disagreed with items related to the limitations of learning tools. This indicated that students perceived the models used in the study were non-fragile and sturdy. However, their perception might change when they are exposed to other organs or body regions with delicate structures such as blood vessels of the brain, cranial nerves attachment to the brainstem, and structures that have a long course (blood vessels and nerves of the limbs). These were the commonly broken or damaged structures observed in the plastinated specimens at the authors' institution. In summary, a better understanding of the advantages and disadvantages of anatomy tools would be valuable for refining the learning materials or supplementing them with other established resources.

CONCLUSIONS

This study developed a psychometrically valid and reliable four-factor instrument that measured the students' perceptions of learning satisfaction, self-efficacy, humanistic values, and limitations of learning tools toward plastinated and 3DP anatomy models. This instrument may guide the evaluation of other contemporary tools for anatomy teaching and learning. It may also help educators build new strategies to improve the anatomy delivery, leverage the strengths, and correct the deficiencies of contemporary anatomy tools.