Teaching students how to interpret animal emotions 2: putting research into practice
Sandra L Nicholson
Thursday, March 2, 2023
Part one of this article discussed the importance of teaching veterinary nursing students how to interpret animal emotions, and presented educational strategies for the classroom and clinical placements, informed by Herrington and Oliver's ‘Authentic Learning Framework’. However, translating educational strategies into tangible teaching plans may require considerable effort. This article aims to reduce some of the work involved by sharing an authentic teaching design that was used to support veterinary nursing students, in stage two of the University College Dublin's BSc Veterinary Nursing programme, in learning how to interpret animal emotions. Insights into the student learning process gained from analysing the teaching are also discussed, including student engagement (participation), the use of anthropomorphic descriptors, and the students’ strengths and weaknesses in interpreting animal emotions. Future directions for teaching are also considered. This article is an example of how veterinary nursing students can enhance teaching for future cohorts by participating in educational research.
The first part of this article discussed teaching veterinary nursing students how to interpret the emotions of animals. It also provided educational strategies for the classroom and clinical placements that were aligned with the ‘Authentic Learning Framework’ of Herrington and Oliver (2000). However, transforming teaching approaches into tangible plans may require considerable work. To avoid the unnecessary duplication of effort, the first aim of this second part is to share the design of a specific teaching and assessment plan on interpreting animal emotions. The second aim is to analyse the outcomes of this teaching and discuss any insights gained, as well as their significance.
This article focuses on a stage two module (animal behaviour for veterinary nursing) within the Bachelor of Science in veterinary nursing degree at the University College Dublin. This 4-year honour’s degree programme is recognised by the Veterinary Council of Ireland and the Accreditation Committee for Veterinary Nurse Education. Animal behaviour for veterinary nursing ran in the spring trimester and had a capacity of 45 students. It included education on normal animal behaviour, animal welfare assessment, learning theory and training methods, and animal behaviour problems. The animal emotions teaching and assessment were fully developed for the 2019-20 academic year. The teaching was planned for delivery at the beginning of the module, so that student understanding could then be built on. Emotions were explored in dogs, cats, cattle and sheep, and taught by this author. Equine behaviour was taught by a different lecturer and did not take the same approach, while the behaviour of exotics was not included in the module. The associated learning outcome was to ‘correctly identify/interpret the behaviour, body language and vocalisations of domestic animals’.
Teaching and assessment strategy
Compiling behavioural signals and images
Before designing the teaching, it was necessary to identify the behavioural signals of animal emotions, and source appropriate images and YouTube films to provide an authentic context (Herrington and Oliver, 2000). In previous research, the lecturer had identified and defined valid animal emotions (anxiety/fear, anger/rage, joy/play, contentment, interest), and created a directory (ethogram) of the behavioural signals of feline primary emotions (Nicholson and O’Carroll, 2021). An early version of the directory was selected for teaching and a further search of the literature was performed to identify the behavioural signals of emotion for dogs, cattle and sheep. This search was less systematic than the search for the signals of feline emotions, as the start of term was approaching. Table 1 shows the behavioural signals of emotion that were identified for dogs, cattle and sheep. However, it must be noted that these may need to be updated in future as the research base develops.
Table 1. Behavioural signals of emotion identified for dogs, cattle and sheep
Table 1. Behavioural signals of emotion identified for dogs, cattle and sheep
Once the behavioural signals of emotion had been identified, a search was undertaken to find representative images of each emotion for each species. It was necessary to obtain some images depicting single emotions and others depicting mixed emotions, to be more ‘true to life’ and to collect images of varying difficulty levels. A combination of still images (photographs) and films were sought, as still images were considered to be easier to interpret and films more challenging. As the context of an image or film has been found to act as a clue to an animal’s internal state (Bahlig-Pieren and Turner, 1999), context-free images were selected to encourage students to focus closely on the animals’ behavioural signals. However, the importance of context should be explained to students.
Google Images was used to search for appropriate photographs. Only those with a Creative Commons licence were selected. Most photographs were obtained from https://commons.wikipedia.org and https://pixabay.com. YouTube films were sourced to showcase interactions between individuals and for complex interpretation tasks, and the links were saved. The collected images were screened to ensure that they were clear and representative of the emotion(s) depicted.
Teaching and assessment design
The animal emotions and their behavioural signals were integrated into the PowerPoint slides for each animal behaviour lecture, and illustrated using representative images and/or films. This provided the ‘scaffolding’, discussed in part one (Herrington and Oliver, 2000). The collected images and films were also used to create whole-group authentic learning activities for use in class. Students were asked to identify the behavioural signals present in images and films labelled with the emotion(s) being displayed. They were then asked to identify the emotion(s) present in unlabelled images and films. Questions were directed to the whole class and students answered verbally or by raising their hands. An example of a typical authentic learning task is given in Figure 1.
Figure 1. Sample authentic learning tasks. Students were asked to identify the behavioural signals of emotion in similar images.
Next, three online discussion boards (‘Canine body language’, ‘Feline body language’, and ‘Farm animal body language’) were established in the virtual learning environment. The purpose of the discussion boards was to encourage students to reflect and articulate knowledge, as well as enable feedback/coaching to be provided (Herrington and Oliver, 2000). Students were invited to post images and films of animal emotional states as new threads under the appropriate emotion in each discussion board. They were asked to justify their interpretations by including description of the relevant behavioural signals. Students were required to obtain appropriate consents before using their own images. Use of the discussion boards was strongly promoted as practice for the online quiz, but was not incentivised with grade credit.
Finally, the remaining images and films were used to design an open-book, online quiz (hosted in the virtual learning environment), that was worth 15% of the module’s grade. The quiz contained 15 questions: five on dogs, five on cats, three on cattle and two on sheep. The majority of questions featured still images but four questions featured films. Question formats included the multiselect type (13), the true/false type (1) and the multiple choice question type (1). Each question was worth one mark and negative marking was not used. Multiselect questions comprised five options (anger/rage, anxiety/fear, contentment/relaxation, interest and joy/play) and a proportion of the mark was awarded for each correct selection. The typical wording of a multiselect question was: ‘What emotional/motivational state(s) is this (dog/cat/cow/sheep) signalling?’ An example of this question type can be seen in Figure 2.
Figure 2. Example of a question type used in the online quiz.
To be practical and timely, results and feedback were set to be automatically available to students immediately after quiz completion. The quiz was open for the majority of the second trimester and students could choose when to complete it. However, some safeguards to reduce the risk of student collaboration were built in. Only one attempt was allowed and students were required to complete the quiz in one sitting of 75 minutes. The right-click function was disabled for the duration of the quiz. In addition, the order of question presentation was randomised and multiselect/multiple choice question options were randomised. In future, increasing the number of available questions, so that each student receives a different set, could further reduce the possibility of collaboration.
Evaluating the strategy and outcomes
The University College Dublin’s Human Research Ethics Committee approved the reporting of student data relating to this authentic teaching strategy (ethics reference number LS-E-20-22-Nicholson). A total of 13 (34%) of the 38 students enrolled in VNUR20350 (animal behaviour for veterinary nursing) in 2019-20 provided electronic informed consent for use of their data.
Engagement (participation) during lectures and with the in-class authentic learning activities was subjectively assessed. Online discussion board engagement was assessed using data from the virtual learning environment on the number of times each participant posted and the number of times they read the posts of others. The discussion board posts of participants were also qualitatively reviewed; correct and incorrect observations/interpretations, the rationales supplied by the student and any omissions were noted. The participants’ online quiz results were anonymously extracted from the virtual learning environment and analysed. Data were stored in password-protected Microsoft Excel (Version 16.63.1) files on an encrypted computer and will be held for 6 years. IBM SPSS Statistics (Version 27) was used to calculate descriptive statistics. The European General Data Protection Regulations (European Council and the Council of the European Union, 2016) were strictly adhered to during the handling and processing of student data.
Findings and discussion
Student engagement (participation)
Engagement with the online discussion boards was much higher than with the in-class learning activities. Most students appeared reluctant to volunteer information in the large group setting and comments were usually made by the same individuals. In future, participation may be promoted by using anonymous polling (Draper et al, 2002; Stowell and Nelson, 2007) or groupwork (Kinsella et al, 2017) instead of questions and verbal responses or hand raising.
In contrast to the in-class tasks, the majority of participants (11/13) contributed to the discussion boards (by posting threads, reading threads or both) between January and March 2020. Study participants posted a median of two times (median=2; range 0-8) and read posts a median of 15 times (median=15; range 0-118).
The students may have found the discussion boards less intimidating than the in-class activities (Douglas et al, 2020). They may also have felt more confident in their responses, having had time to reflect on their learning (Douglas et al, 2020). In addition, students who did not attend lectures could still contribute to the discussion boards. This may explain the higher level of engagement. However, engagement could be further enhanced by offering grade credit for participation. Student engagement was also influenced by the species involved in the activity. The first discussion board (canine body language) had the highest level of engagement (nine students posted threads and 11 students read posts), the second (feline body language) had moderate engagement (six students posted threads and seven students read posts) and the final one (farm animal body language) had the lowest level of engagement (two students posted threads and four students read posts). Students posted threads in the canine and feline discussion boards while the farm animal board was running. Therefore, the lower level of engagement with the farm animal board cannot be explained by fatigue. It may have been caused by difficulties in sourcing images or a lack of confidence in interpreting farm animal behaviour (owing to a lack of experience with these species). A survey found that many Irish veterinary nursing students do not have prior cattle handling experience (Dunne et al, 2018). In future, the lecturer may need to directly encourage students to post more threads for cats and farm animals. In addition, placement supervisors could assist students by encouraging them to observe and interpret the emotions of feline or farm animal patients whenever possible. If their practice does not deal with these species, students could be tasked with sourcing online images and films of them for discussion. All students completed the online quiz and did so between the final week of March and the final week of April 2020.
Insights into the student learning process
The authentic learning tasks provided useful insights into the student learning process and enabled specific, timely and actionable feedback to be provided to them. As a result, performance in the online quiz was excellent. All student participants achieved between 71% and 100% (median= 88.62%, standard deviation =9.07). However, the ability to transfer learning from the classroom or online environment to real situations should be assessed in future.
The first insight gained from the authentic learning tasks related to terminology. In the online discussion boards, some students used lay or anthropomorphic terminology to describe animal behaviour signals, although this was not a barrier to success. For example, in the canine discussion board, ‘cowering’ was used to describe a lowered head and body position and ‘sad anxious eyes’ was used to describe tension in the brow and/or dilated pupils. In the farm animal discussion board, ‘calf zoomies’ was used as the title for a thread on play behaviour in cattle. Interestingly, this did not occur with cats. Indeed, research within the lay population (Bahlig-Pieren and Turner, 1999) found that cat images are less likely to be described anthropomorphically than dog images.
A second insight was that in nearly all threads, students correctly recognised the predominant emotion in an image but failed to notice a more subtle coexisting emotion (such as mild fear presenting with overt anger). Students could identify the principal emotion in an image irrespective of the species depicted and the emotion involved. This is surprising, as non-veterinary professionals often failed to correctly identify feline emotional state (Dawson et al, 2019). In addition, even people without canine experience were found to correctly identify happiness (contentment) in dogs, but canine professionals were most successful in identifying fear (Wan et al, 2012). However, the students had received instruction on animal emotions that lay individuals would not have received. In many cases, the students correctly identified the predominant emotion in an image without listing all of the behavioural signals displayed. They may have forgotten to report them or they may have observed them subconsciously. Alternatively, they may have studied some but not all parts of the animals body, missing some behavioural signals. This may be why they did not recognise mixed emotions. A study found that dog owners used fewer body parts to interpret canine emotions than dog professionals (Wan et al, 2012). Therefore, both lecturers and placement supervisors should encourage students to view all parts of the animal’s body systematically (the face, eyes, brow, ears, neck, body, tail and limbs) and consider all their behavioural signals together before evaluating emotional state. This is also important because some behavioural signals may feature in more than one emotional state (Nicholson and O’Carroll, 2021).
Translating educational strategies into specific teaching plans requires much work. To avoid the duplication of effort, this article reported on the design and outcomes of authentic classroom-based teaching and assessment on interpreting animal emotions. Learning was ‘scaffolded’ by lectures, reinforced through the use of classroom-based authentic learning tasks (involving the use of images and YouTube films of animals) and online discussion boards, and assessed in an online quiz. As a result of student involvement in educational research, insights were gained into the learning process that can be used to enhance teaching. Engagement with the in-class learning activity was low but could be enhanced by using anonymous polling or groupwork. As contributions to the farm animal and feline discussion boards were suboptimal, more emphasis should be placed on observing/analysing the behavioural signals of these species in future. In addition, all students had difficulty in identifying mixed emotions. Therefore, students should be advised to observe the whole animal before making judgements about the emotional state.
- Authentic learning principles can be enacted to successfully teach students how to interpret animal emotions.
- By participating in research, students can help lecturers to enhance veterinary nurse education.
- This research found that students may need particular encouragement and support to engage with activities on the emotional state of farm animals or cats.
- This research also found that students can identify predominant animal emotions while missing subtle coexisting ones. Therefore, students should be advised to observe the whole animal before making evaluations.
- Students occasionally used anthropomorphic language to describe canine or farm animal behavioural signals, but this was not detrimental.