Animals in Dangerous Postures Enhance Learning , but Decrease Willingness to Protect Animals

Animals are the most prevalent subjects for photographs in science textbooks. Many of them are potentially dangerous to humans, and visual exposure to potential threat can influence learning outcomes as well as emotional attachment to these animals. We experimentally investigated the influence of animal posture (aggressive-looking vs. neutral-looking) on 10 – 13-year-old pupils’ information retention and willingness to support the protection of these animals. We found that information placed below aggressive-looking animals was retained significantly better than information placed below neutral-looking animals. Survival-relevant information was retained better than survival-irrelevant information. Willingness to protect aggressive-looking animals was lower than willingness to protect neutral-looking animals. This suggests that aggressivelooking pictures receive stronger attention, but do not distract pupils from learning. Aggressive animals may have a negative influence on animal conservation efforts.

Animals are the most prevalent subjects of the photographs in science textbooks (Link-Pérez, Dollo, Weber, & Schussler, 2010).This can be explained by the natural affinity of humans to animals.Historically, animals have been important sources for the human diet, but also pose a physical danger (Treves & Palmqvist, 2007).This is why animals receive increased attention by humans (the animate monitoring hypothesis, see New, Cosmides, & Tooby, 2007).Indeed, animals are visually preferred by two-day-old babies more than non-biological targets (Simion et al. 2008) and are detected more rapidly than plants (Balas & Momsen, 2014).Predatory animals are detected faster than flowers or mushrooms (LoBue & DeLoache, 2008;Öhman, Flykt, & Esteves, 2001) and tracked by the eyes for a longer time than non-predatory animals (Penkunas & Coss, 2013;Yorzinski, Penkunas, Platt, & Coss, 2014).Observations by children and adults revealed that physical interaction with live animals was more frequent than with toys (LoBue, Bloom Pickard, Sherman, Axford, & DeLoache, 2013).This evidence suggests that humans have an innate preference for animals and predatory animals in particular receive stronger attention.
Pictures in textbooks have not only a cognitive, but also an affective influence on learning (Carney & Levin, 2002;Lenzner, Schnotz, & Müller, 2013;Levie & Lentz, 1982).By means of looking at textbook pictures, increased pupil interest may enhance willingness to learn, and activate cognitive strategies that lead to more effective learning (Schraw & Lehman, 2001;Vollmeyer & Rheinberg, 2006).Most of the research was carried out on "positive" aspects, e.g. on enhancing mood (Lenzner et al., 2013).Animals are special, however, as some of them may pose a danger to humans.Perceived danger increases fear and fear memories are deeply stored in the brain (Johansen, Cain, Ostroff, & LeDoux, 2011;McGaugh, 2000).Indeed, Chapman, Johannes, Poppenk, Moscovitch and Anderson (2013) revealed that items scoring high in fear are remembered better than neutral items.Štefaniková and Prokop (2015) found that information concerning dangerous animals was retained better than information about non-dangerous animals in all probability because retaining survival-relevant information is ultimately more advantageous than retaining survival-irrelevant information (Nairne, 2010;Nairne, Thompson, & Pandeirada, 2007).
It has been speculated that danger from animals is elicited by certain harmful shapes, such as teeth, claws or spikes (Štefaniková & Prokop, 2015;Souchet & Aubret, 2016).Presentation of animals in science textbooks may contain these aggressive cues, but it is not clear whether these cues influence learning.On the one hand, fear may enhance learning (e.g., Chapman et al. (2013), but on the other hand, if aggressive pictures attract more attention, it can be argued that they also distract the pupil from learning (Sanchez & Wiley, 2006).Furthermore, visual exposure to colour photographs of animals positively influenced the willingness to protect them (Štefaniková & Prokop, 2013).Willingness to protect animals is, however, influenced by their perceived aesthetic value (Gunnthorsdottir, 2001;Knight, 2008;Prokop & Fančovičová, 2013a;Soga, Gaston, Yamaura, Kurisu, & Hanaki, 2016) and perceived fear (Prokop & Fančovičová, 2010;Johansson, Sjöström, Karlsson, & Brännlund, 2012).It is again not clear how dangerous cues may influence willingness to protect animals.
In the present study, we presented various colourful pictures of animals to pupils and investigated whether aggressive-looking animals influence 1) information retention about them and 2) willingness to support animal protection compared with the control, neutral-looking animals.

Participants
A convenience sample of 150 pupils was used in this research, but the final sample comprised 91 pupils (48 females) who completed both test and retest.These pupils were fifth and sixth graders from three higher primary schools and seven classes.The age of the pupils was 10 -13 years (M = 11.23,SE = 0.08, n = 91).Participants were unaware of our hypotheses.

State of the literature
• Animals are frequently the subjects of photographs in science textbooks.
• Fear elicited by animals may influence both cognitive and emotional processes.
• Fear may distract pupils from learning.

Contribution of this paper to the literature
• Some information about aggressive-looking animals was retained better than information about neutrallooking animals.
• Survival-relevant information was retained better than survival-irrelevant information.
• Willingness to protect aggressive-looking animals was lower than willingness to protect neutral-looking animals.

Species Selection
Animals which pose a potential danger to humans were used in this research: dingo (Canis lupus dingo), common lancehead (Bothrops atrox), fossa (Cryptoprocta ferox), Indian red scorpion (Hottentotta tamulus), spotted hyena (Crocuta crocuta), snow leopard (Panthera uncia), common hippopotamus (Hippopotamus amphibius), collared peccary (Tayassu tajacu), wolverine (Gulo gulo) and hamadryas baboon (Papio hamadryas).Pictures of all these species were found in both aggressive (or at least with visible weapons dangerous to humans such as teeth) and a neutral posture with Google (Figure 1a, & 1b).The pictures had a similar size, contrast, brightness and all had removed the background.We chose species that do not occur in Slovakia to avoid previous experience and familiarity of the participants with the presented animals.

Procedure
Participants were randomly divided into two groups (A and B).Both groups were shown species in both aggressive and neutral postures in both the test and retest.If, for example, a participant in Group A was shown the fossa in a neutral posture, participants in Group B were shown the fossa in an aggressive posture.Both test and retest were administered online and participants filled out the questionnaire during biology lessons.
In the test (first trial), participants were asked basic demographic information such as grade, gender, age and name.There was a need to do this research non-anonymously, because participants of this age frequently lose their ID numbers and, moreover, if the ID number is required for the next trial, they could have the suspicion that they were going to be examined again.The participant's name was the only way to pair data from the test and retest.The participants were then shown a series of 10 pictures.Each picture contained one animal (a frog, reptile, bird and mammal) which was presented individually for 1 min.Over this time, the participants rated their perceived fear (How dangerous would you consider this animal?) on a 10-point scale (1 = not at all dangerous, 10 = extremely dangerous) and the willingness to protect the animal (Do you think that this species should be protected by laws?) (yes or no).The reliability was acceptable (Cronbach's alpha = 0.64 and 0.74, respectively).Ratings of perceived danger and willingness to protect animals were applied only once (in the test, not in the retest).Some basic information regarding the name of the animal, its occurrence, food habits and danger to humans then appeared on the slide below the picture.This information was standardized to a similar length of text and there were only four types of information mentioned above.Participants were instructed to read this information.After the ratings of all 10 animals was completed, a surprise memory test appeared.Each of the 10 slides contained only a picture with an animal, but instead of the text with information, four open-ended questions regarding the animal name, occurrence, food and dangerousness to humans appeared.To examine which items had passed into long-term memory, the retest (second trial) was administered one week later and contained the same four questions regarding 10 animal species.Similar procedure can be found elsewhere (e.g., Barrett & Broesch, 2012;Prokop & Fančovičová, 2014).The participants were then debriefed and dismissed.
The participant's responses were coded as correct (2 points), partly correct (some information missing, 1 point) and incorrect/do not know (0 points).Example: Where does the animal [collared peccary] occur?Both North and South America (correct); South America or North America (partly correct); Asia, Africa, or do not know (incorrect).Average scores were calculated for animals presented in aggressive and neutral postures separately.

Statistical Analyses
Linear Mixed Model (MIXED) with repeated measures was used to examine the obtained data.The participant's ID and the group of participants (A or B) were treated as random factors.These variables showed no influence on the dependent variables (Wald Z = -1.38 and 1.43, p = .17and .15,respectively).Test -retest measures were treated as repeated measures.The participant's sex, the type of presented animal (aggressive vs. neutral posture) and the type of question (naming, occurrence, food and danger) were defined as categorical predictors and age was a covariate.Since females may perceive animals somewhat differently than males (see Herzog 2007 for discussion), we included gender into statistical analyses.Scores from four questions were dependent variables.Test memory scores were significantly higher than retest memory scores (Wald Z = 26.04,p < .001).No significant interactions between test-retest scores and other variables appeared, thus this variable was not further examined.Statistical analyses were performed with IBM SPSS ver.22.

The Influence of Animal Appearance on Information Retention
The results of the Linear Mixed Model on memory scores (mean scores of four knowledge questions) are shown in Table 1.Aggressive-looking animals received higher memory scores compared with neutral-looking animals (analysis of contrasts, t = 3.36, p = .001).
The type of question also influenced memory scores; the highest scores were received for items regarding danger and food, while the lowest scores were received for naming of animals and occurrence.These patterns were very consistent in both test and re-test (Figure 2a, & 2b).
The interaction term Type of animal × Type of question suggests that pupils received higher memory scores from questions regarding animal dangerousness and food particularly when an animal looked dangerous (Figure 2a, & 2b).The remaining two questions (naming and occurrence) received similar memory scores irrespective of the type of animal.
The interaction term Gender × Type of question suggests that females retained more information regarding animal danger than males, but males scored higher in questions regarding animal occurrence.

DISCUSSION
This study investigated the influences of aggressive animal postures on pupil's learning outcomes and their willingness to protect these animals.We found that aggressive-looking animals significantly influenced both these variables.Pupils retained more information about aggressive-looking animals than about neutral-looking animals, but were less likely to support protection of aggressive-looking animals compared with neutral-looking animals.

Information Retention Concerning Aggressive-Looking Animals
The same animal species produced different information retention scores in pupils, because more information about aggressive-looking individuals was retained than about neutral-looking individuals.These results are in agreement with previous research (Štefaniková & Prokop, 2013(Štefaniková & Prokop, , 2015)).Certain neural processes which are activated in fear-learning (Chapman et al., 2013;Johansen et al., 2011;McGaugh, 2000) would at least partly influence these results.Both pupils and adults, for example, detected the photos of snakes displaying a striking posture faster than the photos of resting snakes (Masataka, Hayakawa, & Kawai, 2011).It can be suggested that aggressive-looking pictures could attract more attention (Penkunas & Coss, 2013;Yorzinski et al., 2014) and stimulated situational interest enhanced information retention (Harp & Mayer, 1997;Hidi & Baird, 1988;Schraw & Lehman, 2001).This possibility can be further investigated by combining an eye-tracking technique with selfreports regarding interest in particular animals.Importantly, pictures with aggressive-looking animals do not appear to distract pupils from learning (Sanchez & Wiley, 2006) as higher information retention scores were obtained compared with neutral-looking animals.
The type of question significantly influenced information retention.Questions about dangerousness and food, in particular, received higher scores than questions regarding the naming of animals and their diet.These results are, again, compatible with previous research on both animals (Barrett & Broesch, 2012;Štefaniková & Prokop, 2013Štefaniková & Prokop, , 2015) ) and plants (Prokop & Fančovičová, 2014;Prokop et al., 2016), suggesting that survival-relevant information is retained better than survival-irrelevant information (e.g., Nairne, 2010;Nairne & Pandeirada, 2010;Nairne, Pandeirada, Gregory, & Van Arsdall, 2009).This is, however, true particularly for dangerousness, not for an animal's diet.Although previous research also found that scores for diet were higher than scores for occurrence and naming (Barrett & Broesch, 2012), it is not fully clear why the score for diet was similarly high as scores for danger.Finally, it is also not fully clear why the score for diet was better retained compared with naming and occurrence, but it suggests that not all information is better retained under an aggressive-looking condition.
Gender differences in information retention were weak, similarly as in earlier research (Štefaniková & Prokop, 2013(Štefaniková & Prokop, , 2015)).Females, however, manifested better information retention regarding animal danger than males.It is possible that females overperceive the danger posed by animals, because of their higher vulnerability of being killed by a dangerous predator (Prokop & Fančovičová, 2013b;Treves & Naughton-Treves, 1999).Males, in contrast, scored higher in questions regarding animal occurrence than females.This can be explained by the higher interest on the part of males in less popular animals such as snails, bats, and rats (Bjerke & Østdahl, 2004) and exotic wild animals (Lindemann-Matthies, 2005).

Willingness to Support Animal Protection
Lower willingness to support animal protection correlates (among other things) with their perceived aesthetic value (Ceríaco, 2012;Gunnthorsdottir, 2001;Knight, 2008;Prokop & Fančovičová, 2013a).Although this variable was not specifically examined in this study, aggressive-looking animals are in all probability perceived as less appealing than the same, but neutral-looking animals.This has important implications for environmental conservation and conservation programmes; aggressive-looking animals may receive lower public support ultimately because their presence may activate fear-motivated behaviour after exposure to danger (Öhman et al., 2001).We further recommend critical evaluation of natural history films about predators and their impact on an individual's motivation to protect them.It is possible that less pleasant, predatory scenes, can have a negative influence on environmental conservation efforts.
In conclusion, differences, not only between-species but also within-species, influence the perception of animals by pupils.We demonstrated that pictures with aggressive-looking animals are associated with better information retention, but the same pictures have a negative influence on willingness to support animal protection.This information is potentially important for teachers and developers of learning material who may alter the appearance of animals in textbook pictures.It is not actually clear as to whether our findings can be applied exclusively on pupils who are more vulnerable to predation threat, and could therefore be more sensitive to aggression cues than adults.Additional, cross-age research in this field is necessary before a final conclusion can be reached.

Figure 1 .
An example of the spotted hyena as an animal presented in a) aggressive and b) neutral posture

Figure 2 .
Differences in memory scores based on four questions in test (a) and retest (b).*** p < .001,NS = not statistically significant P. Prokop & J. Fančovičová / Aggressive Animals, Conservation and Learning

Table 1 .
Results of Linear Mixed Model (MIXED) with repeated measures on an average memory score