You are on page 1of 5

Applied Animal Behaviour Science 169 (2015) 38–42

Contents lists available at ScienceDirect

Applied Animal Behaviour Science


journal homepage: www.elsevier.com/locate/applanim

Development of a standard test to assess negative reinforcement


learning in horses
Line Peerstrup Ahrendt a,∗ , Rodrigo Labouriau b , Jens Malmkvist a , Christine J. Nicol c ,
Janne Winther Christensen a
a
Dept. Animal Science, Aarhus University, Tjele, Denmark
b
Dept. Mathematics, Aarhus University, Aarhus, Denmark
c
School of Clinical Veterinary Science, University of Bristol, Langford, UK

a r t i c l e i n f o a b s t r a c t

Article history: Most horses are trained by negative reinforcement. Currently, however, no standardised test for evalu-
Received 14 March 2015 ating horses’ negative reinforcement learning ability is available. The aim of this study was to develop an
Received in revised form 15 May 2015 objective test to investigate negative reinforcement learning in horses. Twenty-four Icelandic horses (3
Accepted 17 May 2015
years old) were included in this study. The horses were tested in a pressure-release task on three separate
Available online 28 May 2015
days with 10, 7 and 5 trials on each side, respectively. Each trial consisted of pressure being applied on the
hindquarter with an algometer. The force of the pressure was increased until the horse moved laterally
Keywords:
away from the point of pressure. There was a significant decrease in required force over trials on the first
Algometry
Horse behaviour
test day (P < 0.001), but not the second and third day. The intercepts on days 2 and 3 differed significantly
Learning performance from day 1 (P < 0.001), but not each other. Significantly stronger force was required on the right side com-
Operant conditioning pared to the left (P < 0.001), but there was no difference between first and second side tested (P = 0.56).
Pressure-release Individual performance was evaluated by median-force and the change in force over trials on the first
Horse training test day. These two measures may explain different characteristics of negative reinforcement learning. In
conclusion, this study presents a novel, standardised test for evaluating negative reinforcement learning
ability in horses.
© 2015 Elsevier B.V. All rights reserved.

1. Introduction negatively reinforced test. Consequently, investigations of horses’


ability to learn through negative reinforcement and factors influ-
The value of domestic horses is dependent on their ability to encing this ability require a test based on negative reinforcement.
learn the various tasks set by humans. The horse sector would Negative reinforcement is an instrumental learning mechanism,
therefore benefit from the development of standardised methods where the instrumental response turns off an aversive stimulus
to measure learning in horses. Equine learning ability has been (Domjan, 2010). By turning off the aversive stimulus the animal is
investigated mainly in tests based on positive reinforcement (Nicol, rewarded for the instrumental response and will be more likely to
2002), including visual discrimination tasks (Gabor and Gerken, perform the same response when it encounters the same aversive
2010), target touching and following (Christensen et al., 2012; stimulus in the future.
Valenchon et al., 2013) and detour learning (Baragli et al., 2011). In horse training the aversive stimulus is often some kind of
In contrast, most horse training is based on negative reinforcement pressure to the body of the horse, such as leg pressure or rein ten-
(McGreevy and McLean, 2010). Studies have shown that horses’ sion. The force of the pressure is increased until the correct response
general learning ability cannot be evaluated based on perfor- is obtained and rewarded by release of the pressure (McGreevy
mance in one test alone (Visser et al., 2003; Wolff and Hausberger, and McLean, 2010). Currently, no standardised test for evaluating
1996). Rather, learning ability appears to be specific to the learning equine learning ability through negative reinforcement is available,
mechanism targeted, thus, performance in a test based on posi- and the present study aimed to develop an objective test in this
tive reinforcement does not necessarily predict performance in a area. Learning was evaluated in a pressure-release task, where the
pressure was applied in a standardised manner by an algometer,
which also measured the force of the required pressure.
We hypothesised that the force required for the horse to respond
∗ Corresponding author. Tel.: +45 8715 4279. correctly would decrease with repetition within as well as between
E-mail address: peerstrup@gmail.com (L.P. Ahrendt). days.

http://dx.doi.org/10.1016/j.applanim.2015.05.005
0168-1591/© 2015 Elsevier B.V. All rights reserved.
L.P. Ahrendt et al. / Applied Animal Behaviour Science 169 (2015) 38–42 39

2. Materials and methods

The experiment was conducted during November 2012 and


conformed to the Guidelines for ethical treatment of animals
in applied animal behaviour and welfare research [http://www.
applied-ethology.org/ethical guidelines.html].

2.1. Animals

Twenty-four Icelandic horses (3 years old) participated in this


study (12 mares and 12 geldings). The horses came from 14 differ-
ent owners, but had been at the research facility for 3 months prior
to this experiment. All horses had only received routine manage-
ment before arrival at the research facility. At the research facility
the horses were kept at pasture divided by sex into two groups.
They had ad libitum access to haylage, water, grass from the pas-
ture, and a shed for shelter. Upon arrival all horses were trained in
basic handling, such as wearing halter, being caught on pasture and
being led (Marsbøll and Christensen, 2015). The horses had partic-
ipated in another behavioural experiment, where half the horses
had been trained in five handling exercises (training group), e.g.
start and stop, backing, walking sideways and walking in circles (see
Marsbøll and Christensen (2015) for details). During these exer-
cises, the horses were mainly handled from the front, i.e. they were
not more accustomed to being handled from one side. The exer-
cises did not include hind leg yielding as the task applied in the
present experiment. The other half of the horses (control group)
were not trained in these exercises, but spent the same amount of
time in the experimental arena as the trained horses (Marsbøll and
Christensen, 2015). All horses were sound and healthy.

2.2. The algometer

Pressure was applied and its force measured by a ProD algome-


ter (TopCat Metrology Ltd., UK (http://www.topcatmetrology.
com)) with a pressure-surface-diameter of 16 mm. The force was
measured in Newton (N). The algometer was manually operated
with a red and a green light indicating if the force was increased
according to the chosen rate of 4 N/s. The actuator of the algometer
was calibrated up to 30 N.

2.3. Methods

The handler held the horse in halter and lead rope, while the
experimenter applied the pressure on the horse’s hindquarter. The
handler and experimenter were always the same two persons. Prior
to testing, the horses were habituated to the surroundings and
stood calmly when the experimenter approached and stroked the
hindquarter of the horse. Fig. 1. Point of pressure. The yellow dot indicates the spot on the horse’s hindquarter
The horses were tested on three separate days with four and five where the pressure was applied (For interpretation of the color information in this
days between. Each test day consisted of 10 (Day 1), 7 (Day 2) and figure legend, the reader is referred to the web version of the article.).
5 (Day 3) trials on each side of the horse. The number of trials was
reduced on the second and third day because these young horses trial (Fig. 1). In each trial, the horse was placed so that all four legs
appeared to become impatient and less attentive with increasing were weight bearing. The experimenter then placed the algome-
number of trials within days, i.e. they were increasingly difficult to ter on the marked spot and increased the force according to the
get to stand still between trials, were nibbling the handler more set rate. When the horse responded correctly the algometer was
and pawed more. In each trial pressure was applied until the horse immediately removed to release the pressure and thereby reward
responded correctly by moving at least one hind leg laterally away the horse. When the actuator limit of 30 N was reached without
from the point of pressure, i.e. if pressure was applied on the left correct response, the experimenter used the hand, not operating
side the horse should move to the right. The side tested first was the algometer, to apply the extra force necessary to obtain the cor-
randomly chosen for each individual, but balanced according to rect response. This was done to prevent a connection between not
sex. Sides of an individual horse were tested in the same order on responding and the reward of removing the pressure which would
all three test days. The first side was always finished with all trials have disrupted the learning process. In trials where the actuator
before continuing with the trials on the second side. A mark was limit was exceeded a “max” was recorded for the force. Regard-
drawn on the approximate middle of both sides of the hindquarters less of the side of the horse the experimenter faced the point of
to ensure that the pressure was applied on the same spot in each pressure and used the right hand to operate the algometer. Between
40 L.P. Ahrendt et al. / Applied Animal Behaviour Science 169 (2015) 38–42

each trial, the handler repositioned the horse for the next trial. The Finally, we investigated if the proposed performance values
inter-trial interval was approximately 15–60 s, depending on how ranked the horses similarly. All correlations were calculated by
long it took the handler to get the horse in the right position for the Spearman’s rank correlation rho.
next trial. All analyses were done in R version 3.1.2 (R Core Team, 2014)
with significance evaluated at P = 0.05 and tendencies between
2.4. Statistics 0.05 < P < 0.1.

2.4.1. General learning


The necessary force to obtain the correct response exceeded 3. Results
the actuator limit (30 N) in 28% of the trials. Since the probability
of having a censored observation depends on the same explana- The applied force ranged from 0.04 N to the actuator limit of
tory variables used to model the force, the censoring mechanism is 30 N (median [25; 75% quartile]: 17.83 N [8.22; 30]). There was one
informative. Informative censoring rules out the use of many classic instance of what could have been interpreted as a threat to kick,
statistical models for censored observations (e.g. models for sur- but generally the horses tried to avoid the pressure by moving side-
vival analysis or event-history analysis (Kalbfleisch and Prentice, ways or forward or backwards, the two latter being prohibited by
2011)). Additionally, the distribution of the force was unknown, the handler. Some horses moved only one leg in response to the
i.e. classic distributions such as the normal, log-normal or expo- pressure, whereas others took several steps sideways; both types of
nential did not adjust to the distribution of the force. Therefore, response were rewarded. The horses had between 0 and 86% trials
the force was modelled using a semiparametric accelerated fail- exceeding the actuator limit of 30 N during the entire test (median
ure time model (AFTM; function aftgee from the R-package aftgee [25; 75% quartile]: 24% [7; 40]). Two horses had no trials exceeding
(Chiou et al., 2014b)) inferred via least-squares and generalised the limit in the entire test, whereas four horses had all trials on one
estimating equations (Chiou et al., 2014a; Jin et al., 2006). This side exceeding on the first day. In general, the frequency of exceed-
model accounted for the right censoring and the repeated meas- ing trials decreased over the test days, with 35% the first day and
ures on the same horse. As explanatory variables, test day (1–3), 22% on the second and third day.
side (right/left), first or second side tested, sex of the horses
(mare/gelding) and treatment group in a previous experiment 3.1. Learning through negative reinforcement
(training/control) were included as discrete variables and trial as
continuous variable. The logarithm of the force depended on a linear combina-
tion of an interaction between day and trial and additively of
2.4.2. Individual performance side and treatment in the prior experiment. There was no differ-
Individual performance was evaluated by frequency of max- ence between the first and the second side tested (AFTM: z = 0.59,
trials, i.e. trials where the required force exceeded the actuator P = 0.56) or between mares and geldings (AFTM: z = 1.18, P = 0.24).
limit of 30 N (#max), median-force (Median-force), and the change The interaction between trial and day implies that the expo-
in force over trials on Day 1 (Slope1). The median was chosen nential decay in the force over trials was not the same on all
over average because it is independent of the censoring, when less three test days. There was a significant exponential decay in force
than 50% values are censored. If more than 50% were censored the along the first day (AFTM: z = −3.70, P < 0.001, Fig. 2), but not along
median was assigned 30 N. #max and Median-force were calculated the second (AFTM: z = 1.12, P = 0.26, Fig. 2) and third day (AFTM:
for each horse, day and side to investigate correlation between days z = 1.23, P = 0.22, Fig. 2). The intercept for the exponential regres-
and sides. As the slope on day 1 approximated linearity, Slope1 was sion curve was significantly higher on Day 1 compared to Day 2
calculated by a linear regression, where trials exceeding the actua- (AFTM: z = −6.02, P < 0.001, Fig. 2) and 3 (AFTM: z = −5.39, P < 0.001,
tor limit were set equal to 30 N. A value for each side was calculated Fig. 2), whereas the intercept on Day 2 and 3 was not different from
for Slope1. each other (AFTM: z = −1.18, P = 0.24, Fig. 2). The right side of the

Day 1 Day 2 Day 3


35
30
25
Force (N)
15 10
5
0 20

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 1 2 3 4 5

trial trial trial

Fig. 2. Development in force over trials within test days. Dots indicate estimated means for force per trial for the left side on non-trained control horses. The right side
required 1.4 times stronger force (P < 0.001) and the trained horses required 0.6 times lighter force (P = 0.09). The lines show the regression curve for the development over
trials. Only the slope on Day 1 is significantly different from 0 (P < 0.001). The intercepts on Day 2 and Day 3 is significantly different from Day 1 (P < 0.001), but not each other
(P = 0.24).
L.P. Ahrendt et al. / Applied Animal Behaviour Science 169 (2015) 38–42 41

Table 1 horses’ learning ability in a negative reinforcement situation, which


Median and quartiles of measures of individual learning performance in a negative
is relevant for training of horses.
reinforcement based task.
The lack of further decrease in force on day 2 and day 3 may
Performance value Median [25; 75% quartile] indicate that the horses had reached an asymptotic level. However
Median-force on Day 1 (N) 21.05 [15.62; 27.43] it is, of course, unknown if further learning might occur on subse-
Frequency of max trials 2.5 [1; 5.5] quent days. We chose to reduce the number of trials on the second
(#max) and further on the third day because we noted some impatience in
the horses, which could have reduced their attention towards the
Right Left
task.
Change in force over trials −0.45 [−1.21; 0.03] −0.41 [−1.46; 0.22] Horse trainers and riders may argue that at least some horses
on Day 1 (Slope1)
can be trained to respond to lighter pressure, than what is pre-
sented in this study (1 N is equivalent to 0.102 kg). It is very likely
that prolonged training can lead to horses responding to very light
horses required 1.4 times [95% CI: 1.3; 1.5] stronger force than the
tactile cues accommodated by other signals, i.e. a visual signal that
left side (AFTM: z = 3.57, P < 0.001) and the horses trained in the
precedes the pressure, through classical conditioning. From studies
previous experiment tended to require 0.6 times [95% CI: 0.4; 0.8]
investigating avoidance learning it has been shown that horses can
lighter force than the untrained control horses (AFTM: z = −1.72,
learn to connect a conditioned stimulus and an aversive, uncon-
P = 0.09).
ditioned reinforcer, like air puffs (Visser et al., 2003) and electric
shocks (McCall et al., 1993). Although, the ability to form these
3.2. Individual performance connections may be important when considering horses’ trainabil-
ity, this study aimed to investigate negative reinforcement learning
Both median-force and frequency of max trials correlated sig- and, thus, we tried to prevent the horses from responding to signals
nificantly between sides (Spearman rank correlation rho: Median- other than the pressure from the algometer.
force: rho = 0.63, P < 0.001 and #max: rho = 0.60, P < 0.001) and days Contrary to our expectations, we found a significant effect of the
(Spearman rank correlation rho: Median-force: rho = 0.56–0.79, side of the horse; the right side always requiring more force than the
P < 0.01 and #max: rho = 0.64–0.77, P < 0.001) and we, thus, con- left side. The literature does not offer a general population tendency
tinued with one value for day 1 averaged over sides (Table 1). Five towards right or left handedness/sidedness in horses (McGreevy
horses had more than 50% censored trials and were, thus, assigned and Rogers, 2005; Murphy et al., 2005), which could explain our
30 N in the Median-force. Slope1 did not correlate between sides results. McGreevy and Rogers (2005) proposed that motor laterality
(Spearman rank correlation rho: rho = 0.32, P = 0.13; Table 2). The in horses is affected by age and/or training; older horses gener-
left and right side slopes also did not correlate on the second and ally showing stronger laterality tendencies than younger horses.
third day (Spearman rank correlation rho: P = 0.49 and P = 0.58, Icelandic horses are typically led from the front and the horses in
respectively). present study were minimally handled at arrival at the research
centre, which would suggest them to be more ambidextrous. An
3.3. Correlations between performance values alternative explanation is the experimental setup, where the same
right-handed experimenter performed all the trials. Because the
We found a high degree of correlation between Median-force experimenter used her right hand on both sides, this may have
and #max (Spearman rank correlation rho: P < 0.001, Table 2). caused a slight difference in position on each side, which may have
Slope1 on the left side correlated significantly with Median-force caused the side difference. This should be investigated further and
(P = 0.02, Table 2) and tended to correlate to #max (Spearman rank needs to be considered to obtain valid results.
correlation rho: P = 0.097, Table 2), whereas the slopes on the right It is interesting that there appeared to be no transfer of learning
side did not correlate with Median-force or #max (Spearman rank from one side to the other. In humans it appears well established
correlation rho: P > 0.1, Table 2). that after practicing a motor learning task with one limb sub-
sequent performance of the same task with the untrained limb
is facilitated (Kirsch and Hoffmann, 2010). Similarly to humans
4. Discussion (Gibson, 1939), horses appear to easily discriminate between tac-
tile stimuli given in different places on the body (Dougherty and
The results confirmed our main hypothesis. There was a sig- Lewis, 1993), but results in the present study suggest that horses
nificant decrease in force through the trials within the first day do not transfer the information about what the stimulus means. The
and the regression intercept was significantly higher on the first critical brain commissure for interhemispheric transfer is the cor-
day compared to the second and third day. This indicates that the pus callosum (Gooijers and Swinnen, 2014). There is no evidence
horses learned the task through negative reinforcement. Learning that horses’ corpus callosum should be less developed than in other
occurred primarily on the first day whereas the lower intercept on species (Olivares et al., 2001), and Hanggi’s study (Hanggi, 1999)
day 2 compared to day 1 indicates that the horses remembered on a limited number of horses has suggested that horses are able to
the task. Hence, we present a novel, standardised test to evaluate transfer a learned visual discrimination from one eye to the other.
From a practical perspective, however, the result was expected and
Table 2 may be explained by horses’ being very stimulus bound in their
Correlations (Spearman’s rho) between measures of individual negative reinforce- learning (Andrew McLean, personal communication).
ment performance of Icelandic horses. Significant (P < 0.05) values are marked with We did not find any difference between mares and geldings,
bold and tendencies (0.05 < P < 0.10) are in Italics. which was expected and in line with most available literature
Median-force #max Slope1 right (e.g. Visser et al. (2003) and Wolff and Hausberger (1996)). It was
Frequency of max (#max) 0.84
unexpected, however, to find that the previous training of half the
Change in force over trials 0.24 0.13 horses tended to influence the performance in this study, because
(Slope1): right side none of the exercises in this training included hindquarter yielding
Change in force over trials 0.49 0.35 0.32 explicitly (Marsbøll and Christensen, 2015). One of the exercises
(Slope1): left side
consisted of yielding to the side with the entire body, which may
42 L.P. Ahrendt et al. / Applied Animal Behaviour Science 169 (2015) 38–42

suggests that some transfer of behaviour between exercises on the For future studies using this method, we first recommend to use
same side of the horse may occur. However, the tendency towards an algometer with an actuator calibrated to measure higher force.
trained horses requiring lighter force could also relate to trained Secondly, we recommend using a combination of median-force and
horses having a more well-developed coordination necessary to slope to evaluate individual learning performance, as these meas-
perform the sideways movement compared to the non-trained con- ures may explain different parts of learning this task.
trol horses, thereby making the exercise in the present study easier
to perform. Acknowledgements

4.1. Evaluation of individual performance We would like to thank the horse owners for kindly allowing
us to use their horses; Anna Feldberg Marsbøll and Maria Vilain
Currently, there exists no gold standard to evaluate equine neg- Rørvang for help with the experiments. This study was funded by
ative reinforcement learning ability. The way learning performance The Independent Danish Research Council|Technology and Produc-
is measured in this test, or any learning test, may influence the eval- tion (grant no. 11-107010) and The Graduate School of Science and
uation of the individual horse’s learning ability and which factors Technology, Aarhus University.
influence the performance. The reason for evaluating performance
through different measures is that it potentially explains different References
characteristics of learning. In this study we proposed three poten-
tial individual performance values, based on the initial analysis of Baragli, P., Vitale, V., Paoletti, E., Sighieri, C., Reddon, A.R., 2011. Detour behaviour in
horses (Equus caballus). J. Ethol. 29, 227–234.
the applied force. Median-force and frequency of max-trials (i.e. Chiou, S.H., Kang, S., Kim, J., Yan, J., 2014a. Marginal semiparametric multivariate
where the force exceeded the 30 N limit of the algometer) on the accelerated failure time model with generalized estimating equations. Lifetime
first day correlated significantly, i.e. a horse with a high median Data Anal. 20, 599–618.
Chiou, S.H., Kang, S., Yan, J., 2014b. Fitting accelerated failure time models in routine
also had many trials exceeding 30 N. This was expected but not
survival analysis with R package aftgee. J. Stat. Softw. 61, 1–23.
theoretically necessary, because the median could take any value Christensen, J.W., Ahrendt, L.P., Lintrup, R., Gaillard, C., Palme, R., Malmkvist, J.,
between 0 and 30 N, when the frequency of max-trials was below 2012. Does learning performance in horses relate to fearfulness, baseline stress
5. In contrast to the median and frequency of max-trials, which hormone, and social rank? Appl. Anim. Behav. Sci. 140, 44–52.
Domjan, M., 2010. The Principles of Learning and Behaviour, sixth ed. Wadsworth
may express individual tactile sensitivity, the slope on Day 1 may Cengage Learning TM, USA, pp. 155.
express the ‘speed of learning’. The slopes on the left and right side, Dougherty, D.M., Lewis, P., 1993. Generalization of a tactile stimulus in horses. J. Exp.
however, did not correlate and only the slope on the left side corre- Anal. Behav. 59, 521–528.
Gabor, V., Gerken, M., 2010. Horses use procedural learning rather than con-
lated with the median. This may be explained by the slopes being ceptual learning to solve matching to sample. Appl. Anim. Behav. Sci. 126,
rather unstable. The slopes were calculated from only 10 trials and 119–124.
may, thus, be easily influenced by a few trials where a high force Gibson, E., 1939. Sensory generalization with voluntary reactions. J. Exp. Psychol.
24, 237–253.
was required. This is further supported by the result that the sides Gooijers, J., Swinnen, S.P., 2014. Interactions between brain structure and behavior:
did not correlate on the second and third day either, both of which the corpus callosum and bimanual coordination. Neurosci. Biobehav. R 43, 1–19.
were calculated from yet fewer repetitions. Hanggi, E.B., 1999. Interocular transfer of learning in horses (Equus caballus). J.
Equine Vet. Sci. 19, 518–524.
An often used, easy to compare, performance value is number Jin, Z., Lin, D.Y., Ying, Z., 2006. On least-squares regression with censored data.
of sessions to accomplish a set learning criterion. In our study this Biometrika 93, 147–161.
was not possible for two reasons; firstly we did not know at which Kalbfleisch, J.D., Prentice, R.L., 2011. The Statistical Analysis of Failure Time Data,
13., second ed. Wiley-Blackwell, New Jersey, USA, pp. 195.
force we should set this criterion, since this approach had not been
Kirsch, W., Hoffmann, J., 2010. Asymmetrical intermanual transfer of learning in a
tried before. The other, perhaps more important reason, was that sensorimotor task. Exp. Brain Res. 202, 927–934.
horses may have different threshold for when the pressure became Marsbøll, A.F., Christensen, J.W., 2015. Effects of handling on fear reactions in young
sufficiently motivating for them to offer a response (McGreevy and Icelandic horses. Equine Vet. J., DOI: 10.1111/evj.12338.
McCall, C.A., Salters, M.A., Simpson, S.M., 1993. Relationship between number of
McLean, 2010). In earlier studies using avoidance learning tasks, the conditioning trials per training session and avoidance learning in horses. Appl.
horses’ reaction threshold to electric shock was established before Anim. Behav. Sci. 36, 291–299.
initiating the task, to ensure the power of the shock to be sufficient McGreevy, P.D., Rogers, L.J., 2005. Motor and sensory laterality in thoroughbred
horses. Appl. Anim. Behav. Sci. 92, 337–352.
to get the horses to react (McCall et al., 1993). However, the aim McGreevy, P.D., McLean, A.N., 2010. Equitation Science. Wiley-Blackwell, Oxford,
in these studies was not to investigate if the power of the shock UK. pp. 91, 89, 105.
could be decreased by learning but rather to use it as a motivator Murphy, J., Sutherland, A., Arkins, S., 2005. Idiosyncratic motor laterality in the horse.
Appl. Anim. Behav. Sci. 91, 297–310.
to teach the horses to react to a conditioned stimulus and thereby Nicol, C.J., 2002. Equine learning: progress and suggestions for future research. Appl.
avoid the electric shock. This study is the first step in the devel- Anim. Behav. Sci. 78, 193–208.
opment of a standardised test for negative reinforcement learning Olivares, R., Montiel, J., Aboitiz, F., 2001. Species differences and similarities in the
fine structure of the mammalian corpus callosum. Brain Behav. Evol. 57, 98–105.
in horses. Further studies are required to validate the test and it R Core Team, 2014. R: A Language and Environment for Statistical Computing. R
should be investigated if the asymptotic level could be used as a Foundation for Statistical Computing, Vienna.
learning criterion, which would ease the analysis of individual per- Valenchon, M., Lévy, F., Prunier, A., Moussu, C., Calandreau, L., Lansade, L., 2013.
Stress modulates instrumental learning performances in horses (Equus caballus)
formance. Further studies are also required to relate performance
in interaction with temperament. PloS One 8, 1–10.
in this test to tactile sensitivity. The performance in this test should Visser, E.K., van Reenen, C.G., Schilder, M.B.H., Barneveld, A., Blokhuis, H.J., 2003.
also be related to other tests of equine learning ability and also to Learning performances in young horses using two different learning tests. Appl.
practical training of horses. Anim. Behav. Sci. 80, 311–326.
Wolff, A., Hausberger, M., 1996. Learning and memorisation of two different tasks
In conclusion, the presented test can be used to evaluate neg- in horses: the effects of age, sex and sire. Appl. Anim. Behav. Sci. 46, 137–143.
ative reinforcement learning in horses in a standardised manner.

You might also like