The impact of kitten training and socialization classes on cat cognitive bias and discrimination learning

Cats rank among the world’s most popular companion animals. Despite their widespread presence in human homes, accessible training and socialization opportunities for kittens remain scarce. In this study, a 6-week training and socialization class was offered for kittens aged 3-8 months old. Class kittens were compared to a group of control kittens that did not take part in the training and socialization class. Both groups of kittens participated in a cognitive bias test to measure discrimination learning and emotional outlook. Discrimination learning was measured in terms of passing criteria on the cognitive bias test, and the kittens’ responses to an ambiguous stimulus were measured to gauge optimistic or pessimistic outlooks. It was predicted that training and socialization would improve learning and lead to more optimistic judgments in class kittens. Results indicate that the training and socialization class helped maintain the ability to learn the task discrimination over time in class kittens, whereas control kittens that lacked the training and socialization experience displayed a decrease in task discrimination over time (Total N = 63; Experimental = 31, Control = 32). There was no significant difference between the groups in cognitive bias latencies (Total N = 36; Experimental = 17, Control = 19). However, interestingly, both groups showed shorter latency to approach an ambiguous stimulus over time, suggesting a naturally optimistic shift in kittens. This study reports the first use of the cognitive bias test in pet cats, and the findings contribute to our understanding of how socialization and training influence feline cognition and emotional well-being.

Introduction

Emotions are biologically relevant states that may help animals pursue beneficial outcomes while avoiding harm (Paul et al., 2005). The emotional experiences of non-human animals are a topic of growing interest to scientists, pet caregivers, and the public (Paul and Mendl, 2018; Mendl et al., 2010a). Affective states include pleasant, low-arousal states (i.e., calm and relaxed) which often correspond to encountering appetitive rewards (Gruen et al., 2019) and low levels of threat in the individual’s environment. On the other hand, negative, high-arousal states (i.e., fear and frustration) often correspond to threats or danger in an individual’s environment (Mendl et al., 2010b). Life experiences may shape these affective states, sometimes resulting in predictable shifts in the way an animal responds to new or ambiguous situations.

Underlying affective states may be evaluated in animals by observing physiological and behavioral responses (Mendl et al., 2009). The cognitive bias test has been used as a measure of emotional well-being and optimism and examines the underlying affective state of the animal, or the “optimistic” or “pessimistic” outlook of the individual when confronted with an ambiguous situation or stimulus (Mendl et al., 2010a). To date, cognitive bias tests have been conducted with several non-human animals such as livestock, domestic dogs, animals in research centers (including one study with cats housed in a research facility), zoological parks, and laboratories (Mendl et al., 2010a; Tami et al., 2011; Bethell et al., 2012; Douglas et al., 2012; Richter et al., 2012; Clegg, 2018; Harding et al., 2004). Cognitive bias may be measured by utilizing a task with a reward/no-reward outcome (Mendl et al., 2009). In the cognitive bias test, the animal is conditioned to learn the difference between a location that provides a reward and a location where no reward is provided, known as discrimination learning. If the animal learns to discriminate these locations, they are then faced with a stimulus in an ambiguous location. The animal’s approach to this ambiguous stimulus is used to determine their outlook The speed at which the animal approaches the ambiguous stimulus can be compared relative to their approach times to rewarded and non-rewarded stimuli, in relation to the scores of other individuals, and in relation to previous approach speeds at other timepoints (e.g., see Duranton and Horowitz, 2019; Vieira de Castro et al., 2020). A shorter approach latency to the ambiguous stimulus is aligned with a more optimistic outlook (i.e., faster approach due to expectation for reward), and a longer approach latency is aligned with a more pessimistic outlook (i.e., slower approach due to lack of expectation for reward).

Life experiences such as training have been shown to influence emotional state and cognitive bias. In one study, dogs were placed into groups depending on the training techniques implemented by their caregivers (Vieira de Castro et al., 2020). Dogs that received reward-based training were compared to dogs that were trained using aversive methods. The training group of the dogs (reward vs. aversive) and the location of the bowl (rewarded, non-reward, and intermediate location) had a significant effect on how long it took the dogs to reach a bowl. Dogs approached the rewarded location quickly compared to the negative location, which was not rewarded. Dogs in the aversive training group were significantly more pessimistic, as noted by their slower approach to ambiguous stimuli (Vieira de Castro et al., 2020; Casey et al., 2021). Additionally, dogs that participated in a nosework training class, which promoted the species-typical behavior of sniffing, displayed more optimistic outlooks (i.e., decreased latency to approach ambiguous stimulus), compared to dogs that engaged in heel work training classes, which did not promote natural, olfactory behaviors (Duranton and Horowitz, 2019). In all, these findings suggest that prior training, specifically reward-based training that highlights biologically relevant behaviors, may positively influence the affective outlook of animals.

Training experience may not only impact an individual’s affective outlook but may also impact an individual’s future learning and problem-solving success. For example, in prior research with dogs, those with general training experience were shown to later be more successful in solving a food puzzle task (Marshall-Pescini et al., 2008). In a different study, dogs with prior training experiences were faster at solving a string pulling task (Osthaus et al., 2003). Socialization to social partners may also impact affective outlook and learning. Socialization is the process in which appropriate intraspecific and interspecific social behaviors are developed (Serpell, 1988; Turner, 2000). Like dogs, cats, are born with the capacity to learn both species-specific and interspecies social skills (Crowell-Davis et al., 2004). Intraspecific and interspecific socialization are separate processes, and cats must be exposed to members of their own species as well as to other species, like humans, in order to develop typical social behavior and become less fearful. With these considerations, both prior training experiences and socialization may be important factors to consider in terms of animal cognitive outlook and learning success. However, to date no research has examined the impact of training and socialization classes on kitten cognition and learning.

Although there is a wealth of training and socialization opportunities available to dogs and dog caregivers (Howell et al., 2015), organized opportunities for socialization and training of cats are more scarce, despite the benefits of these opportunities. Research with cats indicates that carrier training may reduce cat stress during travel to veterinary exams (Pratsch et al., 2018), that clicker training promotes exploratory behavior and time spent at the front of the cage for shelter cats (Grant and Warrior, 2019), and that additional training improves the performance of cats on visual perceptual tasks (Sasaki et al., 2010). The initial sensitive period for socialization in kittens appears to occur around the 2-7 weeks of life (Karsh and Turner, 1988), which means that during this developmental period, kittens are thought to be most sensitive to initial learning about stimuli associated with social partners as they form early attachments. During this time period, 5-week-old kittens that had been socialized to several different humans displayed reduced fear toward people (as measured by approach/retreat from the human) (Collard, 1967). While socializing kittens during this sensitive period plays an important role in their development, ongoing socialization and developmental experiences are also important to future behavior and cognition. Given that veterinarians recommend neutering kittens around 22 weeks old (Murray et al., 2008), some adopters may be bringing home older kittens outside of this window. Additionally, research indicates that exposing young kittens to enrichment toys too early may increase the likelihood of heightened fear responses later in development (Graham et al., 2024). Caregivers may also be hesitant to bring kittens to public cat training classes until some or all of their initial vaccinations have been received. Therefore, an older sample of kittens may be more beneficial and representative of the population of kittens that would be eligible for caregivers to bring to kitten training classes if they were more commonplace. Social interaction also continues to be relevant in the lives of older cats, and many cats even prefer human social interaction over other appetitive stimuli like food rewards (50% of cats tested preferred human social interaction, 37% preferred food rewards, and the remainder preferred a toy or scent item) (Vitale Shreve et al., 2017). Given the benefits of training and human social interaction, it is possible that enrichment in the form of a training and socialization class may also influence a cat’s emotional outlook and their future learning success.

The current body of work suggests that a training and socialization class may be a beneficial intervention for companion cats to improve cat emotional outlook and learning. The aim of this study was to assess the potential influence of a training and socialization class on two aspects of cat cognition: discrimination learning and cognitive bias latency. It was predicted that, compared to the control group and following the class experience, kittens in the experimental class group would be more successful in completing the discrimination task (i.e., pass criteria to reach the end of cognitive bias task). Additionally, prior experience of positive reinforcement training may possibly make kittens more motivated to behave in a way that could result in positive reinforcement. Therefore, it was also predicted that kittens in the experimental class group would show behavioral patterns consistent with a more optimistic outlook (i.e., shorter approach latencies to an ambiguous stimulus) following the class experience.

Materials and methods

Ethics approval

This research was conducted in compliance with regulations set forth by Oregon State University’s Institutional Animal Care and Use Committee (IACUC), under ACUP #4674.

Subjects

Kittens and their caregivers were recruited via flyers, via online postings (e.g., Craigslist and Reddit), and by word of mouth. A total of 100 human–cat dyads were recruited as part of the current study, as well as for additional projects as part of the cat cognition research program (e.g., Vitale et al., 2019). Cat caregivers received up to $75 in Petco gift cards depending on their level of participation in the research sessions. Caregivers received $25 for participating in baseline lab testing, $25 for participating in follow-up lab testing, and $25 for participating in at-home testing at both baseline and follow-up (for a different research study).

For enrollment into groups, kittens had to be aged between 3 and 8 months old and be a companion cat living in a human home. Kittens were pseudo-randomly assigned to one of two groups: (1) 50 cats were assigned to the Experimental group—kitten training and socialization (i.e., “class kittens group”), and (2) 50 cats were assigned to the Control group. Age and sex of cat were considered for consistency between groups; however, no kittens were reassigned based on age or sex. Group assignments were at times influenced by scheduling, as the kitten class ran in sessions that aimed to involve 10 kittens participating at once. There were a few instances in which participants were automatically assigned to the class group as there was a need to increase class enrollment or automatically assigned to the control group as the final class had been offered. The 50 class kittens ranged in age from 12 to 32 weeks at baseline test (mean = 20 weeks, SD ± 5.6 weeks) and included 30 males and 20 females. All kittens were of domestic breed except for 1 Bengal and 3 Siamese kittens based on caregiver report. The 50 control group kittens ranged from 11 to 32 weeks of age at baseline test (mean = 20.46, SD ± 7.5 weeks) and included 30 males and 20 female kittens. All kittens were of domestic breed except for 1 Bengal and 1 Siamese based on caregiver report.

Following baseline assessments, class kittens (50) participated in the kitten training and socialization class group, and control kittens (50) did not participate in this class. After 6 weeks, all dyads were invited back to participate in a follow-up assessment.

The subject sample size depends on the dataset. The dataset for all kittens which participated in both baseline and follow-up test included 39 subjects in the experimental class group and 36 subjects in the control group (N=75). Of the 39 class kittens, 16 were female and 23 were male and ranged in age from 12 to 32 weeks old at baseline test (mean = 19.2 weeks, SD ± 4.9 weeks). Of 36 control kittens, 17 were female and 19 were male and ranged in age from 12 to 32 weeks old at baseline test (mean = 20.4 weeks, SD ± 7.7 weeks).

The discrimination learning dataset included 31 subjects in the experimental class group and 32 subjects in the control group (N=63). Of the 31 class kittens, 10 were female and 21 were male and ranged in age from 12 to 32 weeks old at baseline test (mean = 18.6 weeks, SD ± 4.61weeks). Of 32 control kittens, were 15 female and 17 were male and ranged in age from 12 to 32 weeks old at baseline test (mean = 20.03 weeks, SD ± 7.5 weeks).

The cognitive bias latency dataset included 17 subjects in the experimental group and 19 in the control group (N=23). Of the class kittens, 6 were female and 11 were male and ranged in age from 12 to 26 weeks old at baseline test (mean = 18.2 weeks, SD ± 3.9 weeks). Of control kittens, 9 were female and 10 were male and ranged in age from 12 to 32 weeks old at baseline test (mean = 18.9 weeks, SD ± 7.3 weeks).

Individual kitten demographic data for each analysis is provided in the Supplementary File.

Kitten training and socialization classes

The 50 kittens assigned to the kitten class group participated in one of five course sessions, which were offered between Winter 2016 and Fall 2017. Each kitten training and socialization class met once a week for 6 weeks on Oregon State University campus. Courses included participation of between 7 and 12 kittens per offering (average of 10 cats per course session). Each class session was 45 min and included a 15-min lecture and then 30 min of socialization/training time. Caregivers were also expected to work with their kittens outside of class time, in their own home. Over the 6-week class, caregivers learned to implement positive reinforcement techniques to train basic behaviors, and teach skills that have the potential to improve human–cat communication and reduce cat stress (Pratsch et al., 2018). Behaviors worked on within the class include come when called, sit, go to mat and stay, targeting, and walk on harness and leash. Additional class activities included facilitating social interactions between kittens and unfamiliar humans, as well as between kittens and unfamiliar conspecifics. Kittens were encouraged to explore the novel environment which included cat towers, scratching surfaces, hiding locations, treats, toys, and litter boxes. For additional details on the kitten training and socialization class intervention, see the Supplementary Information Document, Document S1. Supplementary Experimental Procedures in Vitale et al. (2019).

Behavioral assessments/cognitive bias test

Cognitive bias tests were conducted over two sessions, once at baseline (before the kitten class) and once approximately 6 weeks later (after the experimental group had completed the kitten class) to measure discrimination learning and optimism/pessimism in the presence of an ambiguous stimulus.

The test took place in the Human-Animal Interaction Lab at Oregon State University (OSU), which was an unfamiliar location to all kittens at baseline test. Caregivers brought their kittens to OSU campus in a carrier or on a harness and leash. Caregivers waited in a separate room during the cognitive bias test but were able to watch the test via a live camera feed on a computer. An experimenter brought the kitten into the test room. A start line for the kitten was indicated by tape on the ground. Located 4 m from this line were three locations, S+ (positive, reward location), S− (negative, no reward location), and Sn (ambiguous location, see Figure 1). The location of the side for S+ and S− was counterbalanced between subjects.

Figure 1

Figure 1. Cognitive bias test setup. Kittens were released from a start line facing three possible human locations, each 4 m away. Persons #1 and #3 were either S+ or S−, with the location counterbalanced across cats. The ambiguous person (Sn) was always presented in the center.

Each session of the cognitive bias test followed this general outline (described in further detail below). The session began with the Conditioning Phase, and each cat encountered four conditioning trials (two S+ trials then two S− trials). Next, a minimum of six pseudorandom trials were presented to the cat (but only for cats that approached during the four conditioning trials). After the six pseudorandom trials, the experimenters assessed whether the cat had met learning criteria. If so, a final trial presenting the ambiguous human (Sn) was presented. If not, trials continued until the learning criteria had been met or the test was ended due to kitten stress behavior. Data on number of trials each cat took to meet learning criteria are included under Results in the Pseudorandom Trials section.

In the current study, the cognitive bias methods were a modified version of those previously published with domestic dogs (Mendl et al., 2010a). In prior research, to reinforce dog approaches to the positive position, a small amount of food was placed into a bowl while the negative location contained an empty bowl with no food reward. Given that human social interaction is a highly preferred stimulus for cats (Vitale Shreve et al., 2017), in the present study, three separate people served to provide social interaction and/or food directly to the cat, one at a time, without use of a bowl (Figure 2). Before trials began, the experimenter (a person besides S+, S−, or Sn) provided the kitten with a treat or the option to receive petting or play with a feather toy to decide preferred rewards to be used by the person in the S+ location. It should also be noted that across testing sessions, the individual person and their role in the testing session was not kept consistent, and therefore a research assistant may have encountered the cat previously prior to the follow-up test, but would nonetheless have still been relatively novel to the cat compared with an caregiver or regular human companion.

Figure 2

Figure 2. Screenshots from the cognitive bias test. Images show (a) trial in which kitten approached S+, who provided petting and praise as a reward, (b) trial in which kitten approach S−, who ignored cat and provided no reward, and (c) the last trial in which the kitten was presented with the ambiguous human, Sn. A separate experimenter begins the kitten behind the start line.

Conditioning phase

Each trial began once the cat was at the starting line, and released by the experimenter, and could last for up to 35 s or until the cat made a choice by approaching either the S+ or S− and crossing the line associated with proximity to one of these locations (seen in Figure 2). In Mendl et al. (2010a), each trial could last a maximum of 30 s. In the present study, each trial was made slightly longer at 35 s to account for possible species differences in exposure to the experimental setup (e.g., see Uccheddu et al., 2022).

First two S+ trials: During the first trial in the conditioning phase, an unfamiliar research assistant approached the (S+) location, 4 m from the subject, and sat down. The unfamiliar human called the cat’s name twice to gain its attention and then sat quietly and neutrally until the cat made a choice. The cat was then released at the start line by the experimenter. If the kitten approached the person at the S+ location once released, the kitten received a reward (praise, petting, playing, or treats). The cat was brought back to the starting line by the experimenter, and another S+ trial was repeated a second time. As long as the kitten approached the S+ location at least once the trials continued.

First two S− trials: For the next two trials in the conditioning phase, a different unfamiliar research assistant approached the (S−) location, 4 m from the subject, and sat down. The unfamiliar human called the cat’s name twice to gain its attention and then sat quietly and neutrally. If the kitten approached this human at the S− location once released, the kitten received no reward (instead this human always remained neutral independent of the cat’s behavior). At the end of the trial, the kitten was brought back to the starting line by the experimenter, and the next trial began. As long as the kitten approached the S− location at least once, the trials continued.

Pseudorandom trials

Additional conditioning trials continued pseudo-randomly with the condition that no more than two trials of the same type could occur in a row. On each trial, the kittens had up to 35 s to approach the human. If the kitten did not approach the human, then the trial ended at 35 s and a time of 35 s was recorded (Mendl et al., 2010a) along with a note that the cat made “no choice”. If the kitten made four “no choices” in a row the session ended, and these kittens were not presented with the ambiguous stimulus trial (Sn).

Discrimination training (conditioning) criteria

The first discrimination training assessment occurred after six pseudorandom conditioning trials. The learning criteria was met if the longest latency of the S+ approaches was shorter than any of the latencies to approach S−, as in Mendl et al. (2010a). If criteria were not met, conditioning trials continued until the criteria was reached (reassessed after each following trial). Discrimination training stopped when the learning criteria had been met, a kitten made four “no choices” in a row (marked as “NC” in data), or the kitten displayed excessive stress behavior or aggression (e.g., more than one of the following behavioral indicators: dilated pupils, excessive lip-licking, flattened ears, crouched body posture, hissing or growling, marked as “NCM” in data). Only cats meeting the learning criteria moved on to the ambiguous testing phase.

Ambiguous stimulus phase

After a cat met learning criteria in the pseudorandom conditioning trials (i.e., the kitten had learned the association between the S+ human/location and the delivery of reinforcers), they entered the ambiguous stimulus phase of the cognitive bias test. A novel unfamiliar human (Sn) sat in an ambiguous position, again 4 m from the subject, but in the middle of the room, between where S+ and S− had been located (Figure 2). The ambiguous human called the cat’s name twice and then sat neutrally until the end of the trial. The kitten’s latency to approach Sn was recorded in one trial. Only a single ambiguous trial was performed for each cognitive bias test. For comparisons between the experimental class and control group, and for comparisons between baseline (timepoint 1) and follow-up (timepoint 2), shorter median latencies to approach on the ambitious trial were interpreted as a more “optimistic” response. Relative approach time across groups and conditions for S+ and S− were also conducted to evaluate possible alternative interpretations, such as differences in overall speed of approach or general boldness. For kittens that were presented with Sn, but did not approach, a score of 35 s was written, as the full time allotted had also been written in Mendl et al. (2010a).

Analyses

Kittens were excluded from data analysis if the kitten did not participate in both the baseline and follow-up assessments. Additionally, kittens that did not complete the training and socialization class (i.e., did not attend the majority of class sessions) were not eligible to participate in the follow-up testing and excluded from analysis. No kittens were removed from groups due to cat behavior (e.g., human-directed aggression).

All statistical tests had an alpha level of 0.05.

Conditioning phase

The number of individuals approaching the first human presented during the conditioning phase of the test were analyzed using Fisher’s exact tests run in GraphPad. Comparisons were conducted between groups at baseline, prior to the experimental group’s participation in the socialization and training class, and at follow-up, after the experimental group’s participation in the socialization and training class. The number of kittens in each group either passing or failing to pass the conditioning phase were compared (i.e., approached during the conditioning phase or failed to approach consistently during the conditioning phase). A 2×2 contingency table was analyzed by grouping kittens by if they “passed” and approached during the conditioning phase or “failed” and did not approach at all or only approached once during the four-trial conditioning phase. Subjects that failed the conditioning phase were removed from further analyses with discrimination learning and latency datasets.

Discrimination learning data

Data on discrimination learning were analyzed for kittens that passed the initial conditioning phase of the test. Data on discrimination learning were analyzed using Fisher’s exact tests run in GraphPad. Comparisons were conducted between groups at baseline and at follow-up, Additionally, data were compared within groups, to examine if kittens in each group became more successful from the baseline to follow-up testing. The number of kittens in each group either passing or failing to meet criteria during the pseudorandom trials of the cognitive bias test were compared. A 2×2 contingency table was analyzed by grouping kittens by if they “passed” and met discrimination learning criteria or “failed” to meet learning criteria.

Cognitive bias latency scores

Data on cognitive bias latency scores were analyzed using Mann–Whitney U tests for comparisons between groups, and two-tailed Wilcoxon signed-rank tests were used for repeated measures with the same individuals. Both tests were run in the Social Science Statistics Calculator. To examine if cognitive bias latency differed between class and control groups for subjects which had numerical latency scores at both timepoints, analyses were run on latency to approach for the ambiguous condition (Sn) for only subjects that received a final numerical latency score in both the baseline and follow-up tests (i.e., had passed the discrimination training criteria and been presented with Sn) and also had full data on latency scores (to Sn, S+, and S−). This includes kittens that were presented with the Sn, but did not approach, as a numerical score of 35 s was recorded for these trials. In all, a total of 17 class kittens and 19 control kittens had numerical scores at both testing sessions and were included in the latency analysis.

To examine latency change in cognitive bias score in the pre- and post-training periods, the baseline latency score was subtracted from the follow-up latency score (follow-up–baseline). Here, a negative number indicates a decrease in latency to approach the ambiguous human (quicker approach, optimistic trend) and a positive number indicates an increase in the latency to approach the ambiguous human (slower approach, pessimistic trend).

To examine overall change in latency for all kittens, the baseline and follow-up data for all cats were compared, with the class group and control groups combined (N=36) and compared from baseline to follow-up session. The approach speed to S+ and S− was also examined with the class and control groups combined. The average latency to approach S+ and S− were calculated by averaging the pseudorandom trial approaches for both S+ and S− separately.

Results

Conditioning phase

A total of 39 experimental kittens and 36 control kittens participated in the cognitive bias test at both baseline and follow-up sessions. Of these kittens, a portion of them did not approach during the initial conditioning trials, or only approached a single time, and therefore did not pass the conditioning phase. The number of kittens passing or failing the initial conditioning phase were compared across groups at both baseline (Table 1) and follow-up (Table 2). No significant difference was seen between class and control groups in the number of individual kittens passing or failing the initial conditioning phase at baseline (p=0.12) or follow-up (p = 0.51).

Table 1

Table 1. Baseline data for the number of class and control kittens which passed or failed the initial conditioning phase.

Table 2

Table 2. Follow-up data for the number of class and control kittens which passed or failed the initial conditioning phase.

The total number of kittens (experimental and control groups combined) that either passed or failed the conditioning phase was compared between baseline and follow-up. No significant difference was found in the number of individual kittens passing or failing the initial conditioning phase (p=0.099).

A total of eight class kittens and four control kittens failed to approach at one or both of the testing sessions and were removed from further analysis on discrimination learning and latency scores.

Discrimination learning baseline data (before training and socialization class)

The data from class and control kittens were compared for all kittens that passed the initial conditioning phase. No significant difference was seen between class and control groups in the number of individual kittens meeting learning criteria (i.e., passing or failing) in the cognitive bias test (Table 3, p = 0.302).

Table 3

Table 3. Baseline data for the number of class and control kittens which met criteria to encounter the ambiguous person (i.e., passed criteria) and the number of kittens which did not meet criteria to encounter the ambiguous person (i.e., failed criteria).

Discrimination learning follow-up data (after training and socialization class)

Kittens in the class group were compared with control kittens to determine if the training experience influenced the number of kittens that met learning criteria in the cognitive bias test. There were no significant differences between the groups in the number of class and control kittens passing/failing to meet criteria at follow-up testing (Table 4, p = 0.60).

Table 4

Table 4. Follow-up data for the number of class and control kittens which met criteria to encounter the ambiguous person (i.e., passed criteria) and the number of kittens which did not meet criteria to encounter the ambiguous person (i.e., failed criteria).

Data were also analyzed within groups using a 2×2 contingency table to examine if kittens in each group became more successful from the baseline to follow-up testing. There was no significant change in the number of class kittens to meet learning criteria in the baseline versus follow-up sessions (25 passed at baseline, 22 passed at follow-up, p = 0.55). However, when comparing the number of control kittens to meet learning criteria at the baseline and follow-up sessions, significantly fewer control kittens successfully completed the discrimination training portion of the test in follow-up compared with baseline (29 passed at baseline, 20 passed at follow-up, p = 0.016).

Pseudorandom trials

The number of pseudorandom trials it took for cats to reach criteria are included in the Supplementary File. At baseline, it took a median of six trials for cats to reach discrimination criteria with a range in the number of trials from 6 to 10 trials. At follow-up, it also took a median of six trials for cats to learn criteria with a range in the number of trials from 6 to 16 trials.

Cognitive bias latency scores

To examine if cognitive bias latency differed between class and control groups for subjects which had numerical latency scores at both timepoints, analyses were run on latency to approach for the ambiguous condition (Sn). Kittens with numerical scores at both testing sessions and were included in the latency analysis. No significant difference was seen in terms of latency to approach Sn for kittens with numerical scores at both time points at either baseline (U=146.5, z=0.46, p=0.65) or follow-up testing session (U = 129, z = −1.0, p = 0.31).

As mentioned, the latency change was calculated as a comparison of approach to Sn for each individual at both sessions. Each subject’s baseline latency score was subtracted from their follow-up latency score (follow-up–baseline) and a negative number indicates a decrease in latency to approach the ambiguous human (quicker/optimistic) whereas a positive number indicates an increase in the latency to approach the ambiguous human (slower/pessimistic) for that individual. Latency values at baseline and follow-up are reported in Table 5. Class and control groups did not differ significantly in latency change, (U = 115, z = −1.46, p = 0.14).

Table 5

Table 5. Median and range information for the latency of approach to the ambiguous human (Sn) in the cognitive bias test for kittens with numerical scores at both testing sessions. All latencies are reported in sec.

To examine overall change in latency for all kittens, the baseline and follow-up data for all cats were compared, with the class group and control groups combined (N=36). A significant difference was seen from baseline to follow-up session, with kittens becoming more optimistic in their approach to the ambiguous human over time (W = 125, z= −2.4. p = 0.016).

The approach on the first conditioning trial was also examined, before an association between S+ and a reward had yet occurred. This was the first trial each kitten would have experienced in each session and represents each kitten’s latency of approach to an unfamiliar human before any conditioning occurred. No significant difference was found in approach to the initially neutral S+ human in the first conditioning trial when comparing the same individuals from baseline to follow-up (W = 265.5, z = −0.27, p = 0.79), suggesting that increased approach to the ambiguous human (Sn) at the second testing time point cannot be explained by an increased tendency or motivation to approach new humans faster in general.

The average approach speed to S+ and S− (after conditioning) was also examined with the class and control groups combined. The average latency to approach S+ and S− were calculated by averaging the pseudorandom trial approaches for both S+ and S− separately. There was no significant difference found between the baseline and follow-up average latencies for approach to either the S+ (W = 252, z =−1.3, p = 0.20) or S− (W = 249, z = −1.3, p = 0.19).

Data were also examined to compare average latency of approach between the S+ and S− locations at baseline and follow-up for the same cats. As seen in Figure 3, kittens were significantly faster to approach S+ compared with S− at both baseline (W = 1, z = −5.2, p = < 0.00001) and follow-up (W = 0, z =−5.2, p = < 0.00001). Sn data were not directly compared with S+ and S− as Sn is an average of a single value for each individual, whereas S+ and S− are averages of multiple trials for each individual.

Figure 3

Figure 3. The average latency to approach S+ (white bar) and S− (shaded bar). Each dot represents an individual pet cat. The box indicates the interquartile range, with the upper whisker indicating scores above the middle 50% of the data and the lower whisker indicating scores below the middle 50% of the data. The bolded line indicates the median. *p < 0.05. The median value for average approach to the ambiguous person (Sn) is also marked using a circle with an x in the center.

Discussion

The results of this study indicate that the kitten training and socialization class experience had a positive impact on discrimination learning success over time. Class kittens were successful in completing the discrimination task at baseline and maintained the ability to meet the criteria on the discrimination task at follow-up, whereas control kittens, which lacked the class experience, displayed a decrease in discrimination learning success over time. Significantly fewer control kittens completed the discrimination training portion of the cognitive bias test in the follow-up test compared with the baseline test. These findings are in line with prior research, which has found that training experience and human socialization are important factors to consider in terms of animal learning (Osthaus et al., 2003; Marshall-Pescini et al., 2008). For cats specifically, training has been shown to reduce stress (Pratsch et al., 2018) and human socialization reduces fear of humans (Collard, 1967). The present study additionally indicates that training and socialization experiences may also help maintain a cat’s engagement and success on discrimination learning tasks over time. Together, this work indicates that training and socialization classes are a viable and beneficial option for kittens and could be more widely offered to cats and their caregivers to promote cat welfare.

Although it was predicted that kittens in the class group would show behavioral patterns more consistent with an optimistic outlook (i.e., shorter approach latencies to an ambiguous stimulus) than control cats following the class experience, no significant difference was found in the latency to approach the ambiguous human between the class and control groups. Instead, cognitive bias latencies were in fact shorter at follow-up for kittens in both groups, suggesting an optimistic shift overall. Across both groups, approach to the ambiguous human became quicker for 61% (22/36) of kittens at follow-up test.

Although these data suggest that kittens in both groups developed a more optimistic outlook over time, other explanations may be possible. For example, it is possible that the kittens simply became more coordinated, and therefore faster in their approach to Sn, with age. However, kittens display adult-like locomotion at 6–7 weeks of age and have fully developed motor coordination at 10–11 weeks of age (Martin and Bateson, 1986). Given that all kittens in the dataset were over 12 weeks of age (3–8 months old), all kittens in our sample were expected to have developed full motor coordination prior to the start of this study. Therefore, it is unlikely that these differences in approach time across times points were due to differences in motor coordination over testing sessions. Additionally, the findings that average approach speed to the other people in the trials (S+ and S−) did not differ from baseline to follow-up and that no significant difference was found in approach to the first person the cat encountered (S+ in the first conditioning trial) when comparing the same individuals from baseline to follow-up further support the idea that cats became more optimistic over time (i.e., in their approach to Sn), and not that kittens were only faster at approaching people more generally.

The possibility that kittens display a naturally optimistic shift over time is interesting, and worth exploring further, especially given that age has not been found to impact the optimistic or pessimistic expectations of other companion animals, such as young and old dogs, on cognitive bias tests (Piotti et al., 2018). In the future, a longer-term longitudinal study could follow kittens into adulthood to determine whether cats continue to shift to more optimistic outlooks within the same environment, or at broader populations to determine if cats in general are more likely to adopt more optimistic or pessimistic outlooks into adulthood or in advanced age.

Considerations and limitations

The cognitive bias test may have limited utility in informing recommendations for some species or in certain contexts. In a study with captive grizzly bears, cognitive bias results were not influenced by either the type of enrichment or the duration of interaction with enrichment items, whereas another behavioral measure (pacing duration prior to testing) was associated with optimistic outlooks (Keen et al., 2014). This suggests that the cognitive bias test did not effectively show differences in affective state after the enrichment intervention. However, other work supports that the cognitive bias task provides an accurate measure of an individual’s affective state, which may aid professionals in creating recommendations for animal husbandry and welfare (Gruen et al., 2019), such as to assess the efficacy of environmental enrichment. For example, it was found that pigs’ responses to an ambiguous auditory cue were influenced by environmental conditions, either enriched and barren environments. Pigs, when housed in the enriched environment, were more likely to approach a hatch, and did so more quickly, after hearing the ambiguous auditory cue, compared with when the same pigs were housed in a barren environment (Douglas et al., 2012). Additionally, captive rhesus macaques are more likely to engage with ambiguous stimuli following a period of environmental enrichment (Bethell et al., 2012) and captive European starlings are less likely to engage with stimuli following the removal of environmental enrichment (Bateson and Matheson, 2007). In all, this suggests that at times alternative measures to the cognitive bias test may be more appropriate; however, the cognitive bias test has been used in a number of species to assess the efficacy of environmental enrichment and individual affective state. Finally, some researchers have raised concerns about interpretation of the cognitive bias test and whether additional processes outside of the affective state, such as an animal’s hormonal state, impact responses on the test (Perdue, 2017). Given these perspectives, additional work in this area is greatly needed to more holistically assess the efficacy of the cognitive bias measure of affective states and its use to create applied recommendations.

Some considerations and limitations exist for this specific study. Caregivers of kittens who are more engaged with their pets, or who have kittens that already show more bold or optimistic tendencies, might be more likely to enroll in research opportunities and bring their cats to testing sessions, which may create a potential bias in participant recruitment and testing participation (Webster and Rutz, 2020). Given this, and that significant cultural differences have been identified in cat behavior and the cat–human relationship (Vitale et al., 2024), our results may not be generalizable across all populations. This challenge is not specific to work with cats; it has also been noted in work with other companion animal species including dogs (Koster, 2021) but highlights the need to consider environment and individual experience when discussing the capacity and range of cat behavior and cognition more broadly. This study contributes specifically to our knowledge of pet cat behavior, when under the care of highly engaged human caregivers. The broader field of research on cat cognition should also be inclusive of work with cats living in different environments including cats in shelters and cats in outdoor free-roaming colonies.

It should also be considered that the prerequisite discrimination training required to reach the ambiguous stimulus phase of the cognitive bias test may be challenging for many animals and therefore may limit its value in assessing optimism across full study populations. It is also important to consider that extraneous factors may affect a cat’s response latency on individual trials, especially given the novelty of the environment. Given this, alternative criteria for passing the discrimination training learning criteria could be implemented in the future. In the present study, a percentage of the sample failed to meet criteria for the cognitive bias task, with between 9% and 37.5% of individuals failing (depending on the group and timepoint, see Tables 3 and 4). It is important to note that this challenge is not unique to cats; prior studies with dogs have reported similar numbers of subjects failing to reach the testing phase of the cognitive bias test suggesting this is a procedural, not species-specific, challenge (Burani et al., 2020). While we still feel that the data collected provide meaningful insight into the cognitive bias of participating cats comparable with data collected in prior studies with other species, it is possible that alternative methods that are shorter in duration and do not focus on discrete trials may be beneficial to future work with cats (see Vitale Shreve et al., 2017), and perhaps other species as well, to reduce attrition. With that said, the majority of kittens in the sample did readily approach during the conditioning trials with between 74% and 100% of subjects passing the initial conditioning phase (depending on the group and timepoint, see Tables 1 and 2), which supports the use of this test with cats in future research.

Another consideration is that cognitive bias tests have been conducted in a number of different ways, for example with repeated sessions resulting in multiple ambiguous stimulus presentations (Truax and Vonk, 2025) or with a series of ambiguous stimulus probes with different stimulus properties or an intermediate location (McGuire et al., 2018; Mendl et al., 2010a). Data comparisons using multiple positive, negative, and ambiguous trials have also been analyzed in a number of different ways, for example using GLMMs that consider repeated positive, negative, and ambiguous trials in a single model (McGuire et al., 2018). In this study, we implemented a single ambiguous probe trial at each evaluation time point to remove the possible confound of learning across multiple rewarded/non-rewarded ambiguous trials and to reduce testing session length, a factor which has been associated with attrition in cognitive bias tests and in cat cognition studies more broadly (Vitale Shreve et al., 2017). However, future studies could consider evaluating cat performance on cognitive bias tests using alternative methods as well to evaluate if and how such choices may influence testing efficacy or outcomes.

Conclusions

Although kitten training and socialization classes have been historically uncommon, there is a clear growing demand for such opportunities from cat caregivers (Link and Moody, 2025) and there may also be a range of measurable benefits to providing these experiences for cats. The cat cognition research program, as well as the kitten training and socialization classes themselves, were well received by the public. Of the 50 kittens assigned to the training class, 86% of kittens and their caregivers completed the 6-week class. Additionally, kittens readily learned behaviors through the use of reward-based, positive reinforcement training (Vitale et al., 2019). The attendance and feedback from the training classes indicated general interest in future training and socialization opportunities for cat caregivers and their cats. Beyond simply teaching new behaviors, such classes have the potential to deepen the cat–caregiver bond and enhance cat welfare. There may also be cognitive benefits including maintenance of discrimination learning skills as demonstrated here, or enhanced social sensitivity. More research on cat behavior is also needed to help learn about a broader range of experiences that may be beneficial to cat wellbeing, to tease apart the factors of training and socialization, and also to help us better understand how other factors, including development, may influence changes in cat cognition over time. Although the cognitive bias test has been run with cats living in a research facility, to the authors’ knowledge, this study reports the first use of the cognitive bias in pet cats.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Ethics statement

The animal studies were approved by Oregon State University’s Institutional Animal Care and Use Committee (IACUC), under ACUP #4674. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent was obtained from the owners for the participation of their animals in this study.

Author contributions

KV: Conceptualization, Data curation, Formal Analysis, Funding acquisition, Investigation, Methodology, Writing – original draft, Writing – review & editing, Visualization. CM: Investigation, Resources, Visualization, Writing – original draft, Writing – review & editing. MU: Conceptualization, Formal Analysis, Funding acquisition, Methodology, Project administration, Supervision, Writing – review & editing.

Funding

The author(s) declared financial support was received for this work and/or its publication. KV was supported by the National Science Foundation GRFP under Grant No. (1314109-DGE). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

Acknowledgments

Thank you to the caregivers who participated in the kitten research program and the research assistants who helped support the kitten training classes and testing sessions.

Conflict of interest

This study received funding through a Nestlé Purina sponsorship for studies in cat and dog emotional well-being. Nestlé Purina reviewed an earlier draft of this manuscript prior to submission. However, the funder had no involvement in research activities or in the development of the manuscript.

Generative AI statement

The author(s) declare that Generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fetho.2025.1681085/full#supplementary-material

References

Bateson M. and Matheson S. M. (2007). Performance on a categorisation task suggests that removal of environmental enrichment induces “pessimism” in Puppy parties and beyond: the role of early age socialization practices on adult dog behavior. (Sturnus vulgaris). Anim. welfare 16, 33–36. doi: 10.1017/S0962728600031705

Crossref Full Text | Google Scholar

Bethell E., Holmes A., Maclarnon A., and Semple S. (2012). Cognitive bias in a non-human primate: husbandry procedures influence cognitive indicators of psychological well-being in captive rhesus macaques. Anim. Welfare 21, 185–195. doi: 10.7120/09627286.21.2.185

Crossref Full Text | Google Scholar

Burani C., Barnard S., Wells D., Pelosi A., and Valsecchi P. (2020). Using judgment bias test in pet and shelter dogs (Canis familiaris): Methodological and statistical caveats. PloS One 15, e0241344. doi: 10.1371/journal.pone.0241344

PubMed Abstract | Crossref Full Text | Google Scholar

Casey R. A., Naj-Oleari M., Campbell S., Mendl M., and Blackwell E. J. (2021). Dogs are more pessimistic if their owners use two or more aversive training methods. Sci. Rep. 11, 19023. doi: 10.1038/s41598-021-97743-0

PubMed Abstract | Crossref Full Text | Google Scholar

Clegg I. L. K. (2018). Cognitive bias in zoo animals: an optimistic outlook for welfare assessment. Animals 8, 104. doi: 10.3390/ani8070104

PubMed Abstract | Crossref Full Text | Google Scholar

Collard R. R. (1967). Fear of strangers and play behavior in kittens with varied social experience. Child Dev. 38, 877–891. doi: 10.2307/1127265

PubMed Abstract | Crossref Full Text | Google Scholar

Crowell-Davis S. L., Curtis T. M., and Knowles R. J. (2004). Social organization in the cat: a modern understanding. J. Feline Med. Surg. 6, 19–28. doi: 10.1016/j.jfms.2003.09.013

PubMed Abstract | Crossref Full Text | Google Scholar

Douglas C., Bateson M., Walsh C., Bédué A., and Edwards S. A. (2012). Environmental enrichment induces optimistic cognitive biases in pigs. Appl. Anim. Behav. Sci. 139, 65–73. doi: 10.1016/j.applanim.2012.02.018

Crossref Full Text | Google Scholar

Duranton C. and Horowitz A. (2019). Let me sniff! Nosework induces positive judgment bias in pet dogs. Appl. Anim. Behav. Sci. 211, 61–66. doi: 10.1016/j.applanim.2018.12.009

Crossref Full Text | Google Scholar

Graham C., Pearl D. L., and Niel L. (2024). Too much too soon? Risk factors for fear behaviour in foster kittens prior to adoption. Appl. Anim. Behav. Sci. 270, 106141. doi: 10.1016/j.applanim.2023.106141

Crossref Full Text | Google Scholar

Grant R. A. and Warrior J. R. (2019). Clicker training increases exploratory behaviour and time spent at the front of the enclosure in shelter cats; Implications for welfare and adoption rates. Appl. Anim. Behav. Sci. 211, 77–83. doi: 10.1016/j.applanim.2018.12.002

Crossref Full Text | Google Scholar

Gruen M. E., Foster M. L., Lazarowski L., Jeffries A., Sherman B. L., and Dorman D. C. (2019). Does the cognitive bias test in dogs depend on spatial learning? J. Veterinary Behav. 33, 1–6. doi: 10.1016/j.jveb.2019.04.002

Crossref Full Text | Google Scholar

Harding E., Paul E., and Mendl M. (2024). Cognitive bias and affective state. Nature. 427, 312–152. doi: 10.1038/427312a

PubMed Abstract | Crossref Full Text | Google Scholar

Howell T. J., King T., and Bennett P. C. (2015). Puppy parties and beyond: the role of early age socialization practices on adult dog behavior. Veterinary Medicine: Res. Rep. 6, 143–152. doi: 10.2147/VMRR.S62081

PubMed Abstract | Crossref Full Text | Google Scholar

Karsh E. B. and Turner D. C. (1988). “The human-cat relationship,” in The domestic cat: the biology of its behaviour. Eds. Turner D. C. and Bateson P. P. G. (Cambridge University Press, Cambridge), 157–177.

Google Scholar

Keen H. A., Nelson O. L., Robbins C. T., Evans M., Shepherdson D. J., and Newberry R. C. (2014). Validation of a novel cognitive bias task based on difference in quantity of reinforcement for assessing environmental enrichment. Anim. Cognit. 17, 529–541. doi: 10.1007/s10071-013-0684-1

PubMed Abstract | Crossref Full Text | Google Scholar

Koster J. (2021). Most dogs are not NATIVE dogs. Integr. Comp. Biol. 61, 110–116. doi: 10.1093/icb/icab016

PubMed Abstract | Crossref Full Text | Google Scholar

Link J. K. and Moody C. M. (2025). Socialising kitties: A quantitative survey of US cat owner attitudes towards kitten and adult cat socialisation programmes. Anim. welf. 34, e39. doi: 10.1017/awf.2025.10013

PubMed Abstract | Crossref Full Text | Google Scholar

Marshall-Pescini S., Valsecchi P., Petak I., Accorsi P. A., and Previde E. P. (2008). Does training make you smarter? The effects of training on dogs’ performance (Canis familiaris) in a problem solving task. Behav. Processes 78, 449–454. doi: 10.1016/j.beproc.2008.02.022

PubMed Abstract | Crossref Full Text | Google Scholar

Martin P. and Bateson P. (1986). “Behavioural development in the cat,” in The domestic cat: the biology of its behaviour (Cambridge University Press, Cambridge, U.K), 9–22.

Google Scholar

McGuire M. C., Johnson-Ulrich Z., Robeson A., Zeigler-Hill V., and Vonk J. (2018). I say thee “neigh”: Rescued equids are optimistic in a judgment bias test. J. Veterinary Behav. 25, 85–91. doi: 10.1016/j.jveb.2018.03.009

Crossref Full Text | Google Scholar

Mendl M., Brooks J., Basse C., Burman O., Paul E., Blackwell E., et al. (2010a). Dogs showing separation-related behaviour exhibit a ‘pessimistic’ cognitive bias. Curr. Biol. 20, pR839–pR840. doi: 10.1016/j.cub.2010.08.030

PubMed Abstract | Crossref Full Text | Google Scholar

Mendl M., Burman O. H. P., Parker R. M. A., and Paul E. S. (2009). Cognitive bias as an indicator of animal emotion and welfare: Emerging evidence and underlying mechanisms. Appl. Anim. Behav. Sci. 118, 161–181. doi: 10.1016/j.applanim.2009.02.023

Crossref Full Text | Google Scholar

Mendl M., Burman O. H. P., and Paul E. S. (2010b). An integrative and functional framework for the study of animal emotion and mood. Proc. R. Soc. B: Biol. Sci. 277, 2895–2904. doi: 10.1098/rspb.2010.0303

PubMed Abstract | Crossref Full Text | Google Scholar

Murray J. K., Skillings E., and Gruffydd-Jones T. J. (2008). Opinions of veterinarians about the age at which kittens should be neutered. Veterinary Rec. 163, 381–385. doi: 10.1136/vr.163.13.381

PubMed Abstract | Crossref Full Text | Google Scholar

Osthaus B., Lea S. E. G., and Slater A. M. (2003). Training influences problem-solving abilities in dogs (canis lupus familiaris). Proc. Br. Soc. Anim. Sci. 2003, 103–103. doi: 10.1017/S175275620001262X

Crossref Full Text | Google Scholar

Paul E. S., Harding E. J., and Mendl M. (2005). Measuring emotional processes in animals: the utility of a cognitive approach. Neurosci. Biobehav. Rev. 29, 469–491. doi: 10.1016/j.neubiorev.2005.01.002

PubMed Abstract | Crossref Full Text | Google Scholar

Paul E. S. and Mendl M. T. (2018). Animal emotion: Descriptive and prescriptive definitions and their implications for a comparative perspective. Appl. Anim. Behav. Sci. 205, 202–209. doi: 10.1016/j.applanim.2018.01.008

PubMed Abstract | Crossref Full Text | Google Scholar

Perdue B. (2017). Mechanisms underlying cognitive bias in nonhuman primates. AB&C 4, 105–118. doi: 10.12966/abc.08.02.2017

Crossref Full Text | Google Scholar

Piotti P., Szabó D., Bognár Z., Egerer A., Hulsbosch P., Carson R. S., et al. (2018). Effect of age on discrimination learning, reversal learning, and cognitive bias in family dogs. Learn Behav. 46, 537–553. doi: 10.3758/s13420-018-0357-7

PubMed Abstract | Crossref Full Text | Google Scholar

Pratsch L., Mohr N., Palme R., Rost J., Troxler J., and Arhant C. (2018). Carrier training cats reduces stress on transport to a veterinary practice. Appl. Anim. Behav. Sci. 206, 64–74. doi: 10.1016/j.applanim.2018.05.025

Crossref Full Text | Google Scholar

Richter S. H., Schick A., Hoyer C., Lankisch K., Gass P., and Vollmayr B. (2012). A glass full of optimism: Enrichment effects on cognitive bias in a rat model of depression. Cognit. Affect. Behav. Neurosci. 12, 527–542. doi: 10.3758/s13415-012-0101-2

PubMed Abstract | Crossref Full Text | Google Scholar

Sasaki Y., Gold J., and Watanabe T. (2010). Perceptual learning: cortical changes when cats learn a new trick. Curr. Biol. 20, R557–R558. doi: 10.1016/j.cub.2010.05.004

PubMed Abstract | Crossref Full Text | Google Scholar

Serpell J. A. (1988). “The domestication and history of the cat,” in The Domestic Cat: The Biology of its Behaviour. Eds. Turner D. C. and Bateson P. (Cambridge University Press, Cambridge, UK), 151–158.

Google Scholar

Tami G., Torre C., Compagnucci M., and Manteca X. (2011). Interpretation of ambiguous spatial stimuli in cats. Anim. welf. 20, 185–189. doi: 10.1017/S0962728600002669

Crossref Full Text | Google Scholar

Truax J. and Vonk J. (2025). Assessing judgment bias in ambassador animals: two case studies. Int. J. Comp. Psychol. 38, 1-20. doi: 10.46867/ijcp.35431

Crossref Full Text | Google Scholar

Turner D. C. (2000). “The human-cat relationship,” in The domestic Cat: The Biology of its Behaviour. Eds. Turner D. and Bateson P. P. G. (Cambridge University Press, Cambridge), 195–208.

Google Scholar

Uccheddu S., Miklósi Á., Gintner S., and Gácsi M. (2022). Comparing pears to apples: unlike dogs, cats need habituation before lab tests. Animals 12, 3046. doi: 10.3390/ani12213046

PubMed Abstract | Crossref Full Text | Google Scholar

Vieira de Castro A. C., Fuchs D., Morello G. M., Pastur S., de Sousa L., and Olsson I. A. S. (2020). Does training method matter? Evidence for the negative impact of aversive-based methods on companion dog welfare. PloS One 15, e0225023. doi: 10.1371/journal.pone.0225023

PubMed Abstract | Crossref Full Text | Google Scholar

Vitale K. R., Behnke A. C., and Udell M. A. R. (2019). Supplementary information: supplementary experimental procedures for attachment bonds between domestic cats and humans. Curr. Biol. 29, 18PR864–R865. doi: 10.1016/j.cub.2019.08.036/attachment/015e0f76-237d-425b-a7a2-922026be393f/mmc1.pdf

Crossref Full Text | Google Scholar

Vitale K. R., Takagi S., Arahori M., Chijiiwa H., Udell M. A. R., and Kuroshima H. (2024). A cross-cultural comparison of cat–human relationships in the United States and Japan. Anthrozoös 37, 813–831. doi: 10.1080/08927936.2024.2395127

Crossref Full Text | Google Scholar

Vitale Shreve K. R. V., Mehrkam L. R., and Udell M. A. (2017). Social interaction, food, scent or toys? A formal assessment of domestic pet and shelter cat (Felis silvestris catus) preferences. Behav. processes 141, 322–328. doi: 10.1016/j.beproc.2017.03.016

PubMed Abstract | Crossref Full Text | Google Scholar

Webster M. M. and Rutz C. (2020). How STRANGE are your study animals? Nature 582, 337–340. doi: 10.1038/d41586-020-01751-5

PubMed Abstract | Crossref Full Text | Google Scholar