Our main goal in this study was to gather information about which characteristics people find important when choosing a dog and investigate the possible influence of the owner's family composure, previous experiences with dogs and the purpose of acquiring the dog on this decision. One of the largest difference between our study and previous research was the methodology. Contrary to previous studies which used close-ended questions with predefined answer options, we used oped-ended question with a free-text answer. This difference lead to certain differences in the descriptive statictics. For example, in our study, less than 1% of the owners indicated that the dog's guarding ability was an important factor in their decision. Compared to that, in the study of Cohen and Todd [21], where prospective owners were asked to indicate whether or not they had a preference on 13 predefined behavioural and physical traits for dogs, 66% of the prospective owners indicated a preference for protection. This implies a differences in what people think they should consider important when inquired, and what people actually self-report spontaneously. The larger pool of answer options, characteristic of the open-ended question type over the close-ended question, was also evident from our dataset, where only 16 from the initial pool of 23 traits were mentioned by at least 5% of the responders (the minimum criterion of eligibility we set for the analyses). On the other hand, open-ended questions have their own disadvantages aside from a more complicated and time-consuming data-cleaning procedure, and chief among them is unintentionally leaving out traits that are intuitively obvious to the respondent. This bias can be behind the extremely low (< 2%) frequency of basic characteristics of the dog, like sex, age, and health status. Another difference between our method and those of previous studies was that in our research, we asked the owners after they acquired the dog why they chose it. Although we expected the owners to be able to recall the factors that had influenced their decision to choose that dog somewhat accurately, it is likely that they would recall only the most important, most decisive factors or those that are still somewhat relevant. At the same time, the owners may not have listed certain originally important factors because, although they influenced their decision at the time of selection, they lost their importance later on. This could be behind the fact that less then ¼ of our owners indicated any type of appearance traits, contrary to the much higher rate found in other studies [42, 23, 43]. Interestingly, however, some descriptives were similar to other studies, like the 25% ratio of owners choosing their dogs on a whim [20]. Nevertheless, the methodological differences mentioned above, coupled with the possible differences between the owner population sampled (i.e., many of the above-cited studies investigated only owners who wanted to adopt dogs from shelters), could be at least partly responsible for the rather large discrepancy in the ratio of showing preference for different traits between our study and other studies.
One additional advantage of our methodology was that since we asked the owners post-hoc about their dog choice, this allowed us to investigate choice-consistency over time and across dogs. To this end we analysed double entries, that is, answers where an owner filled out the questionnaire for the same dog twice. Of the 16 categories that were mentioned in at least 5% of the answers, 11 were at least fairly consistent. Among the not repeatable were the least specific and the least frequent (but still eligible) categories, like general looks, general character, and docility. Their low consistency could be explained by owners giving more specific answers on subsequent occasions. Moreover, it is possible that their low prevalence caused the low consistency in the case of the raw categories. Our results also showed that the owners' preferences in dogs showed consistency across dogs, at least for 8 out of the 11 consistent characteristics. Traits related to appearance and behaviour were consistent, while reasons related to the origin of the dog were not. These results suggest that owners do have a general preference for the behavioural and physical characteristics of their (multiple) dogs, and so had chosen their dogs on the basis of similar characteristics. This seems to contradict the idea that each dog is chosen by the owner on the basis of different characteristics because they play a different role in the owner's life [14], but supports the notion that the psychological mechanisms behind dog choice resemble those of preferential partnership [39].
Past studies have indicated that the preferences of individuals aspiring to become dog owners may differ depending on their familial circumstances [21]. Our results showed that the family composure indeed affected the likelihood of mentioning the majority of traits we obtained in the questionnaire. The number of adults in the household was associated with a preference for trainability, friendliness and choosing the dog on a whim. Owners living alone were more likely than other non-single owners to adopt a dog on a whim, without any self-reported expectations or reasons, similar to previous results [44]. Adopting a dog because one likes it can lead to conflict if the animal only partially meets the expectations of several people living together [45]. Of course, it is also possible that because adults living alone may be more lonely [46], their efforts to eliminate loneliness may override most of their preferences more easily, such that the choice can follow from a few or even a single criterion. Our results also showed that dog owners who live in multi-person households were more likely to mention any sport or work feature of the animal, in particular, a smart, easily trainable dog. They also more often wanted a friendly family dog. Trainability and friendliness are both characteristics that would enable the dog to adapt and conform to the needs and expectations of a bigger group [45, 47]. These traits, as highlighted in previous studies [22, 24], often prioritize characteristics over impulsive behaviors of the dog among potential dog owners. Nonetheless, in households with multiple adults, the harmonious coexistence of all members may take precedence over achieving exceptional sporting performance [45].
From all owner characteristics we investigated the most associations were found with whether there is a child in the household or not. In our sample, owners with a child were less likely to take in their dog on a whim, which fits well with the explanations provided above about the same category for multi-adult households. Owners with children were also less likely to adopt a shelter or rescue dog or choose a dog out of pity. Pets adopted from rescue organisations or found on the street often have unknown ancestry, so their possible hereditary diseases are not known. However, there can be many problems with adoption - dogs can bring with them trauma from a previous experience [48, 49], which can cause problems in a larger family later on [50, 51]. With young children, the potential danger to them - for example, a resource guarding dog [52] - can be a concern that parents with children may want to avoid - but in this case, the age of the children can be a determining factor for possible conflicts [53]. However, it is worth noting that animals that come from the wrong breeder can also present problematic and unexpected behavioural issues for owners due to premature weaning, inappropriate socialisation or inappropriate genetic background [54]. Nevertheless, it is conceivable that the dog's predictable temperament is more important for owners with a child. Parents also more often mentioned that they were looking for a dog with 'family dog' traits, especially getting a friendly, easily manageable, calm dog compared to childless owners, similar to previous findings [35, 36]. Although we did not ask whether the dog or the child came first in the family, it is possible that, even when choosing a dog, those without children were already making sure that they were not ruling out the possibility of having children, so they chose a dog accordingly. Since most dog attacks are suffered by children in the family [55], dogs can also play a very important role in a child's life [56] it may be important for a family to ensure that the four-legged friend around the child is as tolerant as possible of any extreme or abusive behaviour by the child [35, 36, 57]. Parents were also more likely to highlight the ease with which the dog can be trained, which, in a household with a child and a fast-paced dynamic, may facilitate training and adaptation to the family. Regular training increases the animal’s manageability and reduces the likelihood of potential behaviour problems, which can lead to easier coexistence [58]. Interestingly, having a child in the household did not influence the likelihood of mentioning any type of appearance traits, including size, even though one might have expected that smaller sized dogs -which are becoming more popular every year- would be preferable for parents because they present lower risk to a child and they are easy to physically control [30, 35, 36]. On the other hand, according to dog breed stereotypes, small dogs are typically considered overly energetic and nervous [59], while large dogs are associated with being good “nannies” (i.e. calm and child-friendly). This discrepancy may have discouraged owners from having a preference towards a specific size category.
Depending on whether a household had a dog when the newest one arrived, different preferences were found when analysing the questionnaire. Owners who had no dogs in the household were more likely to choose the dog on a whim, which, again, agrees with the general explanation that owners who live alone have no need to consider the needs and preferences of others in the household, including another dog. Prospective owners who didn't have a dog before adoption were inclined to express a preference for traits falling under the Composite friendly/manageable category. Interestingly, this preference did not specifically emphasize the narrow friendly/family dog category. This nuanced distinction suggests that their inclination leaned more towards a desire for calmness and manageability. This preference contrasts with prior research indicating the paramount importance to dog owners of having pets that are both easy to handle and friendly with both people and animals [35, 36]. Additionally, these prospective owners were more likely to seek a dog with easy trainability. It is possible that the link between the number of dogs in the household and easy manageability is more indirect and mediated by other lifestyle characteristics of the owner [60]. For example, people who generally have less time for the dog, cannot invest too much effort in dog training, or do not consider themselves particularly good in dog training may prefer to have only one, easily manageable dog at a time. A recently acquired dog characterized by a more amiable and easygoing temperament or displaying reduced assertiveness, as noted by Wallis et al. [61], may find smoother integration into a new household. In contrast, owners who already had a dog at home were more likely to consider the sport or work potential of the next dog. With an increasing number of owners embracing practical roles, such as involving their pets in sports activities as a shared hobby [62], there's a possibility that an owner initiated such pursuits with their initial dog and subsequently sought a new canine companion specifically tailored for those activities. Sporting and working dogs often exhibit higher energy levels and independence compared to their regular pet counterparts [63], demanding more commitment from the owner. This may not make them an ideal choice for a first dog, but in households with multiple dogs, they can engage in activities together, helping to expend their energy [64]. It's noteworthy that individuals with working and sporting dogs may prioritize companions with a 'good drive' [40], although this trait can pose challenges to harmonious coexistence.
Previous research has shown that one's previous experiences with partner choice can influence the partner one chooses [65]. Since certain processes in choosing a dog are similar to those in choosing a mate [21, 39], we wanted to find out whether the presence or absence of previous dog-keeping experience influences the prospective owner when choosing a dog. We found that owners who chose their first dog were more likely to consider the appearance, especially the size of the animal than owners with more dog-keeping experience. The importance of external characteristics when choosing a dog can be paramount for owners choosing their first dog, as they may not know exactly what behavioural traits to look for [66]. The lack of experience may be replaced by ideas gained from social media or from various movies, as can be seen in the emerging trends favouring a particular breed of dog [67]. However, the importance of appearance factors often differs depending on the country [35, 36]. It's plausible that the emphasis on appearance, particularly size, in dog selection is not predominantly driven by inexperienced owners. Instead, these features may have diminished in significance for more seasoned owners, with the latter group possibly regarding size as an inherent and obvious characteristic not warranting specific mention. Considering that past experiences as a dog owner can shape the traits individuals seek or aim to avoid in future dogs [41], it's conceivable that those with prior dog ownership experience might have contemplated qualities as significant in choosing a new dog that hadn't previously crossed their minds. When it comes to owners who already had dog(s) in the past, we found that they were more likely to mention any sport or training-related traits. More experienced owners were also more likely to mention their previous, direct experience with the breed or type of the dog, similar to Menchetti's findings [68].
Aside from other household members and the previous experiences of the owner, the purpose of keeping the dog was also related to what traits the owners consider when selecting their dog. Previous findings showed owners who do sports with their dogs were more confident and conscious about the behaviours they were looking for [69]. Our results showed that owners who kept their dogs for sport or work were more likely to mention any sport/work or training-related characteristics (Composite sport), including both the work/sport skills and smart/trainable category. These relationships do not require additional detailed explanation. Interestingly, a distinct correlation between keeping a dog solely for companionship and the qualities that define a good companion dog, such as a friendly and easy-to-handle nature, was not readily evident. While the acquisition of knowledge prior to bringing a dog into one's life can significantly enhance the long-term ownership experience [45], studies suggest that individuals often invest relatively little time in this critical aspect of preparation [70, 71]. It is conceivable that individuals seeking a pet without a specific working or sporting purpose may not have stringent criteria in mind beforehand. Another plausible explanation could be that, besides their practical roles, sporting/working dogs are also valued for companionship, making these aforementioned characteristics equally vital in their case.
The other association we found with regard to the purpose of the dog-keeping was with the origin of the dog. In particular, adopting a shelter or rescue dog or taking in the dog out of pity was more likely mentioned by owners who keep their dogs only for companionship. As we already mentioned, dogs found on the street or adopted from shelters have less predictable behaviour and, in the case of adopted puppies, also less predictable adult appearance in terms of body size and shape. When adopting from shelters, individuals frequently gravitate towards dogs that exhibit joy in playing or prefer lying in close proximity [22]. In contrast, a dog displaying intense jumping behaviors may be perceived as less appealing than a quiet and calm companion, aligning with the preferences outlined by Wells & Hepper [24]. This preference may persist, even though the more restrained dogs might lack the heightened 'drive' often sought after by owners of sporting dogs [40]. Moreover, higher-level competitions are often only open to dogs with pedigrees and most shelter and rescue dogs are mixed-breeds or have unknown ancestry, so adopted dogs could achieve success only at hobby sporting events. All these factors could discourage owners who wish to participate in competitive sports or work to adopt rescue dogs. It is also possible that the owner may not want to spend as much for a dog to be kept solely for their companionship as for a high-end sporting dog.
The use of the free-text method and post-hoc data collection pose some limitations. The results are not easily comparable with those obtained from close-ended questionnaires used in other studies. Additionally, it is worth mentioning that, like many other studies, our investigation was also conducted on a convenience sample, and as such, it likely skews towards urban owners who are more interested in dogs and dog ownership while probably under-representing owners who are less satisfied with their dogs or those who do not keep their dogs as companions.
Several exciting research questions remain for future work. Although many dog breeds were involved in our research, it would be interesting to see if we can see any differences in the awareness of owners in terms of breeds. For example, it would be interesting to see which breeds were more likely chosen for which specific characteristics (e.g., smartness or friendliness), and whether these associations are based solely on typical breed stereotypes or would also hold up to scientific evidence. In addition to the dog characteristics, it could also be worth looking at whether there are any characteristics of the owners (besides those we have investigated in the current study) that may play a role in the selection of the desired animal. Previous studies have suggested some good candidates in this direction, including the gender and income of the owner [72, 73], although other studies found no significant relationship between the source of the dog and the income category [74]. Finally, previous studies have suggested that assortative mating and similarity in personality is an important factor in the dog-owner relationship [14, 39]. Thus, an additional direction for future studies could be investigating the owner's psychological characteristics, like personality and attachment style, and linking them to their dog-choosing preferences to investigate if owners indeed prefer to choose dogs that are similar to themselves.