6. Review

With regard to the changes in mood before and after conversation, while the male students reported higher feelings at the beginning than female students, their feelings dropped even below that of the females at the end. Given that more males have a real pet in the real world, the drop in mood reported by the males may be due to the gap between the high expectations of conversation with a virtual pet and the real responses from it.

With regard to the relationship between the frequency of conversation and the attachment to a virtual pet, the high frequency conversation group (6 times or more) indicated a stronger favorability rating, while the low frequency conversation group (5 times or less) indicated a weaker rating. It is likely that the frequent conversation increased with the degree of attachment and inversely, that the high frequency resulted from the attachment to the pet. In terms of the relation between the frequency of conversation and the degree of comfort from pets, the high frequency conversation group rated higher scores for 10 items of the 11 items on the Virtual Pet Comfort Scale in Chesney & Lawson (2008) and Lawson & Chesney (2008), indicating their high comfort level with virtual pets. (The 10 items are ? "My pet is a companion", "Keeping a pet gives me something to care for", "My pet makes me experience pleasant activities", "My pet is the source of stability in my life", "My pet makes me feel needed", "My pet gives me pleasure and laughter", "Touching my pet comforts me", "I am happy to see my pet", "My pet makes me feel loved" and "My pet makes me feel trusted".) This result does not show whether the comfort level with virtual pets increased the conversation frequency or vice versa, however, it can be presumed that more people will have virtual pets based on the result that the more frequently they talk with a virtual pet, the higher the comfort attained, although there are not many who have a virtual pet at present.

As Kidd (1995) argues that the type of real pets, whether a dog or a cat, will not make much difference and that the same effects will occur as long as the pet-owner has a favorite pet. Virtual pets, however, differ much more in function than real ones do. According to the results of the survey regarding impression of virtual pets by function, no significant differences were found among the mouse-driven type, the type randomly reading blog entries, the type using memorized dialogues. However, the virtual pet that talks and plays with an avatar was rated lower than these above for "is likable", "elicits attachment", and "gives comfort". The chief cause is presumably that high-spec computers and high-speed lines are required for avatars and pets to smoothly chat and play. Under incomplete conditions, chatting and playing between avatars and pets become awkward, which is probably a cause of the low rating. With regards to expectations of virtual pets, this might have also caused the result that 41 percent do not expect the virtual pet to function with an avatar.

On the other hand, the result that 85% expect a virtual pet to have a conversational function suggests a high expectation for a virtual pet to be a conversation partner. Due to this, it is likely that an uncomfortable conversation with the virtual pet will lead to a lower mood, while a comfortable one to a higher mood.

7. Conclusion and future prospects

With the increase of elderly people living alone and the youth feeling lonely due to remote interpersonal relationships, pets are becoming an increasing presence in our lives. We conducted this investigation about the possibility of having virtual pets on the assumption that the virtual pet will become increasingly important for people in future given that keeping a real pet will become more restrictive and our society will become more computerized. The results showed that the high-frequency conversation group was more favorable toward pets and received stronger comfort from pets. The inverse was found to be true for the low-frequency groups. Furthermore, the result that 85 percent expect a virtual pet to have a conversational function shows they expect a virtual pet to become their conversation partner. It is also assumed that due to high expectations for a virtual pet to have a conversational function, uncomfortable conversation with a virtual pet might have led to a lower mood. Accordingly, virtual pets that converse with more fluently would be expected in the future.