Robot companion cats for people at home with dementia: A qualitative case study on companotics

Methodological considerations

This is a small-scale study that seeks to inform and enlighten rather than to discover a ‘truth’ as might be more the case in a positivist approach (Denzin & Lincoln, 1998). Therefore, an interpretive turn to the data analysis has been taken, supported by verbatim quotes from participants to demonstrate rigour. Interpretation will always be limited by the horizons of the interpreter, or as Heidegger (1962) referred to it, ‘situatedness’ or ‘forestructure of understanding’. The forestructure of understanding presented here is of the participants in their own context. The data are interpreted by a Nursing Academic with knowledge and experience in the field, therefore interpreted through a nursing lens (Peacock, 2001). As such, the reader will be able to judge for themselves the trustworthiness (Lincoln & Guba, 1985) of the results as underpinned by a shared understanding of the context within which the interpretation has taken place and is presented.

Ethics and enrolment of participants

Ethical approval for this research was applied for and granted by the Local Ethics Committee of the University. Subsequently, an advertisement was placed in the local media asking people to apply to take part in the study. They could either self-refer or refer their loved ones to take part. Inclusion criteria were that they were considered by themselves or their family to be living with dementia symptoms, that they were living at home and that they were cared for by family, friends and/or carers. There was no stipulation of age or gender but they needed to be fairly local to the geographical area of the University in order that the researchers could visit readily. Where the cat was accepted, it was not removed following the study since the participants may have connected with it (as will be seen below) and it may have been distressing to remove it. This was considered more important than the potential of study bias, due to the cat possibly being regarded as a ‘gift’.

Participants were visited by a researcher with clinical knowledge and experience of nursing people living with dementia. A dementia assessment tool was not administered as this research is not of a clinical nature, since the phenomenon of concern here is the interaction with the cat and its effect on behaviour rather than the medical condition itself.

Participants were enrolled to the study if they wished to take part, and if either they or their carers were able to converse with the researcher. All participants were able to give their consent, or their families were able to do so legally on their behalf. Where participants rejected the cat outright at the first meeting with the researcher, the researcher did not return to undertake an interview, since the purpose of this interview was to discuss the impact of the cat.

At the first meeting, following introductions and initial discussion of the research, the researcher demonstrated the cat to the participants and their families and research information sheets were given. With their permission, the researcher then gave the cat to the participant to hold should they wish to, or placed it by their side if that was preferable. Written consent to take part was not taken until all introductions had been made, and participants fully understood the aim of the research. Following this, instructions in use of the cat were given, i.e. how to switch it on and off and where the batteries were inserted, and details left of who to call should there be any concerns or questions.

Profile of the participants

The age of the participants was not recorded, since the researchers were only interested in the impact of the robot cats for people with dementia. However, all but one of the 12 participants were female and all were above retirement age. Of the four participants whose family told us they did not have a diagnosis, but appeared to have early symptoms of dementia, three were female and one male. These participants all rejected the cat, but each for different reasons. Either they did not like cats, they did not think it would help them, they did not like the robot cat or were resistant to any interventions. Other than those participants who rejected the cat, three participants were lost to the study, since two died and the other was admitted to the hospital, and the family did not re-establish contact with the research team subsequent to their admission. Table 1 displays the participant group profiles.

OPEN IN VIEWER

Table 1.

Participant profiles.

Participant No.	Gender	Accept (A)Reject (R) Lost to study (L)	Reason for rejection(paraphrased)
1	M	R	Does not like cats
2	F	R	Prefers her teddies
3	F	R	Not interested
4	F	R	Resistant to support
5	F	A	–
6	F	A	–
7	F	A	–
8	F	AL	–
9	F	L	–
10	F	A	–
11	F	A	–
12	F	L	–

In summary, participants 1, 2, 3, and 4 rejected the cat; participants 9 and 12 were lost to the study before they could contribute, due to death; participants 5, 6, 7, 10, and 11 contributed fully to the study; and participant 8 contributed to the first interview of the study, but was lost subsequently, due to hospital admission.

Table 2 displays the contributing participants’ living circumstances, as well as the key contact(s) who participated in the interviews.

OPEN IN VIEWER

Table 2.

Contributing participant circumstances.

Participant No.	Living circumstances	Professional care support	Key contact/interviewee(s)
5	Lives with husband	None	Husband
6	Lives in sheltered accommodation	Paid carers	Daughters/grand-daughters
7	Lives with daughter	None	Daughter and son-in-law
8	Lives in sheltered accommodation	None	Daughter
10	Lives alone	None	Daughter
11	Lives in sheltered accommodation	None	Husband and family

Ageless innovation® robot cat

The companotic device used for this research was a commercially available companion cat (Figure 1). It was originally developed by the US toy manufacturer Hasbro, but since 2018 is produced by a specialist companion pet company named ‘Ageless Innovation’.

OPEN IN VIEWER

Figure 1.

The ‘Ageless Innovation’ companion cat.

This product is marketed specifically for older people, and ‘… with a focus on improving quality of life for aging loved ones, their families, and caregivers’ (Ageless Innovation, 2019). The cat vocalises with meows and purrs in reaction to light and touch, via its on-board sensors. It moves its mouth, eyes, head and also rolls over, again in response to human interaction. Strictly speaking, it is not a robotic device, as in its off-the-shelf form, it is not programmable. However, it operates in two modes: mute or sound. In both modes, the cat’s movements continue to function. It is battery-powered, and the researchers have tested that the replaceable battery life is long (months, not weeks).

A fundamental criterion for accessibility and inclusion, especially given the demographics for this user population, is affordability. Other companion robots are much more expensive, and this is one of the main reasons we selected this device for our research study, as at the time of writing, it is priced in the region of £100 GBP. Although the cat is relatively simplistic in its operation, we postulated that its functionality would be enough for the purposes of this research, and for the engagement of our participants. This assumption was confirmed in preliminary pilot analysis.

Theme: Acceptance and rejection

This first theme brings together the discussions around acceptance of the cat, presenting the interpretation of the results supported by direct quotes from the participants discussing this. Interestingly, there was no guarantee when presenting the robot cat to a person with no diagnosis but with apparent symptoms of dementia, that it would be accepted. Asking screening questions, such as ‘Do you like cats?’ did not entirely predict the likelihood of the person engaging with the cat unless they strongly disliked cats, in which case rejection was immediate on this basis (participant one, male). People who did not dislike cats were more amenable to accepting it. However, there was also the possibility of an immediate dislike of the fact that the cat looked so realistic, which could possibly be attributed to the well-documented ‘Uncanny Valley’ phenomenon (Mori et al., 2012). This is discussed in more detail later in this paper. Even with the cat ‘turned off’, the rejection by one participant on the grounds of its ‘spooky’ appearance meant that the cat had to be put away in its box and the box hidden out of sight. In this case, the participant noted that she would much rather ‘stick to her teddies thank you’ as they were ‘much more of a comfort’ (participant two, female: research diary notes). Even though it was not possible to enrol this participant to the study, and no interview was able to be carried out, the data gleaned here were powerful, as the disgust registered on participant two’s face, and the negative body language (turning away from the cat and eyeing it with distrust) demonstrated to the researcher that immediate rejection of the cat was possible on this basis. This had not been previously considered, but the reaction was so strong, that the researcher did not feel she could leave the cat with participant two to try, for fear of upsetting her.

Participant three also rejected the cat, asserting that she did not think it would help her. She lived independently, and had very mild dementia symptoms, as reported by her daughter. On the day of the first meeting, during which the research was explained, the cat shown, and an appointment made to return for interview, her daughter was present. There was no immediate rejection, and she was happy to ‘try it’, but on returning on the appointed interview date, the participant met the researcher at her front door saying, ‘I don’t think it will help me’ (participant three: taken from research notes). On this basis, and because she said, ‘No, I’ve not switched it on’, no interview was possible.

At this point, what seemed to be predictive of rejection of the cat, therefore, was a dislike of cats, finding the cat ‘spooky’ and whether there was usually resistance to interventions or offers of support. Participant four’s family asked the researcher not to mention the ‘D word’ (dementia) in their mum’s earshot, as she did not accept the diagnosis. During the discussion, the family thought that there was a link between this and the acceptance of any interventions or help. Participant four’s family noted how she did not want the cat, discussing:

She rejected it, it never really worked for her. The daughter (Granddaughter of participant four) has a dog which she brings and she sits there and strokes it. That’s why I thought it would work for her. (Daughter four A)

Of the possible reasons for accepting the cat, those observed here were being a cat lover, and finding the cat ‘cute’ or ‘pretty’.

Theme: Distraction

Some participants, ‘enjoyed having something there’ (participant five) whatever dementia symptoms they displayed. Although participant five knew that the cat was not real, she enjoyed its presence and its actions. While this lady did not want the cat on her lap, and preferred the cat not to ‘meow’ too much, she had it close to her on the chair and stroked it often during the interview. Her husband also was pleased, saying the cat was always by his wife and that she seemed to enjoy its ‘company’ (husband of participant five).

Participant five noted:

It’s calmed me down a bit, yes. I used to get ‘het up’. The purr helps calm me down a little bit.

Participant eight’s daughter asked for a cat for her mother as she felt it would help her. This lady had unpredictable behaviour, and her daughter thought the cat could be a distraction from that behaviour and might help to stabilise her moods. This lady accepted the cat immediately, initially knowing it was a robot cat as the researcher had demonstrated where to put the batteries, but she seemed to forget that straight away, stroked it and talked to it as though it were real. She was very responsive to the cat. It was unfortunate that this lady was admitted to hospital and therefore the short initial interview was the only one that was carried out. During this interview, other than discussing how much she liked the cat, it was not possible to determine the impact of the cat on this lady’s behaviour, as it was very early stages in the process for her. Potentially, the planned later interview may have elicited further information regarding any changes in patterns of her behaviour.

Theme: Communication

For some participants, the cat became a subject of conversation between the participants and their families and carers.

Husband 11 said:

Well, we can have a bit of fun with it when she’s (the cat’s) with us, especially with the carers. They think it’s real, and then when it moves they jump!

In this way, communication was stimulated, and families particularly noted this where communication had previously decreased. This appeared to be partly a function of the novelty of the cat, but was protracted, lasting at least three months (between the first and the second interviews). The impact on Family 11 was striking. This lady had quite severe dementia symptoms and had almost totally withdrawn from her husband, family and carers. She needed help with all aspects of care, including prompting to eat and to drink. She was displaying repetitive behaviour, calling out during the evenings, which fluctuated with periods of silence. Her family found it very difficult to reach her, and could not ascertain her wishes when asked simple questions, such as ‘would you like a drink?’. Participant 11 was less withdrawn when the cat was beside her and would respond in crooning tones to the cat when it meowed and would focus on it when it moved. The cat was also ‘good when she’s shouting’. Granddaughter 11 said:

I try to get Grandad to bring the cat out when Gran is shouting, to see if it will help. It’s a bit worrying for me, because I can see Grandad getting really fed up and tired, especially in the evenings. I worry that Gran’s shouting gets him down. Yes, the cat does help a bit, for a short time, but I think Gran’s at such a stage that it’s difficult for anyone or anything to reach her. It’s funny though, that the cat reaches her where we can’t always. You wouldn’t think it would, would you?

Theme: Connecting with the cat and connecting with others

Some participants and families found that there was a real stimulation occurring that went beyond motivating conversation. One of the strongest reactions came from participant six, whose interview was the longest as her family were delighted with the increase in her engagement with them since having the cat. This lady had exhibited a recent deterioration of mood, functioning and communication. She had been referred to the researchers by the care staff in the supported living complex where she resided, and her family consented to their mum/grandma’s participation in the study. On visiting, the lady was said to be really fond of cats but would be unable to care for a real cat. She loved a toy cat that was in the communal space in the complex, and often spent time petting it. This may well have been a predictor of the immediate and strongly positive reaction on first meeting the robot cat. As the cat was brought into the room, she held out her hands to the cat saying, ‘Where have you been?’ addressing the cat, and took it into her arms and straight to her lap where she stroked it and spoke to it with affection.

The subsequent interview visit to participant six with her family present demonstrated a continuing strongly positive reaction to the cat. The most telling of the positive reactions came from Daughter six who, discussing the effect of the cat on her mother’s wellbeing, said it had ‘given her a new lease of life’ (Daughter six). Her granddaughter went further, stating that it had, ‘Given her a purpose’ (Granddaughter six). When asked to discuss this further, her granddaughter was emphatic, saying, ‘Oh yeah’, and her daughter said again, ‘Yeah, definitely. It’s like as if she’s got a purpose now’. This is a result beyond that which might have been predicted. This lady had a much deeper emotional response to the robot cat than simply being calmed or being stimulated to converse with others. Further discussion during the interview elicited that this lady’s clinical assessment prior to having the cat had led to advice that she may have to move to residential care. A later clinical assessment (after having had the cat for several weeks) had been undertaken, during which the clinician had noted a considerable improvement in her condition, and that there was no need to move to residential care and could stay in her current home. This delighted the family, and was, they felt, confirmation of the cat’s impact on their mother’s general wellbeing. The carers were able to confirm the positive impact of the cat (research notes) noting that she was more amenable to getting up in the morning and to sticking to mealtime routines than she had been, and they used the cat as prompts to get out of bed in the morning saying phrases such as ‘(The cat’s name) is waiting for you’. This was a prompt used to avoid resistive behaviour, and tended to work for participant six most of the time. One episode was upsetting for the family, when participant six was particularly unwell:

Granddaughter six: So we had to get the carers in, she was hallucinating, thinking she was in a taxi, in a hospital bed, and… just horrendous.

Interviewer: Terrible. (All murmur in agreement).

Granddaughter six: It was, it was, her mood changed instantly when she saw the cat.

Interviewer: Really?

Daughter six: Yeah, definitely.

Granddaughter six: We were like, ‘Thank you, (cat’s name)!’

(All laugh)

Interviewer: So, it was like, she saw (cat’s name), and her mood changed?

Granddaughter six: Yeah

Daughter six: Straight away.

Interviewer: Straight away?

Daughter six: Straight away.

Granddaughter six: You can’t really describe that, unless you’ve seen it through your own eyes,

Daughter six: Yeah

Granddaughter six: because it was just like…

Interviewer: A switch?

Granddaughter six: Yeah

Daughter six: Yeah, definitely!

The effect of the cat on mood and behaviour (communication and distraction)

The benefits of the robot cat were clear in terms of improving communication and providing a distraction from repetitive behaviour, and although this was a small-scale study, where they were accepted, the cats provided positive benefits for the participants and their families. Similar benefits for people with mental health problems have been found with companion pets, though research shows that there are practical and emotional burdens associated with pet ownership (Brooks et al., 2018). In our study, one daughter in particular discussed the difficulties her mother had experienced with worrying where her pet cat was at night, and how the robot cat had replaced the real cat very well. Some surprisingly strong changes to the participants were noted in terms of improvements and stability in behaviour, but further research will need to be undertaken in order to determine whether this was a sustained change over a longer period of time. Stability of mood and interruption of repetitive behaviour has been noted and this was of particular benefit to the family carers.

Carers of people living with dementia may become stressed (Svendsboe et al., 2018) and their quality of life can be adversely affected (Farina et al., 2017). Evidenced by the reaction of some participant’s families, the robot cat appears to have had a positive effect on the carers’ wellbeing, through its ability to modify mood and its calming effect. Though this was variable, the most positive benefits appear to have occurred for carers of those participants experiencing moderate to severe dementia symptoms, distracting from repetitive symptoms and providing an anchor to reality.

The benefits of the cat were also related to stimulating conversation between the family and their loved one. It appeared to bring the family together and to stimulate conversation, where it had been previously difficult. The strongest reaction again came from those participants who were experiencing moderate to severe symptoms.

Interaction with the cat (acceptance and rejection, and connecting)

Where participants accepted the cat, many of the participants responded to it as though it were real, stroking it and talking to it in response to its movements, purrs and meows. There was evidence of zoomorphism in that they spoke to the cat, thereby demonstrating a degree of suspension of disbelief that the cat was not alive. Furthermore, an emotional connection seemed to be made with the cat, as the participants appeared to talk lovingly to the cat, sometimes even when conversations could not be held with other people.

What became clear in our interviews and observations were that the cat’s level of interaction had a variable effect on the participants. Some accepted its repetitive meowing, others did not. Also, the cat’s tendency to roll over on its back occasionally caused distress for some, especially when this action was perceived as the cat falling from the participant’s arms or lap. In some cases, it was also important that the cat’s vocalisation sounds could be ‘turned off’. The meow appeared to some participants as the cat asking for something, and where the participant was unable to respond to the perceived need, some distress could be caused.

The cat’s appearance also had a variable affect. This might relate to the phenomenon of the ‘Uncanny Valley’, first proposed by Mori in the 1970s (Mori et al., 2012). This phenomenon considers how a person might react emotionally to objects as they increasingly resemble a human being. The ‘valley’ effect occurs when the object is very similar to a living human, but not identical. Mori hypothesised that there would be an emotional disturbance at that point, using a ‘zombie’ and a corpse as examples (Figure 2).

OPEN IN VIEWER

Figure 2.

The Uncanny Valley.

Whilst the Uncanny Valley focuses specifically on humans, it is clear that such a phenomenon can apply to other living beings in a zoomorphic as opposed to anthropomorphic scenario, and this has indeed been considered by researchers of social robotics (Dalibard et al., 2012).

Although the Uncanny Valley applies specifically to the visual senses, there is no reason why this could not be extended, and this has been proposed in the aural senses by Grimshaw (2009) in the context of computer games sounds. To our knowledge, it has not been researched in the domain of robotics, and this is an area in which we are planning to undertake further experimental studies, particularly in the light of some participants in our study preferring the cat with its sounds deactivated.