=Paper=
{{Paper
|id=Vol-3794/paper01
|storemode=property
|title=How do Lonely, Young Adults Perceive Interactive Technologies
With Varying Human-Likeness? An Experimental Lab Study
|pdfUrl=https://ceur-ws.org/Vol-3794/paper1.pdf
|volume=Vol-3794
|authors=Aike Horstmann,Jacqueline Boußard
|dblpUrl=https://dblp.org/rec/conf/rfh/HorstmannB24
}}
==How do Lonely, Young Adults Perceive Interactive Technologies
With Varying Human-Likeness? An Experimental Lab Study==
https://ceur-ws.org/Vol-3794/paper1.pdf
How do Lonely, Young Adults Perceive Interactive Technologies
With Varying Human-Likeness? An Experimental Lab Study
Aike C. Horstmann1,∗ , Jacqueline Boußard1
1
University of Duisburg-Essen, Forsthausweg 2, 47057 Duisburg, Germany
Abstract
For young adults in particular, it is proposed that interactive technologies can help alleviate loneliness. However, while state loneliness
was found to lead to a more positive evaluation of interactive technologies, general loneliness was found to have the opposite effect.
Since interactive technologies vary regarding their human-likeness in appearance and behavior, it needs to be investigated how this
affects the perception of young adults while considering potential effects of their loneliness. In an experimental lab study with a
2x2 between-subjects design, 101 participants aged 18-35 years interacted either with a social robot or a voice assistant which either
displayed a rather human-like or machine-like communication style. General and state loneliness were assessed alongside evaluations
of the interactive technology. Overall, the participants appeared to be more comfortable the more their interaction partner looked
and talked in a human-like manner. The self-reported state loneliness was very low and appeared to have no influence on evaluations.
General loneliness hints towards a trend towards a more negative evaluation of the interactive technology’s social attractiveness - a
finding that should be further investigated in future studies.
Keywords
human-machine interaction, human-likeness, social robots, voice assistants, loneliness, young adults,
1. Introduction vary regarding their human-likeness.
The intense social isolation and distancing measures that
we experienced during the COVID-19 pandemic are antici-
1.1. Perception of Interactive Technologies
pated to have negative consequences in terms of increased and Loneliness
loneliness [1]. Loneliness is defined as the perceived gap Numerous studies show that interactive technologies such
between desired and actual social relationships and asso- as robots, virtual agents, voice assistants, and chatbots trig-
ciated with mental health issues such as depression and ger social reactions in people as soon as they fulfill a few
physical health problems [2]. Young adults are particularly conditions (interactivity, natural language, fulfilling a so-
susceptible to loneliness, a vulnerability exacerbated by the cial role; [10, 11, 12, 13]). Since interactive technologies are
pandemic, which emphasizes the need to address this issue able to take over social roles and are frequently perceived
urgently [3, 1] . Considering that young adults are often and treated as social interaction partners, they are often
tech-savvy and engaged, interactive technologies that of- proposed to be used to satisfy social needs – particularly
fer features akin to a human interaction partner may be for individuals suffering from loneliness [4]. When experi-
used to alleviate acute loneliness and further train social encing a dissatisfying social need state (e.g., “I feel lonely”),
skills to diminish general loneliness. For instance, previ- individuals are motivated to take action to resolve it (e.g.,
ous studies explored the potential of using a social robot “I will talk to a friend”; [14]). However, if this action is not
to aid young adults facing loneliness as a companion or able to satisfy the need, the individual is still experiencing
a social skills coach [4, 5, 6] . However, there is evidence the need dissatisfaction and the motivation to take action to
suggesting that particularly individuals suffering from se- resolve it [9]. For instance, young adults often turn to social
vere loneliness perceive social technologies and their effects media platforms as a means of seeking connection and alle-
differently – mostly more negatively (e.g., social robots [5], viating feelings of isolation. Paradoxically, intensive social
video call technologies [7], social media [8]). This could media use was found to amplify their sense of loneliness
be because they do not experience the same level of social [15, 16]. This two-sided nature of the relationship between
need satisfaction by using these technologies as others do technology and loneliness underlines the complexity of its
[9]. Even worse, using these technologies and not receiving impact on young adults. The question arises what factors
the need satisfaction they are seeking, may even amplify determine whether interactive technologies can satisfy so-
their feelings of loneliness as their social need dissatisfac- cial needs of lonely young adults. With this aim in mind, it
tion becomes more salient. The perception of interactive needs to be further investigated how lonely young adults
technologies by lonely young adults is a multifaceted issue perceive interactive technologies in general and how this is
that requires a nuanced understanding of how these tech- affected by the appearance and behavior of the technologies.
nologies can do both, alleviate and exacerbate feelings of There is extensive research on how artificial entities’ appear-
loneliness. The current study therefore employs an experi- ance and behavior are generally perceived by their human
mental study design to deepen our understanding of what interaction partners (e.g., [17, 18, 19, 20, 21] . In this study,
role different types of loneliness (state and general) play however, the influences of current and general loneliness in
in how young adults perceive interactive technologies that young adults are specifically taken into account.
Workshop Robots for Humans 2024, Advanced Visual Interfaces, Arenzano, 1.1.1. State Loneliness
June 3rd, 2024
∗
Corresponding author. In situations where social needs are not satisfied and human
Envelope-Open aike.horstmann@uni-due.de (A. C. Horstmann); interaction partners are not available, people may turn to
jacqueline.boussard@stud.uni-due.de (J. Boußard) interactive technologies as an alternative strategy. The re-
Orcid 0000-0003-3412-1639 (A. C. Horstmann)
© 2024 Copyright for this paper by its authors. Use permitted under Creative Commons License sults of various studies suggest that social robots (e.g., Aibo,
Attribution 4.0 International (CC BY 4.0).
CEUR
ceur-ws.org
Workshop ISSN 1613-0073
Proceedings
Paro, Vector) are quite effective in reducing current feelings a positive impact from interacting with them, but only if
of loneliness [22]. Also voice assistants can be perceived their level of loneliness is not on a generally high level.
as companions that may improve social connectedness and Therefore, we assume that people who are more strongly
alleviate loneliness [23]. Previous research suggest that affected by general loneliness react differently to the in-
when feelings of loneliness are activated, individuals tend teractive technology’s human-likeness, resulting in a more
to anthropomorphize interactive technologies more strongly negative evaluation compared to people that are less af-
[24, 25] . For instance, they reported to feel a stronger social fected by general loneliness:
presence than other people while interacting with social H3: General loneliness has a negative effect on an inter-
robots [26]. This could be the case because socially dissatis- active technology’s a) perceived sociability, b) perceived
fied individuals are more sensitive to social cues [27]. An- competence, and c) overall evaluation.
other explanation is that Anthropomorphism helps to fulfill H4: The positive effect of an interactive technology’s
social needs by offering a humanlike interaction with non- human-likeness (in behavior and appearance) on their a)
human entity [28]. State loneliness further seems to result perceived sociability, b) perceived competence, and c) over-
in a more positive evaluation of a technological interaction all evaluation is diminished by general loneliness.
partner, for instance regarding its warmth, friendliness, and
sociability [25]. We therefore hypothesize:
H1: Individual’s state loneliness positively affects an interac- 2. Method
tive technology’s a) perceived sociability, b) perceived com-
An experimental lab study with a 2x2 between-subjects
petence, and c) overall evaluation. It is assumed that higher
design was conducted. The study was approved by the
human-likeness of interactive technologies leads to more
local ethics committee. Supplementary study material
positive effects in social settings. For instance, a robot’s
(data set, questionnaires, script) can be found online:
more human-like appearance led to a stronger perception
https://osf.io/nzs5v/.
of mind [8]. Human-likeness in appearance also increases
social conversation chances: participants were observed
to speak and respond more to a social robot than a voice 2.1. Sample
assistant. They further reported to feel more interpersonal A power analysis (conducted with the software G*Power;
warmth, to enjoy the conversation more, and to feel less .80 power, an effects size of f² = 0.15, and .05 alpha error
lonely with the social robot than the voice assistant [29]. probability) recommended a minimum of 55 respondents.
More human-like behavior – in terms of speaking styles – In total 105 participated in the study. Two datasets were
were also found to result in more positive evaluations of the excluded due to incompletion, one by request of the partic-
technological interaction partners, for instance regarding ipant, and one due to a suspicious answering pattern. Of
warmth and sociability [30] . Therefore, we hypothesize: the remaining 101 participants, 37 identified as male and 64
H2: More human-likeness of an interactive technology’s a) as female. Participants had to be at least 18 years old and
appearance (social robot vs. voice assistant) and b) behavior no older than 35 years. On average, participants were M =
(human-like vs. machine-like communication style) leads 23.12 (SD = 3.63) years old. With 88.1 %, most participants
to less state loneliness. were students. Accordingly, 70.3 % of participants stated to
have a university entrance qualification and 23.8 % to have
1.2. General Loneliness a university degree. Participants’ general enthusiasm for
technology was rather high (M = 3.55, SD = 0.56). Of the 50
As outlined before, previous research reports a link between
people in the social robot condition, only ten had person-
state loneliness and anthropomorphizing tendencies. In
ally interacted with a social robot before, 14 had observed
contrast to state loneliness which describes a temporary,
someone else interact, and 27 had seen a report about social
short-term experience of feeling alone that can be relieved
robots. Of the 51 in the voice assistant condition, 35 had
once the situational factors causing it are resolved, gen-
personally interacted with a voice assistant before, 42 had
eral or chronic loneliness is defined as a prolonged and
observed someone else interact, and 26 had seen a report
persistent state of feeling alone, even when surrounded by
about voice assistants.
others [31]. People with longer lasting loneliness, however,
were found to attribute less human traits to an interactive
technology (i.e., humble, broadminded, polite), which may 2.2. Procedure
discourage them from developing anthropomorphic infer- Following informed consent and reviewing the study mate-
ences (e.g., social response, warmth, competence) [32, 33]. rials, participants answered questions about their sociode-
In a different study, a robot that was proposed to alleviate mographic background and technical affinity on a laptop.
loneliness by functioning as companion or as social skill The experimenter then introduced either the voice assis-
coach was evaluated as less socially attractive the higher tant or the social robot and explained the interactive task
the raters’ self-reported general loneliness [5]. There have (Figure 1). A cover story was used claiming that the pur-
been similar observations by researchers in the context of pose of the study was to test an interactive technology for
computer-mediated communication: During the pandemic, everyday personal use and to improve their speech and com-
people reported to feel even more unsatisfied in terms of munication skills. The experimenter pretended to start the
social needs after interacting with others using video call interaction program by saying ”start interaction program”
technologies [7]. Also social media is frequently found to and left the experiment room, allegedly so that the partici-
have more negative effects the more lonely the users report pants would not feel observed during the interaction. From
to be [16]. Apparently, longer-lasting general loneliness an adjacent room, the experimenter was able to control
leads to more negative reactions towards social technolo- the voice assistant’s or robot’s outputs by using a webcam
gies. Thus, a currently lonely person may be inclined to that was installed in the lab to see and hear the participant
anthropomorphize interactive technologies and experience
Table 1
Regression Analysis Results With State Loneliness as Predictor
Criterion b SE B 𝛽 p 𝑅2 𝐹(1,99)
Sociability 0.14 0.10 .15 .138 .02 2.24
Competence -0.09 0.09 -.11 .283 .01 1.17
Overall eval. 0.01 0.11 .01 .951 .00 0.00
items; e.g., “I feel left out.” with the pre-face “The following
Figure 1: Study Setup With Voice Assistant Alexa (left) or Social questions relate to your feelings of loneliness at the present
Robot Nao (right). moment.”; 1 = “not at all” to 5 = “completely”;𝛼 = 0.70; M
= 1.66, SD = 0.70). General loneliness was assessed via the
short version of the Social Loneliness Scale ([38]; 5 items;
and letting the voice assistant or robot react accordingly e.g., “I do not have any friends who understand me, but I
(Wizard of Oz design; see [34]). The webcam was justified wish I did.”; 1 = “strongly disagree” to 7 = “strongly agree”;
by explaining that in case of errors the developers could 𝛼= 0.75; M = 2.71, SD = 0.78). 14 data sets had missing values
track what went wrong. During the interaction, the voice for the social loneliness scale and were excluded from the
assistant or robot displayed a rather human-like (expression analyses that involved this measure.
of emotions and intentions such as “this makes me happy”,
using terms from humans’ everyday life such as “I work…”) 2.3.3. Interactive Technology Evaluation
or machine-like (e.g., more functional, command-based lan-
The interactive technology’s perceived sociability and com-
guage and phrases attributing to the technical processing
petence were assessed via adjective pairs that were rated
such as ”your response has been saved and processed”) com-
on a 5-point semantical differential ( [39, 40] ; sociability: 6
munication style. First, the voice assistant or robot asked
items; e.g., “friendly – hostile”;𝛼 =0.77; M = 3.33, SD = 0.66;
the participant to provide background information about
competence: 6 items; e.g., “professional – amateur”; 𝛼=0.70;
themselves, including their name, a brief description of their
M = 3.26, SD = 0.58). For the overall evaluation, four items
profession, age, leisure activities, the reason why they are on
from Burgoon and Walther [41] were adapted (e.g., “I was
campus today and anything else they would like to disclose.
enjoying the interaction with the voice assistant Alexa/the
Next, the robot or voice assistant asked about wishes for the
social robot Nao.”;𝛼 = 0.77; M = 3.20, SD = 0.79).
future. Last, the participant was asked about happy, unpleas-
ant, and finally sad experiences. To facilitate self-disclosure
via reciprocity [35], the voice assistant or robot always dis- 2.3.4. Manipulation Checks
closed the desired information about itself first. Three dif- Participants were asked to rate the external appearance
ferent answers were prepared, depending on whether the and the communication style of the voice assistant or robot
test subject answered the question, did not want to answer (1 = “more machine-like” to 6 = “more human-like). Two
the question, or could not think of an experience. At the MANOVAS show that the experimental manipulations sig-
beginning and the end of the interaction, the participant nificantly predicted participants’ perceptions regarding the
was asked about their current mood. After the interaction, appearance, 𝐹 (1, 99) = 33.74, 𝑝 < .001, 𝜂2𝑝 = .25, but not
participants were sent back to the laptop for the second regarding the communication style, 𝐹 (1, 99) = 0.09, 𝑝 =
part of the questionnaire. They were asked to state their .771, 𝜂2𝑝 = .00.
current sense of loneliness, evaluate the interaction and
their interaction partner, and answer manipulation checks
as well as questions about their person. Finally, participants 3. Results
were debriefed and compensated (either course credits or 5
€). The interaction lasted about 10 minutes and the entire Statistical analyses were conducted with IBM SPSS Statistics
experiment about 45 minutes. 29 and the PROCESS macro v4.3, significance was deter-
mined using the standard 𝑝 < .05 criterium.
2.3. Measurements
3.1. State Loneliness (H1-H2)
2.3.1. Personal Background
To investigate H1 (state loneliness positively affects an in-
Participants reported their sociodemographic information teractive technology’s a) perceived sociability, b) perceived
(age, gender, education, occupation), their previous experi- competence, and c) overall evaluation), three linear regres-
ences with robots or voice assistants (frequency of personal sion analyses were conducted. State loneliness was al-
or observed contact, reception of reports; 0 = “never”; 1 = ways the predictor, the criterion was either the interac-
“very rarely” to 5 = “very often” [19]), and their technical tive technology’s perceived sociability, its perceived com-
affinity (TA-EG; [36]; 19 items; e.g., “I enjoy trying an elec- petence, or its overall evaluation. The results, presented
tronic device.”; 1 = “does not apply at all” to 5 = “applies in Table 1, show that state loneliness had no significant
completely”; 𝛼 = 0.76). effect on any of the interactive technology’s evaluation
measures. Consequently, the hypothesis H1 needs to be
2.3.2. Loneliness rejected. For H2 (more human-likeness of an interactive
technology’s a) appearance and b) behavior leads to less
Participants’ state loneliness was measured with the short
state loneliness), an ANOVA was conducted with type of
scale for measuring loneliness by Hughes et al. ([37]; 3
Table 2 Table 4
Regression Analyses With General Loneliness as Predictor Moderation Analyses With General Loneliness as Moderator and
Communication Style as Predictor
Criterion b SE B 𝛽 p 𝑅2 𝐹(1,85)
Criterion 𝑅2 F(3, 83) p Δ𝑅2 𝐹(1,83) p
Sociability -0.17 0.09 -.21 .057 .04 3.71
Competence -0.09 0.08 -.12 .261 .02 1.28 Sociability .15 5.49 .002 .00 0.10 .759
Overall eval. -0.17 0.11 -.16 .128 .03 2.36 Competence .03 0.76 .519 .02 1.13 .290
Overall eval. .07 1.82 .149 .00 0.30 .583
Table 3
Moderation Analyses With General Loneliness as Moderator and
Type of Technology as Predictor ogy, V = 0.00, F(3, 80) = 2.73, p = .049, 𝜂2𝑝 = .09, as well as
communication style, V = 0.15, F(3, 80) = 4.51, p = .006, 𝜂2𝑝
Criterion 𝑅2 F(3, 83) p Δ𝑅2 𝐹(1,83) p = .15, when controlling for general loneliness. For type of
Sociability .07 2.61 .057 .01 0.69 .408 technology, separate univariate ANOVAs on the outcome
Competence .03 0.54 .656 .01 0.27 .607 variables reveal a marginally significant effect on perceived
Overall eval. .06 1.80 .154 .01 0.21 .645 sociability, F(1, 82) = 3.26, p = .074,𝜂2𝑝 = .04, and overall eval-
uation, F(1, 82) = 3.26, p = .074, 𝜂2𝑝 = .04, but no significant
effect on perceived competence, F(1, 82) = 0.27, p = .608,
technology (social robot vs. voice assistant) and communi- 𝜂2𝑝 = .00. For communication style, the separate univariate
cation style (human-like vs. machine-like) as factors and ANOVAs reveal a significant effect on perceived sociability,
the participant’s state loneliness as criterion. There was F(1, 82) = 12.17, p < .001,𝜂2𝑝 = .13, and a marginally significant
no significant effect found, neither for type of technology, effect on overall evaluation, F(1, 82) = 3.76, p = .056,𝜂2𝑝 = .04,
𝐹 (1, 97) = 0.18, 𝑝 = .672, 𝜂2𝑝 = .00, nor for communication but again no significant effect on perceived competence, F(1,
style, 𝐹 (1, 97) = 0.18, 𝑝 = .670, 𝜂2𝑝 = .00. Therefore, the 82) = 0.22, p = .641,𝜂2𝑝 = .00. Summing up, the additional
assumptions of H2 are not supported. analyses regarding the main effects suggest that the more
human-like a technology appears and behaves, the more
3.2. General Loneliness (H3-H4) sociable and generally positive it is evaluated (see Table 5).
To investigate H3 (general loneliness has a negative effect
on an interactive technology’s a) perceived sociability, b) 4. Discussion
perceived competence, and c) overall evaluation), three re-
gression analyses were calculated. As can be seen from the The aim of the current study was to deepen our understand-
results presented in Table 2, no significant effect of general ing of the roles that state and general loneliness play in
loneliness could be found. It is noteworthy that there is a how young adults perceive interactive technologies that
marginally significant effect on sociability: Higher general vary regarding their human-likeness. For this purpose, an
loneliness appears to lead to a more negative evaluation of experimental lab study with a 2 (robot vs. voice assistant)
the interactive technology’s sociability. Hypothesis H3 is x 2 (human-like vs. machine-like communication style)
not supported. To test H4 (the positive effect of an inter- between-subjects design was conducted.
active technology’s human-likeness on their a) perceived
sociability, b) perceived competence, and c) overall evalu- 4.1. State Loneliness
ation is diminished by general loneliness), six moderation
analyses with general loneliness as moderator variable were Against our assumptions, the results suggest that state lone-
calculated using the PROCESS macro by Hayes [42]. Boot- liness has no effect on the interactive technology’s perceived
strapping with 5000 samples and heteroscedasticity consis- sociability, competence, or overall evaluation. However, it
tent standard errors (HC3) were employed. The first three needs to be noted that the state loneliness was generally very
incorporated type of technology as predictor, with either low indicating a floor effect. Previous studies that found
the interactive technology’s perceived sociability, its per- an effect of state loneliness on the perception of interactive
ceived competence, or its overall evaluation as criterions. technologies employed a setting where state loneliness was
The analyses did not show that general loneliness moderates intentionally primed [25]. Since we were also interested
the effect between the type of technology and its evaluation in whether the differences in human-likeness of an interac-
significantly (see Table 3 for statistic values). The other tive technology’s appearance and behavior influences state
three moderation analyses contained the technology’s com- loneliness, we chose to not influence their state loneliness
munication style as predictor and the same variables as via a priming task. However, we also did not find any sig-
criterions as the previous three moderation analyses. As nificant effect of the interactive technology’s appearance
the results presented in Table 4 show, general loneliness (social robot vs. voice assistant) or behavior (human-like vs.
also does not appear to moderate the effect between the machine-like communication style) on participants’ state
technology’s communication style and its evaluation signif- loneliness. Since state loneliness was low across all groups,
icantly. Consequently, H4 needs to be rejected. Following we assume that the interaction with the technology itself
recommendations by Hayes [42], the interaction term was had a positive effect regarding the currently experienced
disregarded to have a look at the main effects instead. There- loneliness. Although previous research found differences in
fore, a MANCOVA was calculated with type of technology perception in respect to an interactive technology’s human-
and communication style as factors, general loneliness as likeness, our results are very much in line with fundamental
covariate, and perceived sociability, perceived competence, media psychological research findings. According to the
as well as overall evaluation as criterions. Using Pillai’s media equation theory, minimal social cues elicit social re-
trace, there was a significant main effect of type of technol- actions [13]. If a technology is interactive, uses natural
Table 5
Descriptive Statistics of the Interactive Technology’s Perceived Sociability (S), Perceived Competence (C), and Overall Evaluation
(OE)
Voice Assistant Social Robot Machine-like Human-like
M SD M SD M SD M SD
S 3.20 0.64 3.47 0.65 3.11 0.62 3.58 0.61
C 3.27 0.56 3.24 0.61 3.21 0.62 3.30 0.54
OE 3.03 0.72 3.37 0.82 3.05 0.80 3.38 0.75
language, and fulfills a social role, this is sufficient for hu- To get a clearer picture of the influence of general loneliness,
mans to react to them socially. All three criteria were met future studies should consider recruiting two groups – one
in our study by both interactive technologies and in both with high and one with low levels of general loneliness.
communication style conditions. Therefore, the interaction
in all conditions might have been sufficient to bring all par-
ticipants’ state loneliness to a low level (cf. [5]). However, 5. Conclusion
considering the detrimental long-term effects of chronic
If a gap between desired and actual social interactions per-
loneliness [1, 2], the examination of general loneliness as
sists over a longer period, profound consequences on phys-
influencing factor for the perception of interactive technolo-
ical and mental health are likely [1, 2] . To address the
gies is particularly critical.
pervasive issue that is largely affecting young adults, it is
crucial to understand the nuances of loneliness, for instance
4.2. General Loneliness concerning the perception of potential technology-based in-
Against our assumptions, the results show no significant terventions. The findings from our study particularly shed
effect of general loneliness on the evaluation of an interac- light on the question how loneliness (state and general) in-
tive technology. Furthermore, general loneliness was also fluences the perception of human-likeness in interactive
not found to diminish the positive effect of an interactive technologies that are proposed to alleviate loneliness. Self-
technology’s human-likeness on how it is evaluated. As reported state loneliness had little effect on the evaluation
the additional analyses reveal, the more human-likeness of an interactive technology. However, state loneliness was
there is in an interactive technology’s appearance and be- generally low in this study, likely because interacting with
havior, the more socially attractive and generally positive any kind of interactive technology that fulfills a few so-
it is rated. With this results we are in line with previous cial check boxes elicits social reactions in humans. Overall,
research that shows that increasing certain aspects of an participants appeared to feel more at ease with an interac-
interactive technology in terms of human-likeness leads to tive technology the more it appeared and communicated
positive effects in how they are perceived (e.g., [8, 30, 29] in a human-like manner. Since the results further suggest
). However, there was a marginally significant effect with that general loneliness is linked to a less positive view of
regard to sociability: the higher people’s general loneliness, an interactive technology’s social appeal, the underlying
the less socially attractive they rate the interactive technol- mechanisms causing this effect should be investigated in the
ogy. Although this finding needs to be interpreted with future by including young adults who are severely affected
caution, it is in line with previous research that people with by general loneliness and its consequences.
unsatisfied social needs evaluate social technologies (to in-
teract with or through) more negatively [5, 7, 32, 33, 16]. Acknowledgments
In future research, this phenomenon and the mechanisms
behind it need to be further investigated also by focusing We acknowledge support by the internal DFG incentive
specifically on young adults suffering from severe general scheme for initial grant backing and the Open Access Publi-
loneliness. cation Fund of the University of Duisburg-Essen.
4.3. Limitations and Future Research References
Due to the cross-sectional nature of the study no conclusions
regarding causal relationships can be derived. Moreover, [1] M. Luhmann, D. Brickau, B. Schäfer, P. Mohr,
all results are based on self-report which can affect the M. Schmitz, A. Neumann, R. Steinmayr, Einsamkeit
accuracy and reliability of the data. Particularly loneliness unter Jugendlichen in Nordrhein-Westfalen nach der
is highly stigmatized [43] and might therefore be reported Pandemie, Staatskanzlei Nordrhein-Westfalen, 2023.
in a biased way. One of the current study’s strengths is that [2] L. A. Peplau, D. Pertman, Loneliness: A sourcebook of
it employs a real interaction. Future studies should extend current theory, research, and therapy, Wiley, 1982.
this with several interactions over a longer period and by [3] J. Baarck, B. d’Hombres, G. Tintori, Loneliness in
considering other interactive technologies such as virtual europe before and during the covid-19 pandemic,
agents or chat bots as well. Since state loneliness was rather Health Policy 126 (2022) 1124–1129. doi:10.1016/j.
low in this study, we recommend to employ techniques to healthpol.2022.09.002 .
experimentally manipulate state loneliness if the effect of it [4] M. Ghafurian, C. Ellard, K. Dautenhahn, Social com-
is of central interest (see [25] for loneliness priming). In our panion robots to reduce isolation: A perception change
study, we were also interested in investigating the effect of due to covid-19, in: C. Ardito, R. Lanzilotti, A. Mal-
the interactive technologies on individuals’ state loneliness. izia, H. Petrie, A. Piccinno, G. Desolda, K. Inkpen
(Eds.), Human-Computer Interaction – INTERACT behavior effects on perceived mind, morality, and
2021, Lecture Notes in Computer Science, Springer In- trust, International Journal of Social Robotics 13 (2021)
ternational Publishing, Cham, 2021, pp. 43–63. doi:10. 2021–2038. doi:10.1007/s12369- 020- 00692- 3 .
1007/978- 3- 030- 85616- 8_4 . [18] F. Eyssel, D. Kuchenbrandt, S. Bobinger, Effects of
[5] A. C. Horstmann, Hey robot, can you help me feel less anticipated human-robot interaction and predictabil-
lonely? an explorative study to examine the potential ity of robot behavior on perceptions of anthropo-
of using social robots to alleviate loneliness in young morphism, in: Proceedings of the 6th International
adults, in: Companion of the 2023 ACM/IEEE Inter- Conference on Human-Robot Interaction - HRI ’11,
national Conference on Human-Robot Interaction - ACM Press, New York, New York, USA, 2011, p. 61.
HRI ’23, ACM, New York, NY, USA, 2023, pp. 500–505. doi:10.1145/1957656.1957673 .
doi:10.1145/3568294.3580135 . [19] A. C. Horstmann, N. C. Krämer, Great expectations?
[6] S. Rasouli, G. Gupta, E. Nilsen, K. Dautenhahn, Po- relation of previous experiences with social robots in
tential applications of social robots in robot-assisted real life or in the media and expectancies based on
interventions for social anxiety, International Jour- qualitative and quantitative assessment, Frontiers in
nal of Social Robotics 14 (2022) 1–32. doi:10.1007/ Psychology 10 (2019) 939. doi:10.3389/fpsyg.2019.
s12369- 021- 00851- 0 . 00939 .
[7] J. P. Kluck, F. Stoyanova, N. C. Krämer, Putting the [20] A. C. Horstmann, N. C. Krämer, Expectations vs.
social back into physical distancing: The role of digital actual behavior of a social robot: An experimen-
connections in a pandemic crisis, International Journal tal investigation of the effects of a social robot’s in-
of Psychology 56 (2021) 594–606. doi:10.1002/ijop. teraction skill level and its expected future role on
12746 . people’s evaluations, PloS one 15 (2020) e0238133.
[8] A. Abubshait, E. Wiese, You look human, but act like doi:10.1371/journal.pone.0238133 .
a machine: Agent appearance and behavior modulate [21] A. Rosenthal-von der Putten, C. Strassmann, M. Mara,
different aspects of human-robot interaction, Fron- A long time ago in a galaxy far, far away…the effects of
tiers in Psychology 8 (2017) 1393. doi:10.3389/fpsyg. narration and appearance on the perception of robots,
2017.01393 . in: 26th IEEE International Symposium on Robot and
[9] J. A. Hall, N. Pennington, A. J. Merolla, Which Human Interactive Communication - RO-MAN ’17,
mediated social interactions satisfy the need to be- IEEE, 2017, pp. 1169–1174. doi:10.1109/ROMAN.2017.
long?, Journal of Computer-Mediated Communica- 8172452 .
tion 28 (2022). doi:https://doi.org/10.1093/jcmc/ [22] R. A. Ananto, J. E. Young, We can do better! an initial
zmac026 . survey highlighting an opportunity for more hri work
[10] L. Hoffmann, N. C. Krämer, A. Lam-chi, S. Kopp, Me- on loneliness, in: Companion of the 2021 ACM/IEEE
dia equation revisited: Do users show polite reactions International Conference on Human-Robot Interac-
towards an embodied agent?, in: Z. Ruttkay, M. Kipp, tion - HRI ’21, ACM, New York, NY, USA, 2021, pp.
A. Nijholt, H. H. Vilhjálmsson (Eds.), Intelligent Vir- 457–462. doi:10.1145/3434074.3447213 .
tual Agents, Lecture Notes in Computer Science, [23] C. F. Corbett, P. J. Wright, K. Jones, M. Parmer, Voice-
Springer Berlin Heidelberg, Berlin, Heidelberg, 2009, activated virtual home assistant use and social iso-
pp. 159–165. doi:10.1007/978- 3- 642- 04380- 2_19 . lation and loneliness among older adults: Mini re-
[11] A. C. Horstmann, N. Bock, E. Linhuber, J. M. Szczuka, view, Frontiers in Public Health 9 (2021) 742012.
C. Straßmann, N. C. Krämer, Do a robot’s social doi:10.3389/fpubh.2021.742012 .
skills and its objection discourage interactants from [24] F. Eyssel, N. Reich, Loneliness makes the heart
switching the robot off?, PloS one 13 (2018) e0201581. grow fonder (of robots) — on the effects of loneli-
doi:10.1371/journal.pone.0201581 . ness on psychological anthropomorphism, in: 2013
[12] C. Nass, Y. Moon, Machines and mindlessness: Social 8th ACM/IEEE International Conference on Human-
responses to computers, Journal of Social Issues 56 Robot Interaction - HRI’13, IEEE, 2013, pp. 121–122.
(2000) 81–103. doi:10.1111/0022- 4537.00153 . doi:10.1109/HRI.2013.6483531 .
[13] B. Reeves, C. Nass, The Media Equation: How People [25] S. Li, F. Yu, K. Peng, Effect of state loneliness on robot
Treat Computers, Television, and New Media like Real anthropomorphism: Potential edge of social robots
People and Places, Cambridge University Press, 1996. compared to common nonhumans, Journal of Physics:
[14] K. M. Sheldon, A. Gunz, Psychological needs as basic Conference Series 1631 (2020) 12024. doi:10.1088/
motives, not just experiential requirements, Journal 1742- 6596/1631/1/012024 .
of Personality 77 (2009) 1467–1492. doi:10.1111/j. [26] K. M. Lee, Y. Jung, J. Kim, S. R. Kim, Are physi-
1467- 6494.2009.00589.x . cally embodied social agents better than disembod-
[15] M. G. Hunt, R. Marx, C. Lipson, J. Young, No more ied social agents? the effects of physical embodi-
fomo: Limiting social media decreases loneliness and ment, tactile interaction, and people’s loneliness in
depression, Journal of Social and Clinical Psychol- human–robot interaction, International Journal of
ogy 37 (2018) 751–768. doi:10.1521/jscp.2018.37. Human-Computer Studies 64 (2006) 962–973. doi:10.
10.751 . 1016/j.ijhcs.2006.05.002 .
[16] E. Marttila, A. Koivula, P. Räsänen, Does excessive [27] C. L. Pickett, W. L. Gardner, M. Knowles, Getting a cue:
social media use decrease subjective well-being? a the need to belong and enhanced sensitivity to social
longitudinal analysis of the relationship between prob- cues, Personality & Social Psychology Bulletin 30
lematic use, loneliness and life satisfaction, Telemat- (2004) 1095–1107. doi:10.1177/0146167203262085 .
ics and Informatics 59 (2021) 101556. doi:10.1016/j. [28] N. Epley, A. Waytz, J. T. Cacioppo, On seeing human:
tele.2020.101556 . A three-factor theory of anthropomorphism, Psy-
[17] J. Banks, Good robots, bad robots: Morally valenced chological Review 114 (2007) 864–886. doi:10.1037/
0033- 295X.114.4.864 . Communication Research 17 (1990) 232–265. doi:10.
[29] J. Nakanishi, J. Baba, I. Kuramoto, K. Ogawa, 1111/j.1468- 2958.1990.tb00232.x .
Y. Yoshikawa, H. Ishiguro, Smart speaker vs. social [42] A. F. Hayes, Introduction to Mediation, Moderation,
robot in a case of hotel room, in: 2020 IEEE/RSJ and Conditional Process Analysis: A Regression-Based
International Conference on Intelligent Robots and Approach, 2nd ed., Guilford Press, 2018.
Systems (IROS), IEEE, 2020, pp. 11391–11396. doi:10. [43] N. A. Kerr, T. B. Stanley, Revisiting the social stigma of
1109/IROS45743.2020.9341537 . loneliness, Personality and Individual Differences 171
[30] L. Hoffmann, M. Derksen, S. Kopp, What a pity, pep- (2021) 110482. doi:10.1016/j.paid.2020.110482 .
per! how warmth in robots’ language impacts reac-
tions to errors during a collaborative task, in: Compan-
ion of the 2020 ACM/IEEE International Conference on
Human-Robot Interaction - HRI ’20, ACM, New York,
A. Online Resources
NY, USA, 2020, pp. 245–247. doi:10.1145/3371382. Supplementary material (data set, questionnaires, script)
3378242 .
can be accessed via https://osf.io/nzs5v/.
[31] E. van Roekel, M. Verhagen, R. C. M. E. Engels, R. H. J.
Scholte, S. Cacioppo, J. T. Cacioppo, Trait and state
levels of loneliness in early and late adolescents: Ex-
amining the differential reactivity hypothesis, Journal
of Clinical Child and Adolescent Psychology 47 (2018)
888–899. doi:10.1080/15374416.2016.1146993 .
[32] S. Li, S. Ni, K. Peng, The role of good human unique-
ness in social robot anthropomorphism influenced
by chronic loneliness, in: 2020 IEEE International
Conference on Human-Machine Systems (ICHMS),
IEEE, 2020, pp. 1–6. doi:10.1109/ichms49158.2020.
9209343 .
[33] S. Li, L. Xu, F. Yu, K. Peng, Does trait loneliness pre-
dict rejection of social robots? the role of reduced
attributions of unique humanness, in: Proceedings
of the 2020 ACM/IEEE International Conference on
Human-Robot Interaction, ACM, New York, NY, USA,
2020, pp. 271–280. doi:10.1145/3319502.3374777 .
[34] N. Dahlbäck, A. Jönsson, L. Ahrenberg, Wizard of
oz studies — why and how, Knowledge-Based Sys-
tems 6 (1993) 258–266. doi:10.1016/0950- 7051(93)
90017- N .
[35] O. Hargie, The Handbook of Communication Skills,
Routledge, 2018.
[36] K. Karrer, C. Glaser, C. Clemens, C. Bruder, Tech-
nikaffinität erfassen: Der fragebogen ta-eg [measur-
ing technical affinity—the questionnaire ta-eg], Der
Mensch im Mittelpunkt technischer Systeme 8 (2009)
196–201.
[37] M. E. Hughes, L. J. Waite, L. C. Hawkley, J. T. Ca-
cioppo, A short scale for measuring loneliness in large
surveys: results from two population-based studies,
Research on Aging 26 (2004) 655–672. doi:10.1177/
0164027504268574 .
[38] E. DiTommaso, C. Brannen, L. A. Best, Measurement
and validity characteristics of the short version of the
social and emotional loneliness scale for adults, Ed-
ucational and Psychological Measurement 64 (2004)
99–119. doi:10.1177/0013164403258450 .
[39] X. Dou, C.-F. Wu, K.-C. Lin, S. Gan, T.-M. Tseng, Effects
of different types of social robot voices on affective
evaluations in different application fields, Interna-
tional Journal of Social Robotics 13 (2021) 615–628.
doi:10.1007/s12369- 020- 00654- 9 .
[40] S. R. Fussell, S. Kiesler, L. D. Setlock, V. Yew, How
people anthropomorphize robots, in: Proceedings of
the 3rd ACM/IEEE international conference on Human
robot interaction, ACM, New York, NY, USA, 2008, pp.
145–152. doi:10.1145/1349822.1349842 .
[41] J. K. Burgoon, J. B. Walther, Nonverbal expectancies
and the evaluative consequences of violations, Human