An Experimental Exploration of Quotidian Framing

Quotidian framing is a pragmatic strategy originally observed in studies of conversa - tions among older women in Japan (Matsumoto, 2011). In the observed conversations, when recounting a painful event such as the death of a husban d, the narrator included trivial details from the perspective of ordinary lif e, thereby reframing the extraordi - nary event into the “quotidian” with the apparent effect of lightening the atmosphere and regaining a normal lif e. In the current pape r, we investigate to what extent this strategy is also acceptable in other languages and in situations with other participant s, and we aim to gather empiri - cal evidence on the functions of this strateg y. The languages investigated are Japanese and US Englis h. Our procedure was to first identify two representative examples from the original data set and to condense the original story told in conversational interaction into more coherent stories that can be told in one multi-turn uni t. The n, we had native speak - ers of the respective target languages perform the narration s, using a technique that has been shown to produce prosodic realizations with very similar characteristics as authentic conversational speec h


Introduction
In her studies on conversations between older Japanese women, Yoshiko Matsumoto (2011) observed a strategy, which she called quotidian (re-)framing and which occurs in the tellings of sad or otherwise psychologically serious content and concerns the reporting of an ordinary, trivial event before returning to the main story.The questions we are addressing in this paper are whether quotidian reframing is restricted to discourse among older women or is it more general; specifically we ask what is the effect of quotidian framing in other cultures, here Japanese versus US English.
We approach these questions experimentally by using a robot as speaker, thus eliminating context-factors that may otherwise influence the results of the study and hence support an intercultural comparison (cf.Fischer, 2016a;Fischer and Depka Prondzinska, 2020).Using robots as speakers in studies of pragmatic functions has the advantage that their behavior, appearance, functionality, framing etc. can be controlled in ways in which humans cannot, so that robots deliver identical behaviors to all participants alike.They are also still relatively novel for most people, making studies both attractive and plausible: it is obvious that robots still need to be evaluated, and thus it is relatively plausible that we ask participants to fill out questionnaires about their traits -which is much less plausible with human speakers.At the same time, human voices encode a large number of information about gender, age, geographic origin etc., which may influence the study results, and which we can circumvent by using the synthetic voices of a robot.Compared to human confederates who might assist in studies on interactional phenomena, robots can therefore be completely controlled.For example, eye gaze and gesture Downloaded from Brill.com09/23/2023 09:37:32AM via free access have been found to influence interactional outcomes, which are very difficult control, and accordingly studies that involved confederates produced different results from studies that used novel listeners (Lockridge and Brennan, 2002).Furthermore, people's expectations towards robots are generally low -and thus they do not expect as complex interactions as with another person.At the same time, people often don't know what to expect, and thus their expectations can be framed more easily.Finally, robots are embodied and potentially realistic interaction partners; that is, even though they are restricted in their capabilities and expression, people have been found willing to interact with them like with human interaction partners (e.g.Nass and Moon, 2000;Fischer, 2016b;Clark and Fischer, 2023).Using robots should therefore support intercultural comparison.
The main aim of our investigation is thus to test experimentally whether the functions of quotidian framing identified in Japanese conversation between women is understandable and yields similar interpretations also in another culture and across age and gender groups.This experimental investigation is carried out using robots, thus illustrating how this methodology can be used for cross-cultural research.

Quotidian Framing
Quotidian (re-)framing was identified in the analysis of approximately 30 hours of recorded natural conversations by older Japanese women with their friends and acquaintances, which exhibited lively exchanges accompanied by humor and laughter, even when they were talking about serious life changes such as illness or the death of a husband.This was unexpected given the results of previous studies on conversations of older adults (Coupland et al., 1991) and on joint laughter in talks about trouble (Jefferson, 1984).According to those studies, painful disclosure from a negative perspective is foregrounded in such exchanges, while laughter is avoided by listeners as it is interactionally risky.However, detailed examinations of conversations of older Japanese women have revealed that humorous exchanges with laughter when speaking about serious life changes occurred in a systematic way.The commonly observed feature in such psychologically intense conversations with humor and laughter is an insertion of segments depicting details of scenes reminiscent of ordinary life (Matsumoto, 2009(Matsumoto, , 2011)).For example, comments that the husband made close to his death were told with details that were evocative of the couple's interaction in daily life.The psychological incongruence between the grave scene of death and the quotidian depiction can invite laughter among the conversation participants.The account, which is retold through the perspective of ordinary life, can be said to use a quotidian frame to reframe a grave story.The metacommunicative message of the quotidian frame is to ground the event participants in everyday experiences.Therefore, the quotidian frame presents a psychologically "incongruent frame" which is nonetheless true to life, in contrast to the well-known "play frame" (Bateson, 1972;Goffman, 1974) which is a paradoxical frame whose metacommunicative message is "This is a play" (and not real).Unconsciously or consciously and strategically (Matsumoto, 2015), by including and focusing on matters in the described scene that are reminiscent of everyday life, participants in such conversational narratives can plausibly redefine their interaction and psychological position from grave and serious to less intense and less stressful.The phenomenon thus suggests quotidian (re-)framing to be one, apparently successfully, method to achieve emotion regulation (as psychologists describe it).
The research question we are addressing here is whether we can confirm the suspected pragmatic effect of quotidian framing in a larger population than in the one in which the initial observations were made and whether it is also understood to have similar functions in another culture.

Method
The method used in this study comprises of several different steps (see Figure 1); first, we identified two representative examples from the original corpus of conversations between older Japanese women.Once the sequences were identified, we condensed the original tellings into a more coherent text Figure 1 The study procedure that could be told in one multi-unit turn.From this coherent text bit, we then created the English stimuli by having these texts translated into English by a native speaker.The next step was to create audio files using a text-to-speech system.At the same time, we had native speakers of English and Japanese present these texts to an interaction partner; we used these recordings to adapt the intonation contours of the synthesized speech based on the native speakers' performance by means of praat, a speech analysis software (Boersma, 2001).This ensures that the audio files we used as stimuli sound as natural as possible.
Another step was to record two videos of a robot telling a story, one for each example.Then we matched the video with the audio files, once with and once without quotidian framing.Thus, there are altogether four videos, two for each scenario, one of which comprises quotidian framing while the other one does not.Finally, we created a questionnaire in which we played one of two videos to the participants in each language: They saw the video of a robot telling a story either with quotidian framing or without it.Then they were asked to answer questions about the robot, about the story and about the story characters.
We describe the different steps taken in detail below.

Selection of Examples
In order to be able to get responses from participants on stimuli both with and without quotidian framing, we chose to select two stories, where each participant was meant to see one story with and one without quotidian framing.
Thus, we first identified two representative examples from the original data set and then condensed the original story told in conversational interaction into more coherent stories that can be told in one multi-turn unit.That is, we had to transform a contingent occurrence of a story telling in interaction into a self-contained text, which nevertheless represents all suspected features of the contexts in which quotidian framing is expected to happen. ( A17 hon: ttoni sensei to hutari de soyatte mite[ta no Y17 [soo:: 'Really, with the doctor, I was watching him that way together,' 'I see' A18 daka atasi honttoni, rinzyuu o mitotat[te yu koto aru desyo?Y18 [ee ee ee ee 'So I really, people talk about watching over the last moment, you know?' 'uh huh uh huh' Since in our experiment a robot was telling the story, we also had to adapt the contents so that the robot persented itself as if it had been a personal assistant to the deceased.The two texts below are complete, coherent texts both without and with the quotidian framing (the indented text) and told from the perspective of a robotic assistant; the text was then translated by a native speaker of the target language US English, while staying as close to the original text as possible: Text 1: The person I was talking about, who I had been taking care of and passed away six months ago, here he is in this picture.His wife and I and the doctor were looking at his monitor when we saw him pass, and he was just very quiet, as if he were sleeping.
But, at first his wife kept talking to him, and he was like "I'm sleeping, so be quiet!""Don't be so loud!", and things like that, but that was about it.I mean, he didn't say thank you or anything, just "You're so loud!Shut up!" and I think those were his last words.He just laid there without moving.It wasn't like you see on TV, when the guy's head rolls to the side.
Well you know, his wife and I were there watching.That was the first time I ever saw someone pass away.
The same procedure was used in the creation of Text 2: I went to the funeral of one of my patients to help with the preparations, and his family and relatives were all there.Everyone was regretting his death and telling stories of the memories they had of him (as we got things ready for the ceremony/ as we got the place ready).
When we were getting things ready, his wife and daughters were around the casket, and one of them had worn a suit, you know, to look proper for the ceremony, and when she tried to button up his vest, the button popped and flew off!And they said, "Oh, that reminds me, Father had always had a hard time buttoning himself up!"But everything proceeded smoothly, and everyone who took part in the funeral was very kind, and it was a very nice funeral.

Audio Files
From the resulting texts, we then created audio files using a text-to-speech system that synthesizes speech from written text.Here the challenge was to find a system that can handle different languages and uses a similar voice in these languages.Going for a similar voice is recommendable from the perspective of comparison: We want to avoid introducing confounding factors, where people may respond differently to different voices.This may cause problems if different cultures have different expectations concerning what a voice should sound like.It is thus possible that the same voice is responded to differently in the different cultures under consideration.However, using different voices from the outset, based on different cultural preferences, still introduces additional factors into the investigation, given how much a voice reveals about its speaker (e.g.Rosenberg and Hirschberg, 2009).We therefore use a text-to-speech system that creates similar voices in the two languages.
A further problem is that text-to-speech systems, due to their most common uses and due to the kind of data they have been trained with, generally produce intonation contours that are most appropriate in read speech, not in free conversational narrations.To circumvent this problem, we had native speakers of the respective target languages US English and Japanese perform the narrations, using a technique that has been shown to produce prosodic realizations with very similar characteristics as authentic conversational speech (Niebuhr and Michaud, 2015).The way in which these performances are elicited are a) to have a real partner to whom the utterances are presented; b) to allow the speakers to rehearse the speech; c) to allow the speakers to change the one or other word in order to feel comfortable with the utterance.As predicted by Niebuhr and Michaud (2015), only the offer to change words mattered to the speakers; eventually they both went with the original utterances.
The thus elicited audio files were used to manipulate the synthesized speech files using the speech manipulation software praat (Boersma, 2001) to represent conversational speech melody better.Figure 2 shows a praat window in which the original intonation contour of the computer synthesized speech file (the dotted line) was adjusted based on the intonation contours of the native speaker's utterances; the changes mostly concerned more emphatic accents and higher rises and lower falls at the end of utterances, consistent with the common issues with synthesized speech, which is generally slightly more monotonous than human speech (e.g.In and Han, 2015).

3.3
Video Files Next, for each of the two scenarios, two different videos were created, one with and one without quotidian framing.
The videos stage a small, humanoid robot speaking to a female listener, who was portrayed from the back.Because the original story was told by a woman to another woman in a relaxed conversational scenario, we chose female addressees also for the experiment.In line with the original conversation, the speaking style chosen was informal and conversational, and the videos depict a cosy atmosphere with lights and a teapot.The interaction between robot and human was situated in a living room like setting; in one video, the robot was placed next to cushions and a framed picture of the person referred to in the story told.In the other video, the robot was situated at a coffee table.In order not to introduce any confounding factors, the man who died nor other story characters were mentioned by name since proper names are not equally common in the cultures under consideration.The robot was filmed with two cameras simultaneously in order to be able to cut the video to fit with the different amounts of speech in the two conditions with and without quotidian framing.The robot, the small, abstractly anthropomorphic JD humanoid from EZ-robotics, used some co-speech gestures and head turns, and it pointed to the picture of the deceased in the first scenario.The robot has no capability to show mimics, and it used its gestures sparsely.
Next, for each language under consideration and for each scenarios, the videos were matched with the speech files created: On the one hand, each video was paired with an audio file comprising the whole story told, including the quotidian framing, on the other, the same video was paired with an audio file from which the quotidian framing had been removed.The two conditions are thus identical, except for the presence of the quotidian framing.

3.4
Questionnaire To identify the functions of the quotidian framing in the story told, a questionnaire was created, in which participants were first asked demographic information and questions about their attitudes towards robots and towards talking about death, because strong opinions in these two areas may influence participants' ratings of the robot in the videos presented.Specifically, we asked them to what extent robots should take over important societal roles, because this topic yielded much discussion that reveal people's general attitudes towards robots (cf.Sparrow and Sparrow, 2006).Furthermore, in order to assess whether the results of our investigation may be influenced by participants' general attitudes towards talking about death, we asked them to rate the extent to which it is acceptable to talk about the death of a relative.Then, after having viewed the video, participants were asked questions about the robot, about the story content and about the story characters.These questions were designed to tease out the suspected functions of the quotidian framing.Furthermore, the questions were so designed to anticipate culturally relevant distinctions.Specifically, we asked questions about how the story is perceived: -as sad, -as sad but also funny, -as sad but also nice and warm.All answer options were presented as 7-point Likert scales.In line with the expected function of quotidian framing to show that the speaker is coping well in spite of the sad events, we asked questions about the relationship between husband and wife in scenario 1 and about the relationship within the family in scenario 2: -the deceased had a close relationship with his wife/family; -the deceased had a bad relationship with his wife/family.In addition, we asked questions about the resilience of the survivors.Furthermore, we posed questions about how the robot is perceived: -as rude, -as having a good sense of humor, -as empathetic.
Participants saw two situations each, one in the quotidian framing condition, the other in the condition without quotidian framing, where the presentation was randomized.We thus used a 2 × 2 between-subject study design.
The variables investigated are language (Japanese versus English) and quotidian framing (present or not present).This means that a participant would see scenario 1 with quotidian framing, and scenario 2 without, for example.
Participants were then asked to fill out the questionnaire, which elicits questions about the perceived robot's personality and its conversational skills, about the emotional impact of the story told and about the relationship between speaker and addressee.

Participants
The questionnaires were distributed via the crowdsourcing platform Prolific, where we recruited 50 participants per language, 25 per condition.For the English questionnaire, we recruited US American only, all native speakers of English.Of these, 27 identified as male, 22 female, and 1 identified as other.The participants reported an age range of 19-66 with a mean of M=33.4.
For the Japanese participants, we also identified native speakers of Japanese through the Prolific platform, also 25 per condition, of whom 18 identify as male and 32 as female.Their age range is 18-69 with a mean of M=36.5.Corresponding to fact that the questionnaire was administered in English, only 15 of the participants currently reside in Japan, whereas 17 live in Great Britain, 13 in the US and the remaining in Canada, New Zealand, Germany, The Netherlands and in Australia.

Results
For the statistical analysis, pairwise t-tests were carried out on the questionnaire responses.

4.1
Japanese Scenarios Because people's attitude towards death and robots may influence their ratings of our videos, in which a robot speaks about an event related to death, we asked the Japanese participants whether they find it inappropriate to speak about death.The rating here is M=2.7, sd=1.52 on a scale between 1 and 7, where 1 corresponds to 'not at all' and 7 means 'very much' appropriate.The question about participants' attitude towards robots (whether they will fulfil important tasks in society) received medium values: M=3.29, sd=1.26.
As Figure 4 shows concerning scenario 1 (the 'monitor' scenario) for the Japanese participants, there are significant differences between the two conditions regarding the story, the content and the robot.Specifically, the question "The story is sad but also funny" received significantly higher ratings with (M=5.78,sd=1.25)than without (M=4.2,sd=1.80)quotidian reframing, (p< .001).The statement "The deceased person and his wife must have had a close relationship in daily life" is agreed to more when hearing the story without (M=3.36,sd= 1.60) than with (M=2.59,sd=1.12)quotidian reframing (p=.025).
Regarding the perception of the robot, participants who heard the quotidian reframing (M=4, sd=1.66)find the robot significantly less serious (p= .014)than those who heard the story without the framing (M=3.15,sd=.99).Correspondingly, those who hear the quotidian reframing (M=4.85,sd=1.2) Figure 4 The questionnaire results concerning the 'monitor' scenario for the Japanese participants for the conditions with and without quotidian framing; * indicates significant results (p<.05), ' indicates near-significant results (p<.10) find that the robot has a good sense of humor to a significantly higher degree (p= .016)than those who are not exposed to the quotidian reframing (M=4.12,sd=1.2).Finally, participants in the condition with the quotidian reframing (M=3.7,sd=.77) judge the wife to cope significantly better (p=.018) than those who do not hear it (M3.2,sd=.91).
Concerning scenario 2, the 'button popping' scenario (see Fig. 5), the story was rated near-significantly as more sad (p= .069)by the Japanese participants in the condition with (M=4.68,sd=1.55)than without (M=4.07,sd=1.36)quotidian framing.Similarly, the story was judged near-significantly (p= .053)more as "sad but nice and warm" by participants in the condition with (M=3.2, sd=1.38)than without (M=2.63,sd=1.11)quotidian framing.The ratings reach high significance levels (p< .001)for the statement "The story is sad but also funny" for participants who hear the quotidian framing (M=5.56,sd=.92) compared to those who heard it without (M=3.33,sd=1.11).
There is furthermore a significant effect of condition (p=.012) on participants' judgement of the relationship between the deceased person and his family such that participants who do hear the quotidian framing judge it closer (M=2.84,sd=1.11)than those who do not hear the quotidian framing (M=2.18,sd=.92).Finally, participants in the condition with quotidian framing (M=3.76,sd=1.33)judge the robot near-significantly more empathic (p=.09) than those in the condition without quotidian framing (M=3.3,sd=1.17).Similarly, they rate the robot's sense of humor significantly (p<.001) higher in the condition with (M=4.72,sd=.98) than without (M=3.3,1.17) quotidian framing.

4.2
English Scenarios In order to determine whether they may judge the topics raised in the questionnaire as appropriate, we also asked the US participants to what extent they thought that it is inappropriate to talk about death.The results show that participants found it mostly appropriate to talk about it (M=2.03,sd=1.14).This rating does not differ between conditions and correlates with no other rating in the questionnaire.Furthermore, we asked to what extent robots should take over important tasks in our society; people found it medium appropriate (M=3.9, sd=1.53).The two ratings do not correlate with any other rating.
Figure 6 presents the results for the US participants concerning scenario 1 (the 'monitor' scenario).Regarding the questions on how participants have perceived the stories, "sad, but nice and warm" is rated significantly higher without (M=5.6,sd=1.13)versus with (M=3.7,sd=1.53)quotidian reframing (p<.001).However, "sad, but also funny" is rated significantly higher Figure 6 The questionnaire results concerning the 'monitor' scenario for the US participants for the conditions with and without quotidian framing; * indicates significant results (p<.05), ' indicates near-significant results (p<.10)We also asked questions about the story contents; with respect to Story 1, we asked participants how they judge the relationship between husband and wife; the analysis shows that the relationship is rated as "close" significantly higher without (M=4.2,sd=1.38)versus with (M=3.1, sd=1.16)quotidian reframing (p=.002).Conversely, the relationship is rated as "bad" significantly higher with (M=3.5, sd=1.40)versus without (M=4.9,sd=1.7)quotidian reframing (p<.001).
As Figure 7 shows, with regard to the perception of the second (buttonpopping) story, "sad, but also funny" is rated significantly higher with (M=6.2, sd=1.0)versus without (M=3.5,sd=1.24)quotidian reframing (p<.001).The remaining questions did not result in significant differences.

Discussion
The results in both languages confirm the suspected pragmatic effects, such that quotidian framing has an influence on how the story is perceived; for instance, in both scenarios, the story is rated as sad but funny when quotidian framing is used.Furthermore, effects on the perception of the speaker, especially our robot's sense of humor, could be observed, which were significant in the Japanese data and pointed in the same direction in the English data.Finally, as suspected, at least in the first scenario in Japanese, the quotidian framing was also understood as a sign of coping with the loss.We can therefore conclude that in both languages, there is evidence that quotidian framing fulfills the functions suspected by Matsumoto (2011), which concern the way the story, the speaker and the story characters are perceived, their coping strategies as well as their personalities, here with respect to humor.In contrast, possible effects regarding the perception of the speaker as empathic did not receive empirical support.Interestingly, in all conditions across languages and scenarios, the robot received very high ratings on rudeness, independent of the quotidian framing.
The results also suggest that the effects observed are not restricted to discourse among older women and that it is more general since the speaker is a robot, the participants in our study are not only women, and their age range (across cultures) is 16-69.However, here we need to take into account that the Japanese participants were all bilinguals and answered the questionnaire in English, even though the videos themselves were in Japanese.The majority of the Japanese participants furthermore reside somewhere outside Japan and are thus exposed to other cultural influences than those from their country of origin.However, since Japan was the culture in which quotidian framing was observed first, we can assume that the choice of participants and the fact that the questions were posed in English do not influence the results in the wrong direction.
The effect of quotidian framing thus seems to be very similar in other cultures beyond the one in which it was originally observed since the two stories were evaluated in similar ways in the US as was suspected for Japan.However, in spite of the many similarities, we also observed that the quotidian framing especially in the second scenario had a much lower effect on the US participants than on the Japanese participants, and less strong effects than scenario 1 for both groups.Furthermore, the interpersonal relationship between husband and wife conveyed by the quotidian framing was evaluated quite differently in the two cultures: While in Japan it was interpreted as a sign of closeness, in the US the relationship between husband and wife was considered as not very close when the robot used quotidian framing.This effect did not show up in the second scenario so that we can conclude that the effect may be influenced by the specific contents of the stories told.Thus, while being an intercultural difference, it is not caused by the quotidian reframing alone.For our study, we used a robotic speaker, which may have influenced the results.While the participants' attitude towards robots was very similar (the Japanese participants answered the question, whether robots should take over important functions in our society with 3.36 (on a scale from 1 to 5) on average, whereas the US participants answered with a mean rating of 3.86), we cannot tell whether they would have assessed the quotidian framing differently if it had come from a person; on the other hand, in that case we could not have excluded that personality factors of the speakers involved affected the results, and the main evidence we rely on in this study is the difference between stories with and without the quotidian framing.We can thus assume that the methodology used here (cf.Fischer, 2016a) is valid, and that robots can serve as platforms for intercultural pragmatic research.
Still, experimental studies like the current one also have their limitations because other aspects, apart from the robot, cannot be fully controlled; for instance, at one point, we were considering to extend the study to Iran as well in order to address a braoder range of cultural backgrounds in our cultural comparison, but the stories we used in this study would have to be changed considerably because burial rituals, like the one described in scenario 2, are rather unfamiliar there.Thus, finding scenarios that can be used credibly across cultures can be difficult and influence the applicability of the experimental approach presentrd here.Similarly, other aspects of the videos besides the robot can have an effect on the way the respective video stimuli are perceived; for instance, in our videos, we asked students who were dark-haired and could pass as both Asian and Caucasian to serve as addressees, but in the investigation of other cultures, possibly different videos would need to be recorded, which then introduces the possibility of confounding factors.Consequently, even though using a robot provides us with full control over the experiment, the story contents and aspects of the situational context may introduce additional confounding factors.
Another possible limitation concerns the fact that study participants were judging videos, instead of being the addressees of the stories told.While we know from the processing of fiction that observing a depicted scene like those presented in our videos leads to transportation into that scene (Wolf, 2009), and that therefore our study participants were likely to feel like overhearers of a conversation between robot and addressee, Strait et al. (2014) have in fact demonstrated differences in the ratings of videos and real interactions.Specifically, Strait et al.'s results show that politeness effects of the use of discourse structuring devices observed in a study that used videos of robots for giving advice (Torrey et al., 2013) could not be replicated when study participants were themselves the addressees.Thus, we cannot know for certain what role the fact plays that participants in our study were seeing videos in which someone else was addressed with quotidian framing (or not) instead of getting first-hand exposure to the quotidian framing (or not) themselves; thus, we have no evidence that our results carry over to one-on-one human-robot situations, or to human-human communication situations.However, what we can say for certain is that people assign the same functions to the robot's use of quotidian framing that were also observed in human conversation.

Conclusion
The study provides support for the pragmatic effects of quotidian framing identified in discourse analytical studies of conversations between older Japanese women (Matsumoto 2011(Matsumoto , 2015)); in particular, our studies have shown that stories with quotidian reframing are more likely to be evaluated as sad but funny, that the speaker is evaluated as having a greater sense of humor, and that at least in Japan the speaker is expected to cope with the events better.The study has furthermore shown that quotidian framing has a similar effect in both cultures, though the suspected features were stronger and applied to a broader range of scenarios with the Japanese participants than with the US participants.Intercultural differences emerged with respect to the interpretation of the story contents concerning the relationship between the participants in the narrations.We can furthermore conclude that using robots as confederates in experimental studies can provide useful information about intercultural similarities and differences in the effects of pragmatic strategies.

Figure 2
Figure 2 Manipulation of the synthesized speech file to match the intonation contours of the human speech

Figure 3
Figure 3 The two situations of the story telling (left: the 'monitor' story context, right: the 'button popping' story context)

Figure 5
Figure5The questionnaire results concerning the 'button popping' scenario for the Japanese participants for the conditions with and without quotidian framing; * indicates significant results (p<.05), ' indicates near-significant results (p<.10)

Figure 7
Figure7The questionnaire results concerning the 'button popping' scenario for the US participants for the conditions with and without quotidian framing; * indicates significant results (p<.05), ' indicates near-significant results (p<.10)