Decoding the Language of Care: A Typology of Caregiver Utterances

and Their Influence on Assistive Technology Use

Takeru Komori

, Dai Sakuma

1, 2 a

, Miki Saijo

and Takumi Ohashi

School of Environment and Society, Tokyo Institute of Technology, Japan, 2-12-1, Ookayama, Meguro-ku, Tokyo, Japan

Faculty of Teacher Education, Shumei University, 1-1 Daigaku-cho, Yachiyo-shi, Chiba, Japan

Keywords: Assistive Technology, Causal Attribution Theory, Caregiver-Recipient Communication, Utterance Typology,

Elderly Care.

Abstract: Amid a global caregiver shortage and a growing reliance on assistive technology, this research investigates

the intricate interactions between caregivers and care recipients in elder care settings, primarily focusing on

caregivers' verbal utterances and the conditions under which these exchanges occur. Drawing on Weiner's

causal attribution theory, we developed a typology of caregiver utterances that prompt shifts in care recipients'

attributions during the use of assistive technology. This typology—comprised of 'praise', 'affirmation

/acceptance', 'confirmation', and 'feedback' categories—illuminates key links between caregiver

communication strategies and care recipients' perception shifts. Notably, 'confirmation' utterances tend to

align with attributions to 'ability', whereas 'feedback' utterances correspond more closely with 'effort'. Our

analysis of temporal fluctuations revealed significant changes in the frequency of these utterances throughout

various stages of assistive technology usage. By offering a holistic understanding of these complex dynamics,

this study seeks to shape the development of more effective caregiver communication strategies. Such

enhancements are pivotal to optimize care recipients' experiences and engagement with assistive technology,

thus addressing the ongoing caregiver deficit. Future research endeavors will expand our dataset and examine

the potential generalizability of our findings to other caregiving environments.

1 INTRODUCTION

The global decline in birth-rates and an increasingly

aging population have resulted in a pressing shortage

of caregivers in many parts of the world. To maintain

the physical, mental, and social well-being of elderly

individuals requiring care, fostering their active

participation in society is important.

Zallio & Ohashi (2022) mention that assistive

technology has become a pivotal tool in this setting.

Such technologies not only support elderly care

recipients in physical activities but also provide a

crucial platform for meaningful interaction between

caregivers and care recipients. Effective

communication within these interactions can

significantly shape the emotional and motivational

states of the care recipients.

https://orcid.org/0009-0007-3638-8229

https://orcid.org/0000-0002-2813-5658

https://orcid.org/0000-0001-5977-5861

However, it is still unclear as to which verbal

interventions prove most influential during the usage

of assistive technology. While previous studies such

as those by Zolnierek et al. (2009) and Street (2013)

have emphasized the impact of healthcare provider

communication on patient outcomes, these studies

often overlook the specific dynamics of

communication during assistive technology use.

In this regard, Weiner's et al. (1989) causal

attribution theory offers a substantial contribution to

our understanding. This theory is a cornerstone in the

study of human motivation and suggests that

individuals interpret and predict the outcomes of

achievement tasks using four elements of attribution:

'ability', 'effort', 'task difficulty', and 'luck'. This

theory posits that these attributions significantly

influence the way individuals react to successes or

286

Komori, T., Sakuma, D., Saijo, M. and Ohashi, T.

Decoding the Language of Care: A Typology of Caregiver Utterances and Their Inﬂuence on Assistive Technology Use.

DOI: 10.5220/0012234800003598

In Proceedings of the 15th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2023) - Volume 3: KMIS, pages 286-293

ISBN: 978-989-758-671-2; ISSN: 2184-3228

failures, affecting their emotions, motivations, and

future performance.

In healthcare, Palmieri et al. (2009) explored the

applicability of attribution theory, illuminating how

healthcare providers' communication methods

influence patients' perceptions and behaviors.

Similarly, Ohashi et al. (2021) applied causal

attribution theory in social services, focusing on

interactions between caregivers and elderly patients

using an electrically assisted four-wheel cycle, a form

of assistive technology. The aim was to detect

emotional and motivational fluctuations in care

recipients during tasks to estimate attributions for

each utterance. However, existing studies primarily

focus on caregivers or care recipients individually,

neglecting the interaction between them.

This study endeavors to elucidate the interactions

between caregivers and care recipients, namely, how

the utterances of caregivers impact the willingness of

care recipients to use assistive technology. Figure 1

represents the conceptual framework of this study. To

clarify the interaction between both parties, we

utilized the user experience (UX) framework by

Hassenzahl & Tractinsky (2006). They stated, “UX is

the result of a user’s internal state (traits, expectations,

needs, motivation, mood, etc.), the characteristics of

the designed system (complexity, purpose, usability,

functionality, etc.), and the context (or environment)

within which the interaction occurs

(organizational/social setting, meaningfulness of the

activity, voluntariness of use, etc.).” The framework

offers a comprehensive approach to understanding

the interaction between technical systems and users,

providing a solid foundation for grasping the

relationship between caregivers' utterances, the state

of care recipients, and involvement with assistive

technology.

We extended the elements of “internal state,”

“context,” and “system characteristics” in the

framework as follows: The state of the care recipient

is positioned as the “internal state.” The “internal

state” of the care recipient indicates motivations and

moods, and applying the attribution theory to this is

important. The attribution theory is a framework that

explores how individuals interpret and attribute

causes to outcomes. By utilizing this, it is possible to

intricately analyze how the utterances of caregivers

affect the psychological state of care recipients,

particularly influencing motivations and the

willingness to use assistive technology. We situated

caregivers' utterances within “context.” The

utterances affect the state of care recipients,

significantly influencing how care recipients

experience assistive technology. Thus, these

utterances correspond to the overall situation or

environment in which a specific UX is formed. The

electrically assisted four-wheel cycle used in this

study was categorized under “system characteristics,”

with its features being influenced by usability and

functionality.

The objective of this study is to focus on the

communication between caregivers and care

recipients within the UX framework, clarifying their

interaction from the perspective of attribution theory.

To this end, the following steps were undertaken.

(i) To create a classification system or “typology”

for the different utterances caregivers use that

influence how care recipients attribute their actions

while using assistive technology.

(ii) To investigate how this typology is related to

changes in the care recipients' attributions, meaning

how the recipients understand and explain their own

behavior.

(iii) To study which types of utterances cause

shifts in these attributions and examine how the

effects of these utterances change over time. Here,

“temporal fluctuations” simply refers to these

changes or shifts over time.

Figure 1: Conceptual framework of this study.

2 RESEARCH METHODOLOGY

2.1 Study Design

To achieve our first objective—constructing a

typology of caregivers' utterances that influence care

recipients' attribution shifts—we utilized Otani's

(2011) qualitative Steps for Coding and Theorization

(SCAT) method. This method is well-suited for the

nuanced analysis of relatively small data sets,

allowing us to isolate and categorize impactful

caregiver utterances during assistive technology

usage. For our second objective—examining the

relationship between the constructed utterance

typology and changes in care recipients'

attributions—we employed quantitative analysis

techniques. This provided statistical insights into how

changes in caregiver communication related with

Decoding the Language of Care: A Typology of Caregiver Utterances and Their Inﬂuence on Assistive Technology Use

287

Table 1: An exemplar of the analytical procedure for Steps for Coding and Theorization (SCAT).

No. Speaker Text

(1) Focused

words from

within the text

(2) Words outside of

the text that are

replaceable with the

words from (1)

(3) Words which

explain the words

in (1) and (2)

(4) Themes and

constructs

Care

recipient

Heavy. It's heavy.

It's heavy.

Heavy, It's

heavy

Body, High load

Pedal operation,

Physical load,

Mention

Mention of physical

load

Care

giver

Yeah? Yeah, ?

Affirmation,

Confirmation

Equipment

operation,

Dissatisfaction

Confirmation of

condition

Care

recipient

You're not doing it

right, are you?

You're not doing

it right, are you?

Manipulation,

Evaluation, Self-

responsibility

Equipment

Operation, Method,

Search for cause

Search for causes of

physical load

Care

giver

Yeah. It starts a

little, you know, with

a little resistance, so

it might be heavy.

Yeah, It starts a

little, A little

resistance, It

might be heavy

Guess, Initial

movement,

Resistance, Body,

high load

Assistive

technology, Initial

resistance, Search

for cause,

Clarification

Suggestion of the

cause to the initial

operation of the

equipment

Care

recipient

This is not going to

work with Silver.

This is not going

to work with

Silver

Initial movement,

High load, Elderly,

Manipulation,

Difficulty

Elderly person,

Pedaling, Difficulty,

Suggestion

Suggestion of

difficulty in

operation by the frail

elderly

shifts in care recipient attributions. Finally, to address

our third objective—exploring the typologies of

utterances that induce shifts in attributions and

analyzing their temporal fluctuations—we employed

a comparative analysis of earlier and later

interactions. This allowed us to ascertain potential

variations in the influence of different utterances as

the caregiving interaction unfolds.

2.2 Studied Data

The data for this study were drawn from Ohashi et

al.'s (2021) research on interactions between elderly

participants using assistive technology and aides. In

this paper, "caregiver" refers to non-professional

individuals facilitating task progression rather than

providing care; we use "care recipient" for those

receiving support. Interactions occurred during a

predetermined course using an electrically assisted

four-wheel cycle for about 15 minutes. We utilized a

subset of their dataset, totaling 2514 utterances,

collected from two sessions of three unique pairings

(pairs A, B, and C), with each session indicated by a

respective number (e.g., A-1). To avoid influence

from existing relationships, each pairing involved

individuals interacting for the first time.

Conversations were recorded via a camera attached to

the technology and an additional handheld camera.

Utterances were chronologically segmented, and,

aligning with Ohashi et al.’s (2021) rules, divided into

four classifications, stemming from Weiner et al.'s

(1989) attribution theory: 'ability', 'effort', 'task

difficulty', and 'luck'.

2.3 Data Analysis

The particular procedures (1-5) used in our study are

outlined below:

(1) Utilizing the SCAT method, we

conceptualized all utterances from both caregivers

and care recipients. This process entailed identifying

phrases of interest, paraphrasing them, explicating

their meaning, and forming emergent concepts and

themes (Table 1).

(2) In order to identify if a caregiver's utterance

induced an attribution shift, we first singled out the

utterance made by the caregiver immediately before

the shift. Considering the context derived from the

storyline described by SCAT, we determined whether

the caregiver's utterance had the potential to impact

the attribution shift.

(3) The process for typifying caregiver utterances

that facilitate attribution shifts in care recipients

involved extracting the SCAT-derived themes and

constructs that were judged in step (2) to have

potentially triggered an attribution shift. We executed

a three-stage abstraction process, encompassing sub-

sub-typologies, sub-typologies, and typologies for

KMIS 2023 - 15th International Conference on Knowledge Management and Information Systems

288

these conceptual frameworks to formulate the

typology.

4) The analysis of the relationship between

attribution shifts and utterance typologies involved

counting the number of shifts incited by each

typology. This provided us with a quantitative means

to examine tendencies in both caregiver utterances

and care recipient attributions.

(5) To scrutinize the connection between the

utterance typologies and the temporal progression of

interactions, we numerically assigned utterances and

computed their ratio within total utterances. Values

less than 0.51 were considered as belonging to the

first half of the interaction, and values of 0.51 or

higher were treated as belonging to the latter half. We

calculated the frequency of typologies that could

potentially trigger attribution shifts in both halves,

thereby gaining insights into not only the evolution of

caregivers' linguistic strategies over time but also the

varying responsiveness of care recipients.

3 RESULTS

3.1 Utterance Typologies of Caregivers

Findings from procedures (1) to (3) unveiled distinct

patterns in caregiver utterances. Out of 1,474

utterances made by caregivers, 309 were singled out

as potential instigators for attribution shifts in care

recipients. We classified these utterances into four

categories as outlined by the SCAT framework:

'praise', 'affirmation/acceptance', 'confirmation', and

'feedback' (Table 2). 'Praise' includes utterances that

laud or amplify the recipients' skills or attitudes.

'Affirmation/acceptance' encapsulates utterances that

recognize or empathize with the recipients' abilities or

attitudes during the use of the assistive technology.

'Confirmation' contains utterances that confirm

essential skills, attitudes, or identify challenges.

Lastly, 'feedback' consists of utterances that provide

guidance on device operation or stimulate attitude

changes.

These findings underscore the diverse

communication strategies utilized by caregivers, and

spotlight various linguistic cues that might sway the

attribution perceptions of recipient.

3.2 Relationship Between Attribution

Shift and Utterance Typology

Table 3 illustrates the identified typologies and

quantifies the instances where caregiver utterances

instigated attribution shifts. To further understand the

relationship between attribution shifts and utterance

typologies, we utilized the Monte Carlo estimation.

This is an unbiased technique that estimates exact

significance levels by repeatedly sampling from a

reference set of tables, which share the same

dimensions and row and column margins as the

observed table. According to IBM Corporation

(2021), it allows estimation of exact significance

without relying on the assumptions needed for the

asymptotic method, which is especially beneficial

when the dataset is too large for exact computations

or fails to meet the assumptions of the asymptotic

method.

Table 2: Typologies and number of caregiver utterances extracted.

Typologies Sub-typologies Examples of caregiver utterance

Number of

utterances

Praise Praise of skills The button controls are good, aren't they? 5

Affirmation/

Acceptance

Affirmation of attitude Oh, yes. Thank you for your hard work.

Affirmation of skills Ah, yes, yes, that's right.

Confirmation

Confirmation of attitude Are you okay?

60 Confirmation of skills Are your legs getting tired?

Confirmation of problems Is it set to low speed? Is it fast?

Feedback

Encouraging attitude change You're getting used to it.

195

Guidance for task accomplishment Next time, it's better to make a wider turn.

Decoding the Language of Care: A Typology of Caregiver Utterances and Their Inﬂuence on Assistive Technology Use

289

Through the Monte Carlo estimation with 10,000

samplings, we discovered a significant deviation

between the observed frequencies of attribution shifts

for each utterance typology and their expected values

(Monte Carlo p < .001). Additionally, we conducted

post hoc tests using the Bonferroni multiple

comparisons method, and their results are

incorporated into Table 3.In our findings, within the

'confirmation' typology, 25 of the 60 utterances

prompted shifts towards an 'ability' attribution,

representing approximately 41.6%. Similarly, within

the 'feedback' topology, 82 out of 195 utterances

triggered shifts towards an 'effort' attribution,

accounting for around 42.1%.

These findings emphasize the complex interaction

between the types of caregiver utterances and the

direction of attribution shifts in care recipients.

Specifically, when the utterance typology is

'confirmation', there is a significant tendency towards

'ability' attribution, and when it is 'feedback', there is

a clear inclination towards 'effort' attribution.

Table 3: Interrelation between the utterance typologies and

the attribution shift.

Effort Ability Luck Task Total

Praise 1a 0a 2a 2a 5

Affirmation

/Acceptance

15a 8a 11a 15a 49

Confirmation 10a 25b 10a,b 15a 60

Feedback 82a 31b 41b 41a,b 195

Total 108 64 64 73 309

Note: Different letters indicate significant differences at p < .05.

The following excerpt illustrates a 'confirmation'

typology of utterance that encourages a shift towards

'ability' attribution. The context involves a care

recipient expressing concern about the height of the

seat when trying out an assistive technology.

Excerpt A-2:

Turn No. [Speaker] Utterance.

7. [Care recipient] Alright, here I go, I'm getting

on.

12. [Caregiver] Seems like the pedal is too low

to cycle. This isn't quite right.

13. [Care recipient] This is fine for me.

While the caregiver is making observations to

confirm the condition of the equipment, remarking,

“seems like the pedal is too low to cycle,” the care

recipient appears to be compromising, asserting,

“This is fine for me.” This suggests that the care

recipient is attributing the cause to their own

'capability.'

Below is an example of the 'feedback’ typology

of utterances, which encourages a shift in attribution

towards 'effort'. This instance occurs when the care

recipient expresses unease regarding the operation of

the equipment.

Excerpt A-1:

133. [Care recipient] I'm a bit scared, it's

leaning to the side.

134. [Caregiver] Just a bit more, this way, this

way.

135. [Care recipient] Oh…

136. [Caregiver] Wide turn. Wide turn. Make a

wide turn this way.

137. [Care recipient] Wide turn.

138. [Caregiver] Yes.

In response to the caregiver's instruction on the

handling of the equipment, “Wide turn. Wide turn.

Make a wide turn this way,” the care recipient

exhibits understanding by repeating the direction,

“Wide turn.” This instance clearly demonstrates the

care recipient's effort to overcome the apprehension

related to the operation of the assistive technology,

prompted by the caregiver's instructive feedback.

This interaction suggests a transition in causal

attribution towards 'effort'.

3.3 Temporal Fluctuation of Effect of

Utterance Typology on Attribution

Shift

Table 4 demonstrates the distribution of utterance

typologies throughout the progression of the

interaction. To delve deeper into the relationship

between utterance typologies and their temporal

progression, we utilized the Fisher-Freeman-Halton

exact test on the data presented in Table 4. The

analysis unveiled a significant difference between the

observed and expected distributions over time for

each utterance typology (p < .001). We also

synthesized the findings from post hoc tests, using

Bonferroni's multiple comparison method, into the

same table. This data synthesis suggests that

'affirmation/acceptance' utterances are more likely to

appear in the latter half of the progression, while

'confirmation' utterances are predominantly observed

in the first half.

As in Table 4, affirmation/acceptance', consisting

of 49 utterances, predominantly appears in the 'latter

half' of the interaction, comprising approximately

60% of the utterances. In contrast, 'confirmation',

totaling 60 utterances, is primarily observed in the

KMIS 2023 - 15th International Conference on Knowledge Management and Information Systems

290

'first half' of the interaction, accounting for about

75%. These observations elucidate a discernible

pattern between the occurrence of utterance

typologies and their placement within the temporal

progression.

Table 4: Number of caregiver utterance typologies in the

first and latter halves of the trial ride.

First half Latter half Total

Praise 1a 4a 5

Affirmation

/Acceptance

21b 28a 49

Confirmation 45b 15a 60

Feedback 109a 86a 195

Total 176 133 309

Note: Different letters indicate significant differences at p < .05.

4 DISCUSSIONS

4.1 Interplay Between Attribution Shift

and Utterance Typologies

This study delved into the complex relationship

between the caregiver's utterance typologies and the

care recipient's attribution shifts amidst the usage of

assistive technology. The principal findings

underscore two salient relationships: 'confirmation'

utterances from caregivers commonly lead care

recipients to attribute their actions to 'ability'. In

contrast, 'feedback' utterances result in care recipients

attributing their actions to 'effort'.

These findings align with the pedagogical models

proposed by Bangert-Drowns et al. (1991) and Shute

(2008). The model by Bangert-Drowns et al. outlines

that learners experience a sequence of stages, from

advice reception to performance adjustment, based on

their own evaluation. In this framework, a caregiver's

advice may act as a catalyst, prompting the care

recipient to re-evaluate their performance and adjust

their actions accordingly. Shute's research further

emphasizes the importance of feedback tailored to the

learner's state, which includes task simplification,

goal adjustment, and the reinforcement of correct

methods. When applied to the caregiving context, a

caregiver's feedback might modulate a care recipient's

motivation, interest, goals, and knowledge, thereby

instigating changes in their behavioral state.

Building on prior research, our study offers novel

insights into the influence of caregiver utterances on

care recipients' assistive technology usage.

Specifically, 'confirmation' utterances, which often

involve checking the care recipient's equipment

operation and their physical and mental state, seem to

bolster the care recipient's acknowledgement of their

own ability. As a result, care recipients are more

likely to attribute their actions to 'ability'.

Similarly, 'feedback' utterances, typically offering

guidance on equipment operation and attitude

improvement, can heighten care recipients' awareness

of their goal achievement progress. This heightened

awareness often prompts an increased effort from

care recipients, making them more likely to attribute

their actions to 'effort'.

By pinpointing these specific utterance-attribution

relationships, our study enhances the understanding

of the interplay between caregiver communication

strategies and care recipients' perceptions and

utilization of assistive technology.

4.2 Relationship Between Utterance

Typologies and Temporal

Fluctuation

The care recipient's test ride time was divided into the

first and latter half, and the trends in utterance

typologies were analyzed from a temporal fluctuation

perspective. The results showed that the utterance

typologies 'affirmation/acceptance' was significantly

more prevalent in the latter half of the test ride, while

the typology of utterances 'confirmation' was

significantly more prevalent in the first half.

Two potential influences may account for the

noticeable discrepancy in the timing of utterances

prompting attribution shifts. The first is the potential

transformation in the caregivers' utterances between

the first and latter halves of the period. The second is

the possible alteration in the care recipients' state,

specifically, their responsiveness.

To discuss the potential shift in the caregiver's

utterances over time, actual examples of utterances

are provided. The utterance example below is from a

scene where the care recipient, seen near the first half

of the test ride, is commenting on the strain on their

legs concerning the operation of the four-wheel cycle

with electric assist.

Excerpt A-1:

224. [Care recipient] This is heavy.

225. [Caregiver] It's heavy, is it? Well, you can

pedal a bit faster then.

226. [Care recipient] Okay, I'll pedal faster. It's

pointless otherwise.

In response to the care recipient's remark, the

caregiver directs the locus of causality toward the

Decoding the Language of Care: A Typology of Caregiver Utterances and Their Inﬂuence on Assistive Technology Use

291

care recipient himself, suggesting “Well, you can

pedal a bit faster then.” In response, the care recipient

attributes the causality to the 'task difficulty',

asserting, “It's pointless otherwise.”

The following presents an analogous scene

occurring in the latter half of the same pair's trial.

Excerpt A-1:

281. [Care recipient] But, this is rather

overwhelming. It's heavy and feels like it's

moving at high speed.

286. [Caregiver] Hmm, I wonder what could be

the reason. Maybe we should ask someone from

Yamaha.

287. [Care recipient] (Laughs)

In a similar situation, in response to the care

recipient's remarks, “overwhelming” and “heavy,”

the caregiver shifts the attribution of the cause to the

assistive technology, suggesting, “Maybe we should

ask someone from Yamaha” (Yamaha:

Manufacturing company name). This statement

directs the cause of the problem not to the care

recipient himself, but to external factors such as the

environment or equipment. The care recipient

responds with a smile, inferring a shift in attribution

of the cause towards 'luck'. Thus, by altering the

method or trend of their linguistic interventions

between the initial and latter stages of the trial ride,

caregivers may have instigated different shifts in the

care recipients' attributions, even under similar

scenarios. This could be one potential explanation for

the observed temporal shift in the occurrence of

specific types of caregiver utterances that trigger

attribution shifts.

Next, in terms of the latter possibility, that the care

recipient's state, namely their responsiveness, shifted

in the first and latter half of the ride, an explanation is

provided with actual examples of conversation. The

following utterance is an example observed in the

first half of the ride, where the care recipient feels

anxious about operating the assistive technology.

Excerpt C-1:

212. [Care recipient] I'd feel terrible if I were to

damage anything here.

213. [Caregiver] No, no, you're all right. You're

absolutely fine.

215. [Care recipient] That won't do.

216. [Care recipient] Indeed, you're okay.

218. [Caregiver] I'm sorry, but I may not be able

to do this. I feel bad for causing everyone

inconvenience.

219. [Care recipient] You're entirely fine, truly.

221. [Care recipient] This, this part is... it's not

going well at all.

In response to the care recipient's expression of

apprehension about operating the device, conveyed

through their words, “I'd feel terrible if I were to

damage anything here,” the caregiver seeks to

reassure by assuring them, “you're all right,”

Contrarily, the care recipient exhibits a negative

reaction towards operating the device, stating, “ I may

not be able to do this,” and “it's not going well at all”.

This indicates that the care recipient's attribution is

directed towards their own abilities.

A subsequent example from a similar situation in

the latter part of the ride with the same pair is

presented.

Excerpt C-1:

545. [Care recipient] No, it wouldn't be good if it

gets damaged. Just a bit more.

546. [Caregiver] It's okay, it's okay. You're

doing fine.

547. [Care recipient] Is it okay?

548. [Caregiver] Yes, it's perfectly fine.

551. [Care recipient] Ah, it's hitting. Is it not

hitting?

552. [Caregiver] Yes, everything is fine.

555. [Care recipient] So, like this.

In response to the care recipient's apprehension

about potentially damaging the equipment, expressed

with phrases such as “it wouldn't be good if it gets

damaged,” the caregiver alleviates the concern by

affirmatively stating “you're doing fine.”

Consequently, the care recipient exhibits a greater

commitment to the operation of the device through

proactive queries like “Is it not hitting?” and assertive

declarations such as “So, like this.” This behavior

indicates a shift in attribution towards their own effort.

Hence, it can be surmised that the temporal

transition occurring within the trial duration, despite

similar situations and identical linguistic

interventions from the caregiver, is driven by a

transformation in the care recipient's reactive state.

This transformation could potentially explain the

observed changes in the frequency of caregiver

utterances that promote attribution shifts.

5 CONCLUSIONS

The aim of this study is to focus on the

communication between caregivers and care

recipients within the UX framework and to elucidate

their interactions from the perspective of attribution

theory. By deepening the understanding of these

dynamics, we can promote more effective verbal

interventions by caregivers and encourage more

KMIS 2023 - 15th International Conference on Knowledge Management and Information Systems

292

productive use of assistive technology by care

recipients.

In alignment with Weiner's causal attribution

theory, we constructed a typology of caregivers'

utterances that influence care recipients' attribution

shifts during the use of assistive technology. We

delineated four main categories of utterances: 'praise',

'affirmation/acceptance', 'confirmation', and

'feedback'. Our analysis revealed that 'confirmation'

utterances were frequently associated with

attributions to 'ability', while 'feedback' utterances

were more often linked with 'effort'. These findings

suggest that caregivers adapt their language to guide

shifts in care recipients' attributions, tailoring their

responses to different situations.

Our temporal fluctuation analysis demonstrated a

significant difference in the frequency of

'affirmation/acceptance' and 'confirmation' utterances

between the initial and later stages of assistive

technology usage. This shift may reflect changes in

caregivers' verbal strategies and the care recipients'

responses over time.

The study acknowledges its limitations, such as

the need for a larger and more diverse dataset to

permit a comprehensive understanding of the

relationships among attribution, utterance categories,

and temporal fluctuation data. Future research will

focus on expanding our dataset and investigating

whether these utterance typologies and shifts in

causality perception can be applied to other

caregiving contexts. Additionally, to verify the

validity of the findings of this study, obtaining

feedback from either caregivers or care recipients is

considered as a subsequent step.

ACKNOWLEDGEMENTS

This work was partly supported by JSPS KAKENHI

Grant Number 20K20257. This work utilized

OpenAI's ChatGPT for initial drafting, which was

thoroughly reviewed, edited, and supplemented by

the authors. We therefore assume full responsibility

for the final content of this publication.

REFERENCES

Zallio, M., & Ohashi, T. (2022). The Evolution of Assistive

Technology: A Literature Review of Technology

Developments and Applications. AHFE Open Access,

vol 37. 85-93. http://doi.org/10.54941/ahfe1001646

Haskard Zolnierek, K. B., & DiMatteo, M. R. (2009).

Physician Communication and Patient Adherence to

Treatment: A Meta-Analysis. Medical Care, 47(8),

826–834. https://doi.org/10.1097/MLR.0b013e

31819a5acc

Street, R. L. (2013). How clinician–patient communication

contributes to health improvement: Modeling pathways

from talk to outcome. Patient Education and

Counseling, 92(3), 286–291. https://doi.org/10.

1016/j.pec.2013.05.004

Schuster, B., Forsterlung, F., & Weiner, B. (1989).

Perceiving the Causes of Success and Failure: A Cross-

Cultural Examination of Attributional Concepts.

Journal of Cross-Cultural Psychology, 20(2), 191–213.

https://doi.org/10.1177/0022022189202005

Palmieri, P. A., & Peterson, L. T. (2009). Attribution theory

and healthcare culture: Translational management

science contributes a framework to identify the etiology

of punitive clinical environments. In G. T. Savage & M.

D. Fottler (Eds.), Advances in Health Care

Management (Vol. 8, pp. 81–111). Emerald Group

Publishing Limited. https://doi.org/10.1108/S1474-

8231(2009)0000008008

Ohashi, T., Watanabe, M., Takenaka, Y., & Saijo, M.

(2021). Real-Time Assessment of Causal Attribution

Shift and Stay Between Two Successive Tests of

Movement Aids. Integrative Psychological and

Behavioral Science, 55(3), 541–565.

https://doi.org/10.1007/s12124-020-09592-7

Hassenzahl, M., & Tractinsky, N. (2006). User

experience—A research agenda. Behaviour &

Information Technology, 25(2), 91–97.

https://doi.org/10.1080/01449290500330331

Otani, T. (2011). “SCAT” A Qualitative Data Analysis

Method by Four-Step Coding : Easy Startable and

Small Scale Data-Applicable Process of Theorization.

Bulletin of the Graduate School of Education and

Human Development. Educational Sciences, 54(2), 27-

44. https://doi.org/10.18999/nueduca.54.2.27

IBM Corporation. (2021). Exact Tests. In SPSS Statistics

25.0 Documentation. IBM Documentation. Retrieved

July 24, 2023, from https://www.ibm.com/docs/

en/spss-statistics/25.0.0?topic=tests-exact

Bangert-Drowns, R. L., Kulik, C.-L. C., Kulik, J. A., &

Morgan, M. (1991). The Instructional Effect of

Feedback in Test-Like Events. Review of Educational

Research, 61(2), 213–238. https://doi.org/10.

3102/00346543061002213

Shute, V. J. (2008). Focus on Formative Feedback. Review

of Educational Research, 78(1), 153–189. https://

doi.org/10.3102/0034654307313795.

Decoding the Language of Care: A Typology of Caregiver Utterances and Their Inﬂuence on Assistive Technology Use

293