1 Introduction

Inclusive design is about designing products, services and environments that “are accessible to, and usable by, as many people as reasonably possible […] without the need for special adaption or specialized design” (Persson et al. 2015). One fundamental aim of inclusive design is to address as many users as possible with only one product while understanding the challenging diversity of heterogeneous user needs (cf. Goodman-Deane et al. 2010; Luck 2013; Steinfeld 2013; Waller et al. 2015). Stockton (2009) states that ‘stigma can be seen as the Achilles heel of Inclusive Design’. The reason for this is that using technological artefacts can impose negative societal perceptions on their users (Stockton 2009). How a product is perceived by a person is thereby highly subjective and depends on various psychological dimensions (Carbon 2019). The term stigmatisation generally describes moments or situations where an individual is treated poorly by others like being ignored or discredited, because he or she deviates from the respective social norm (Carneiro et al. 2015; Vaes 2014). The triggers for a stigmatising behaviour against a person are versatile and range from cultural heritage to individual characteristics such as gender or disabilities (Crocker et al. 1998; Goffman 1963).

Figure 1 shows that product usage can also foster such stigmatisation leading to product-related user stigma (Vaes 2019). However, not every product automatically causes stigmatisation. Some products are more powerful triggers than others. The likelihood of provoking stigmatisation is referred to as stigma-sensitivity which depends on the product type (e.g., protective, assistive or medical products), the visibility (e.g., of the interaction, the product or individual product areas) and the product design (e.g., aesthetics, appearance, symbolism) (Vaes 2014). Assistive devices are particularly affected (Vaes 2014), because their use is usually associated with a person's disability or limitation (Parette and Scherer 2004). Such negative associations often arise from a strong focus on good usability and a negligence of emotional aspects [cf. (Vaes 2019)]. Resulting product designs may be easy to use but have a clunky and old-fashioned look, which is often not aesthetically appealing to the user and highlights the user’s limitations (Bispo and Branco 2009; Scherer 2000; Shinohara and Wobbrock 2011; Wobbrock et al. 2011). Whether a product is used permanently (e.g., hearing aid) or temporarily (e.g., walker) is irrelevant.

Fig. 1
figure 1

Concept of product-related user stigma (Schröppel et al. 2021; Vaes 2019)

Full size image

However, the aim is to create a product design that is both emotional and usable and strengthens the self-efficacy of a user. Promoting a positive perception of usability and emotionality through certain design implications could particularly help in the case of stigma-sensitive products as we know that regular exposure to stigmatising reactions has a negative impact on a person’s self-confidence (Bichard et al. 2007; Shinohara and Wobbrock 2011), self-awareness (Major and O'Brien, 2005; Wright 1983) and/or self-esteem (Shinohara and Wobbrock 2011; Wilkinson and Angeli 2014). Therefore, stigma-sensitive products are associated with lower acceptance [e.g., (Jacobson 2014; Louise-Bender et al. 2002)] and higher rejection rates [e.g., (Parette and Scherer 2004; Rønneberg Næss and Øritsland 2009)]. Product rejection is particularly problematic with respect to assistive devices given their potential to improve quality of life and everyday competence (Schulz et al. 2015). Moreover, the ways in which people think and feel about their own age is strongly associated with product use. For example, subjective perceptions of obsolescence in later life have been found to predict poorer performance when using everyday technology such as blood pressure monitors (Schmidt and Wahl 2019). In addition, a robust body of knowledge exists showing that maintaining positive and avoiding negative aging-related views and attitudes are critical for mental and physical health across life (Wurm et al. 2017).

Based on these arguments, this research explores the role of usability and emotional design as a positive influence for self-efficacy in the context of walkers. Walkers are assistive devices for people with limited walking abilities providing support for mobility and functional independence. Walkers are mainly used in public and their appearance is often old-fashioned and usability-driven which makes them prone to product-related user stigma (cf. Buker 2022). Overall, if a correlation is found between perceived self-efficacy and perceived usability and emotionality, it shows the benefit and the power of product design adaptations and their meaning in terms of strengthening the user's self-efficacy and thus to create a positive user experience while possibly also compensating product-related stigma.

1.1 Related terms and definitions for understanding the relationship between self-efficacy and the constructs of usability and emotionality

To better understand the aim of the study, it is important to agree on the terms used to describe the constructs being studied. In the context of product usage, the concept of self-confidence refers to the psychological construct of self-efficacy. Self-efficacy is defined as the subjective certainty of being able to cope with new challenging situations based on one's own competence (Schwarzer and Jerusalem 2002).

This paper examines the relationship between the constructs of usability and emotionality in relation to self-efficacy. To classify and discuss the observed results, the relationship between usability and emotionality as independent or interdependent constructs is further discussed in the literature. Different design theories consider usability and emotionality as part of the other and thus as possible cause and effect. ISO 9241-11:2018 (ISO 9241-11:2018), for instance, defines usability as an ‘extent to which a system, product or service can be used by specified users to achieve specified goals with effectiveness, efficiency and satisfaction in a specified context of use’. Thereby, the meaning of satisfaction evolved within the standard from referring to the absence of discomfort in 1988 to referring to physical, cognitive and emotional reactions of the user in relation to the user’s expectations in the revised version of 2018 (Buker et al. 2022). Thus, an increasing number of ‘soft factors’ are being integrated into what used to be a primarily ergonomic usability. Considering emotional design, Norman (2005) proposed the emotional processing on three levels: the visceral level refers to the appearance of a product and its perceptible qualities; the behavioural level relates to the effectiveness and ease of use of a product; the reflective level considers the rationalisation and intellectualisation of a product (i.e., self-image, personal values etc.). The behavioural level may thus be seen as an equivalent to usability. Despite a certain overlap, usability and emotional design cannot be entirely equated. The fulfilment of usability-related goals during product development is not enough to arouse the user’s emotions and satisfaction (Nurkka et al. 2009). Emotional design rather goes beyond good usability including attractiveness, approval, desire and pleasure (cf. Desmet 2002; Hekkert and Desmet 2002; Jordan 2002). To meet those emotional driven user demands, factors such as aesthetics or identification with the product have to be taken into account, which are not necessarily related to usability. In selected cases, such as assistive devices, emotionality is often neglected, whereas usability is given special attention (Vaes 2019). Kurosu and Hashizume (2014) therefore recommended a separate consideration of subjective and objective product quality to emphasise the importance of emotionality as part of the subjective quality of a product. As an assistive devise is going to be addressed and analysed within this paper, we adapt this separated examination of both constructs. Based upon the ISO 9241-11:2018 (ISO 9241-11:2018), usability is thus understood as the product’s ability to be used efficiently and effectively by the user in a specific context of use addressing the physiological and cognitive level of the user. Emotional design is contrarily understood as the evocation of positive emotions and the creation of an emotional bond between product and user through factors, such as aesthetics, product loyalty and originality (cf. Kurosu 2015; Thüring and Mahlke 2007). Thereby, emotions are seen as complex patterns of physical arousal that are subjectively perceived and influence the observable behaviour towards a product (e.g., product use or rejection) (Hofbauer and Sangl 2011).

1.2 Related work on reducing product-related user stigma with design activities and the role of self-efficacy in this context

Reducing product-related user stigma is a complex, highly interdisciplinary challenge bringing together social science, psychology and engineering design. While conducting a systematic literature review, Schröppel et al. (2021) identified different approaches to counteract product-related user stigma from an engineering perspective. Thereby, general strategies make an important contribution by addressing user stigma in terms of product intervention, user empowerment and cultural interventions [cf. (Vaes 2019)]. Product intervention strategies try to improve the product itself, e.g., by changing the meaning of the product and creating symbolism (Bispo and Branco 2009; McCarthy et al. 2017; Vaes et al. 2012) or using camouflage or disguise (Barros et al. 2011; Green 2009), whereas cultural intervention strategies aim to transform the perception of stigma, e.g., by providing educational services (Stockton 2009) or using campaigns to address the issue of user stigma (Vaes 2019). User empowerment strategies, instead, focus on strengthening the users’ internal coping resources by giving a person enough confidence to use a stigma-sensitive product. They do not aim to avoid stigmatising reactions of third parties, but to make the affected person less vulnerable, e.g., by making sure the product provides extra abilities for the product user (Vaes 2019) or raising the user’s self-confidence while minimising the level of psychological and emotional distress (Partheniadis and Stavrakis 2019). All these strategies are very general and can be applicable to a wide range of different products. Yet, there is a prevailing lack of knowledge on how to actually apply them through design activities (Schröppel et al. 2021). The realisation of these strategies is correspondingly difficult and challenging, especially for those product developers who are inexperienced in the field of stigmatisation.

Our research agenda aims at improving the implementation of stigma reducing strategies to empower user self-confidence (cf. Partheniadis and Stavrakis 2019). For this purpose, we focus on identifying and examining possible ‘starting points’ for product development. So far, little research has been done on how the user's confidence can be strengthened through the design of physical products. For instance, Jordan (1998) stated that both a feeling of confidence during product usage (e.g., having the feeling to have control over the interaction) and a user’s self-confidence after use can be evoked while providing the user with pleasurable products. In another study, it was found that positive usage experience with assistive devices can improve self-confidence (Scherer 2000). Moreover, Shinohara and Wobbrock’s (2016) examined assistive device usage showing that participants felt competent and reported increased self-confidence when the product was used successfully. Contrarily, feelings of improper operation of a product or loss of control over a product can have the opposite effect. In their study, Tsao and Chan (2011) found evidence that people tend to blame themselves for embarrassing situations that result from the aforementioned aspects. All those findings coincide with theoretic models, like the A.C.T. model by van Gorp and Adams (2012), that suggest a combination of functionality, emotionality and usability as the basis for great user experience. While functionality is usually a core competence of a company (Pahl et al. 2007), the equal consideration of usability and emotional design is still a challenge (Buker et al. 2022).

Taken together, findings from the literature indicate both usability and emotional design to be adequate instruments to boost user self-confidence. Yet, this assumption has not been systematically explored. We do not know whether self-efficacy is empirically associated with usability or emotionality and whether both constructs have a combined influence on self-efficacy.

2 Aim and study objectives

This paper aims to examine whether self-efficacy would be associated with perceptions of a product’s usability and positive emotions. As this paper follows a separate understanding of both constructs, their association with the user’s perceived self-efficacy is tested within separate hypotheses (H1 and H2). To take into account the close relationship between emotionality and usability, possible interaction effects will also be considered (H3). The three derived hypotheses are:

  1. H1:

    Perceptions of better product usability are positively associated with the user’s perceived self-efficacy.

  2. H1:

    Perceptions of an emotionally appealing product design are positively associated with the user’s perceived self-efficacy.

  3. H1:

    A user’s perceived self-efficacy is mostly improved when both aspects are considered in the product.

To test these hypotheses, an empirical study was conducted focusing on the stigma-sensitive product walker that is used among people with limited walking abilities. This study will give useful insights to assess the potential of and prospectively confirm usability and emotional design as valid ‘starting points’ to strengthen a user’s self-efficacy. We assume that self-efficacy is influenced by other factors besides the usability and emotionality of a product. To secure the study results, important covariates such as demographic user attributes, health status and user experiences are included in the survey and will be considered within the statistical analyses.

3 Methods

3.1 Participants

The sample was an ad-hoc sample in which we particularly wanted to represent older adults, because this population is more likely to use walkers. However, even if no walker is currently used, age-related use is still a realistic scenario for most people, which is why an assessment based on imagination seems possible. Recruitment took place via postings on online platforms for older adults. In total, 105 persons ranging from 17 to 92 years of age (M = 62; SD = 21.13) participated in the study (see Fig. 2a). This resulted into a statistical power of 0.81 to identify a medium effect of f2 = 0.15 within a multiple regression model with seven predictor variables. The gender distribution was balanced with 48% female and 52% male participants (see Fig. 2b). The majority of the participants was well educated, 65% had at least an advanced technical college certificate (see Fig. 2c). Most of the participants felt healthy with 47% reporting good health; 25% mentioned very good health; 10% reported excellent health; 19% reported mediocre or poor health. 37% of the participants reported experience with walkers.

Fig. 2
figure 2

Overview of the participants’ demographic data

Full size image

3.2 Study design

3.2.1 Procedure

The online survey consisted of two parts. The first part included the collection of different user-specific information including demographic data (age, gender and educational level), health-related data (walker use and subjective health) and ratings of the respective products usability and emotional appealingness. The second part included the visual product assessment. Based on a between-subjects design, one of four walkers was randomly presented to the participants. The observed walker had to be rated in terms of usability, positive emotional appealingness (positive emotions) and the perceived subjective self-efficacy. This approach allowed us to reduce the risk of potential bias that would occur if participants had different products for comparison (Charness et al. 2012). The participants took an average of 13.5 min to complete the survey (SD = 8.4). This study was based on an online survey, and, therefore, it was only possible to present spatial images of walkers. However, ergonomic features were highlighted during product presentation to ensure that participants got a better understanding for the usability of the respective product. In particular, the respondents were made aware of special features of the models in the form of an information box with textual information, e.g., on weight, type of brake, available height adjustment or ergonomic handles. No further highlighting was used to minimise bias and at the same time allow an initial ergonomic assessment to be made on a purely visual basis. The study design thus focused on visual perception, which showed to be particularly important at the beginning of product usage (Fenko et al. 2010). User tests that consider multisensory product perception are planned in follow-up studies, but are currently not subject of the examination.

3.2.2 Product demonstrator walker

As already mentioned in the introduction, walkers are often used in public settings and have a high-level of stigma-sensitivity. They also often lack an appropriate balance of emotional and usable design whereas emotionality is mostly neglected which may lead to product rejection. Walkers were thus chosen as product demonstrators for the conducted study. The four walkers had different levels of usability and emotionally appealingness (see Fig. 3). To ensure comparable and standardised images within the online survey, the walkers were remodelled in a 3D-CAD-System via reverse engineering on the basis of existing products. The reference product for model A has a rather cheap-looking frame. To move it forward, the rear part must be actively lifted due to the lack of wheels. In addition, there are neither ergonomic handles nor a storage or the possibility to sit down. The reference product for model B has a very clean and appealing look, but does not have a brake, ergonomic handles or height adjustment. A seat is provided in form of a loose plank. The reference product for model C has an old-fashioned look, however, in terms of usability it was rated excellent by the German consumer organisation ‘Stiftung Warentest’ (Nusser 2019). The walker has, for instance, a good rollability, height adjustment, comfortable handling, a brake and a bell. The reference product for model D has a dynamic shape and a stylish look. The brake cables are hidden and the walker has a good power and shock absorption as well as ergonomic handles and a low weight. The reference product won the Red Dot Design Award in 2019 (byACRE 2019). To neutrally confirm the differences in usability and emotional appealingness, the walkers were pre-evaluated by three experts from user-centred design. For this purpose, the experts were shown the standardised images of the four walkers. They were then ranked in terms of their usability and emotional appealingness. Thereby, it was confirmed that model C and D have a good, model B has an average and model A has a poor usability. Furthermore, models B and D were rated well in terms of emotional appealingness, while models C and A scored poorly (see Fig. 3). Fleiss’ kappa was run to determine the agreement of the experts on the ranking of the four walkers. Thereby, Fleiss’ kappa showed substantial agreement (κ = 0.778; p = 0.000) [cf. (Landis and Koch 1977)].

Fig. 3
figure 3

Overview of product variants used during study

Full size image

3.3 Measures

The individual constructs were assessed using validated psychometric instruments from the literature ensuring high quality data. Regarding the assessment of usability and emotional design, a multitude of different instruments exist in literature. The instruments used within this study represent a selection of them. They were chosen with regard to their suitability for the merely visual assessment of the walkers as well as the product context of assistive devices.

3.3.1 Personal variables

3.3.1.1 Educational level

Educational level was specified as a dummy variable (0 = secondary school or below; 1 = university entrance diploma).

3.3.1.2 Experience with walkers

Experience with walkers was measured with questions asking whether participants currently use walkers, had used walkers in the past or whether they ever tested a walker. The respective variable was included as a dummy variable (0 = no experience; 1 = experience).

3.3.1.3 Subjective health

Subjective health was measured on a 5-point-scale ranging from 1 (poor) to 5 (excellent).

3.3.1.4 Meaning of usability and emotional appealingness

In terms of measuring the personal meaning of usability and emotional appealingness on a 5-point-scale from 1 (very unimportant) to 5 (very important) the participants were asked to consider both aspects generally in the context of product usage.

3.3.2 Usability

Usability was measured with the rapid assessment of product usability and universal design (RAPUUD; Cronbach’s α = 0.80) (Lenker et al. 2011). The RAPUUD consists of 12 items (see Table 1) which were merged into the total usability score via the arithmetic mean. Therefore, negative questions were recoded preliminary as recommended by Lenker et al. (2011). For a better comparability, a uniform 7-point scale was used for the construct’s usability, positive emotions and self-efficacy. Agreement was assessed from 1 (do not agree at all) to 7 (completely agree). The mean was calculated with higher values reflecting better usability.

Table 1 Items of RAPUUD
Full size table

3.3.3 Positive emotions

Positive emotions are a fundamental aspect of emotional appealingness. They were collected with the user emotions module of the meCUE questionnaire (Cronbach’s α = 0.94/0.91). The meCUE is a modular tool for measuring user experience (Minge et al. 2016). The dimension of positive emotions was calculated via the mean of six items (see Table 2). Agreement was measured on a 7-point-scale (1 = “do not agree” to 7 = “completely agree”). A higher mean value of the items reflected higher levels of positive emotions.

Table 2 Items of meCUE—positive emotions [English version for understanding; German version see (Minge and Riedel 2013)]
Full size table

3.3.4 Self-efficacy

Self-efficacy was captured with the modified computer self-efficacy scale (mCSES; Cronbach’s α = 0.94), which was introduced by Laver et al. (2012). Within the conducted study, the focus lies upon walker usage during sole use without the help of others. This dimension consists of five items and is calculated by adding up the scaling per item (see Table 3). Agreement was measured via a 7-point-scale from do not agree at all (1) to completely agree (7). Thus, the highest self-efficacy score is 35, the lowest is a value of five. As the study was conducted in Germany, the items of the mCSES were translated into German and adapted to the context of walker use.

Table 3 Items of mCSES—sole use (English version for understanding)
Full size table

3.4 Statistical analyses

Descriptive statistics were used to explore the characteristics of the included participants (see chapter participants). The individual meaning of usability and emotional appealingness of a product was analysed via a bar chart. Hypotheses were tested by conducting correlation and regression analyses. The total usability score and positive emotions were correlated with self-efficacy during sole use applying the Pearson correlation coefficient r to explore basic dependencies between these constructs. A p level of 0.05 or below was interpreted as significant. Regression was used to confirm these correlations taking the participant’s age, gender, educational level, subjective health and experience with walkers into account as important covariates. To test H1 and H2, usability and positive emotions were initially examined independently of each other. In terms of H3, both aspects were reviewed together considering their interaction effect. The variables for the interaction term were centred to make the model more interpretable. All statistical analyses were completed using the statistical software Stata 16 (StataCorp., 2019).

3.5 Research ethics and consent

An online-based survey was conducted as part of the research. Participants were asked to rate statements using various scales and to answer general demographic questions. There is no physical, financial, occupational or social risk for the participants. Participants were free to withdraw from the study at any time. Ethical approval of the study was therefore not required. At the beginning of the study, the participants were informed transparently regarding the purpose of the survey. Moreover, data protection information was provided. The data is anonymised and treated confidentially in accordance with the legal requirements of the European Data Protection Regulation. Each participant consented voluntarily to the processing and use of the data in advance.

4 Results

4.1 Descriptive statistics

The general importance of usability and emotional appealingness of a product varied among the participants (see Fig. 4). Usability was mostly rated as very important (59%) or important (33%), whereas emotional appealingness was less specific. Most participants described it neither as important nor as unimportant (so–so: 38%). Still, 33% of the participants perceived emotional appealingness as important and 8% as very important.

Fig. 4
figure 4

Importance of usability and emotional appealingness of a product

Full size image

Table 4 shows findings from the conducted correlation analysis. It revealed significant relationships between usability and self-efficacy during sole use (r = 0.276; p = 0.002) as well as between positive emotions and self-efficacy during sole use (r = 0.161; p = 0.050). This means that participants who reported higher usability or higher emotional appealingness were more self-efficient to use walkers on their own compared to those participants who rated the walker less usable and less emotional appealing. In the following regression analyses, self-efficacy during sole use is considered in more detail.

Table 4 Pearson correlation analysis including usability, positive emotions and self-efficacy during sole use (N = 105)
Full size table

4.2 Regression analyses

Table 5 shows findings from three regression models. The first model tested whether self-efficacy would be positively associated with perceptions of usability while controlling for covariates (H1). Findings indicated that self-efficacy during sole use was significantly lower for each additional year of age and significantly higher with higher levels of perceived usability. The second model tested whether self-efficacy would be positively associated with emotionally appealing design while controlling for covariates (H2). We found that self-efficacy was significantly lower with increasing age and higher with better health. Moreover, we observed a positive effect of design indicating that participants who rated the walkers as emotionally appealing were more likely to report higher levels of self-efficacy. The third model tested the interplay between usability and emotionally appealing design with respect to different levels of self-efficacy (H3). The model provided a significant interaction effect indicating that different levels of self-efficacy varied, depending on the combined effects of both variables. Against our expectations, the interaction was negative suggesting that the positive association between emotionally appealing design and self-efficacy was lower with higher levels of perceived usability, and vice versa. Figure 5 illustrates the interaction by showing the sample values of usability for which the effect of emotionally appealing design proved significant. We see that the association between emotionality and self-efficacy decreased with higher levels of perceived usability.

Table 5 Multiple regression models predicting product related self-efficacy (sole use) (N = 105)
Full size table
Fig. 5
figure 5

Adjusted effect of emotionally appealing design on self-efficacy across the range of chronological age with 95% confidence interval

Full size image

We performed four supplementary analyses. The first analysis explored bivariate correlations between self-efficacy, emotion, and usability with respect the exposure of the different walker models (see Appendices Table 6). We found that Model A was associated with lower positive emotion and lower usability. No significant correlations were found for the other demonstrators. The second analysis tested whether the effects of usability and emotion on product-related self-efficacy would differ depending on the walker model (see Appendices Table 7). Findings indicated a negative Usability × Model A interaction suggesting that the positive influence of usability on self-efficacy was lower for Model A. The third analysis tested whether the observed relationships would be different depending on walker experience (see Appendices Table 8). The respective interaction terms with walker experience did not reach significance. The fourth analysis explored the factorial structure of the RAPUUD and meCUE questionnaires to ensure that usability and positive emotion would reflect different constructs (see Appendices Table 9). The exploratory factor analysis with maximum likelihood estimation and oblimax rotation provided the expected two-factor solution. However, we observed significant cross-loadings of some of the RAPUUD items on the positive emotion factor.

5 Discussion

In the following, the three hypotheses are discussed within the results of the study. In terms of H1, correlation analysis showed a significant relationship between usability and self-efficacy. This means that higher levels of usability across all walkers were associated with higher levels of product-related self-efficacy beliefs. Regression analysis indicated that usability was a critical influencing factor to explain self-efficacy while controlling for sociodemographic covariates. Supplemental analyses showed that the effect of usability remained robust when controlling for the exposure to different walker models. However, we found a negative Usability × Model A interaction indicating that the effect of usability on self-efficacy was lower for Model A. In the specific use case of the walkers, H1 can thus be confirmed: Perceptions of better product usability are positively associated with the user’s perceived self-efficacy.

Regarding H2, correlation analysis confirmed a significant relationship between positive emotions and self-efficacy. Similar to usability, the participants reported a higher self-efficacy if more positive emotions were associated with the walkers. In addition, regression analysis provided evidence that positive emotions were fundamental to explain self-efficacy while controlling for sociodemographic covariates. Supplemental analyses indicated that this finding was robust when controlling for the potential impact of the different walker models. For the product demonstrator walker, H2 can therefore be confirmed: Perceptions of an emotionally appealing product design are positively associated with the user’s perceived self-efficacy.

In terms of H3, it was initially expected that a person’s self-efficacy is mostly improved if both usability and emotional appealingness are highly rated. However, the regression analysis that considered both aspects and their interaction effect showed that the positive influence of usability on self-efficacy decreases with higher positive emotions, and vice versa. This indicates that positive emotions may be able to compensate the relevance of usability in terms of product perception and shows the importance of an emotionally appealing product design. This is contrary to the individual opinions of the participants, who assessed usability to be more important for a product than an emotional design. Based on the current study design, valid causal relationships cannot be derived for these results. However, other studies that examined the sole correlation between perceived beauty/aesthetics and perceived usability suggest a halo effect as a possible explanation to such outcomes (Hassenzahl and Monk 2010; Tractinsky et al. 2000). Following this assumption, this means that the perceived positive emotions might overshadow minor usability inconveniences of the walkers and make them less noticeable to the user. Given the significant interaction effect between usability and positive emotions, hypothesis H3 cannot be confirmed.

Nonetheless, the affirmation of H1 and H2 shows the positive effects of usability and positive emotions on the participant’s self-efficacy. From the perspective of product development, both aspects should therefore be regarded as suitable starting points to empower the user’s self-efficacy. As mentioned in the introduction, stigma-sensitive products often show deficits in terms of an emotionally appealing product design. Such products often prioritise functionality, reliability and safety (McCreadie and Tinker 2005) or have a strong focus on ergonomic aspects, especially for products where the user has no choice to use it or not (e.g., wheelchair) (Bispo and Branco 2009). An emotionally appealing product design is neglected instead (Shinohara and Wobbrock 2011; Vaes 2019). From an objective point of view, this generally fits the participant’s assessment of the importance of usability and emotional appealingness (cf. chapter 5.1). Yet, the observed interaction effect implies that, in terms of increasing self-efficacy, emotions might be more (or at least equally) important than usability. The potential of emotional product design is thus rather underestimated. A stronger focus on designing for positive emotions may hereby not only lead to a higher self-efficacy but also an overall improved product acceptance [cf. e.g. (Jacobson 2014)]. It would thus be beneficial for product designers to become more aware of the user’s personal concerns about self-image and identity that results from using stigma-sensitive products. These are sometimes more important than improving issues of use such as usability (Kelly and Matthews 2014). This does not mean that no attention should be paid to usability. After all, a usable product is the basis for product usage. However, both usability and emotional design need to be considered together.

5.1 Limitations

Although significant correlations as well as interesting and sometimes surprising results were found for the product demonstrator walker, it cannot be claimed that these results are generalizable. The results currently have no external validity beyond walkers. Therefore, further investigations with different types of products are necessary in the future. This would additionally strengthen the claims of the study and provide the basis for a more representative study. Furthermore, the majority of the participants were older people that had no experience with walkers. Although this needs to be mentioned as a limitation, we submit that walker experience was statistically controlled in the regression models. Moreover, we believe that the inclusion of inexperienced participants contributes to a broader perspective to the studied phenomenon. That is because using a walker in the future may have special relevance in the anticipation of one’s own aging process. The number of participants is already quite good with over 100. Nevertheless, increasing this number would strengthen the statistical power and may enable further sociodemographic-specific analyses. Furthermore, evaluating the usability of a product solely via 2D images is always limited as the participants cannot fully experience product usage. Aspects such as rolling friction or weight, for example, can hardly be considered in such an evaluation. Subsequent examinations should thus also include a physical test component to assess a product in more detail. In particular, a visual assessment of ergonomic factors can be considered critical. However, it can be assumed that, particularly in the early stages of the product development process, e.g., during the concept phase, an initial assessment of ergonomics can also be made by means of an evaluation using images and descriptions. In later phases of the product development process, an ergonomic evaluation on a physical product is crucial. However, the conducted study and its results provide a sufficient initial assessment for product evaluation, as visual stimuli dominate, especially at the beginning (Fenko et al. 2010). We already claimed that self-efficacy is probably influenced by additional factors apart from the usability and emotionality of a product. It is also possible that the participants do not consider the two constructs of usability and emotionality completely separately, but instead evaluate their relative importance in relation to each other. Although selected user-specific influences were already taken into account in the conducted study, further covariates such as the user’s motivation should therefore be examined in follow-up studies. Another potential limitation pertains to the question whether positive emotion and usability reflect distinct or overlapping constructs. Although our supplementary factor analysis provided evidence for the expected two-factor structure, we observed significant cross-loadings from several RAPUUD items on the positive emotion factor. The reason for this may be a possible ambiguity of the analysed items. Particularly in terms of emotionality, personal perception is heavily dependent on the individual experiences and judgements of the users which cannot be directly derived from the data. Further studies that include the personal experiences of users in more detail, for example through personal interviews, could provide additional and meaningful insights. We also submit that such findings should be carefully interpreted given the small sample size in this study. However, future research should explore the relation between both constructs on the measurement level more specifically. Moreover, we note that all of our findings remained robust, when using only those RAPUUD items for scale construction that provided no cross-loadings. Finally, it should be noted that product-related evaluations and usage behaviours typically depend on the interplay between the person and the respective technological product (Wahl and Gerstorf 2018). With respect to this study, this means that expert ratings regarding the usability of the product variants do not necessarily reflect objectively true values. For example, Model A might be more usable for individuals with specific disabilities and limited mobility, because there is a clear affordance to stand within the frame of the walker. However, those users with a greater range of movement may perceive the model as too constrictive and less usable. Future research in this domain should focus more strongly on product evaluations depending on specific levels of physiological, cognitive, and motivational competence of users.

5.2 Practical implications

Despite the limitations and the need for further examinations, the conducted study already provides some useful insights to confirm a good usability and an emotionally appealing product design as beneficial starting points to increase the user’s self-efficacy. Thereby, the study results especially indicate an unused potential to strengthen the user’s self-efficacy by evoking positive emotions through product design. As emotions have only been given limited attention in the development of stigma-sensitive products so far, product developers need to raise their awareness towards this potential.

6 Conclusion

This paper aimed for examining whether self-efficacy would be associated with perceptions of a product’s usability and emotional appealingness. This is a necessary step towards confirming usability and emotional design as valid starting points for product development to strengthen the user’s self-efficacy and thus to empower a person to be more confident while facing product-related user stigma. Correlation and regression analyses verified usability and positive emotions to empower self-efficacy in the context of walkers. Thereby, a negative interaction effect between both aspects implies that, in terms of increasing self-efficacy, emotions are able to compensate the importance of usability. Yet, the potential that lies within designing for emotions often remains unused and should be focused on by product development stronger. Future research should further explore the interaction effect between usability and emotional design in the context of an increasing self-efficacy as well as assess the generalizability of the results by examining different product cases besides the stigma-sensitive walker. In the long term, the examination of the superordinate relationships between usability, emotional design and self-efficacy can be traced back to individual product characteristics. This could ultimately lead to concrete recommendations that foster a user’s self-efficacy through product design.