- Email: [email protected]
Unity enhances product aesthetics and emotion
International Journal of Industrial Ergonomics 59 (2017) 92e99
Contents lists available at ScienceDirect
International Journal of Industrial Ergonomics journal homepage: www.elsevier.com/locate/ergon
Unity enhances product aesthetics and emotion Wu Tyan-Yu a, *, Tsao Chueh-Yung b, Sie Cian-Yu a a b
Department of Industrial Design, Chang Gung University, No. 259, Wenhua 1st Rd, Guishan District, Taoyuan City 333, Taiwan Department of Industrial and Business Management, Chang Gung University, No. 259, Wenhua 1st Rd, Guishan District, Taoyuan City 333, Taiwan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 2 August 2016 Received in revised form 3 January 2017 Accepted 14 February 2017
Although the aesthetic properties of a product are likely associated with users' emotional responses, empirical evidence does not explain how the aesthetic properties of a product evoke an emotional response. This study presents the hypothesis that users' emotions are evoked when they observe an aesthetically pleasing product with unity. The results implied that a product form with more unity had a greater likelihood of affecting users' positive emotions compared with those with less unity, and that the unity aesthetic may act as a mediator in evoking emotion. In addition, the results confirmed that products composed of curvier elements tended to evoke a stronger pleasure response compared with those defined by straight lines. A systematic approach, namely the decision tree method, acts as a unity design guideline for the enhancement of product aesthetics, which may evoke users’ pleasure responses further. Relevance to industry: Our findings imply that a product form with more unity had a greater likelihood of affecting users’ positive emotions compared with those with less unity. In addition, a systematic approach, namely the decision tree method, acts as a unity design guideline for the enhancement of product aesthetics, which may apply for designing products with pleasures. © 2017 Elsevier B.V. All rights reserved.
Keywords: Aesthetics Unity Emotions
1. Introduction Unity, an aesthetic principle, is frequently adopted for enhancing a product's aesthetic features, and aesthetics may play a critical role in consumers' purchase decisions (Hirschman and Holbrook, 1982; McDonagh et al., 2002; Bloch et al., 2003; Blijlevens et al., 2009; Hoyer and Stokburger-Sauer, 2012). Moreover, the perception of a pleasingly designed product can provide users with sensory pleasure in terms of positive emotions (Hekkert, 2006; Chang and Wu, 2007; Kumar and Garg, 2010). For instance, a harmony chair may convey the aesthetic taste of its users (Dhar and Wertenbroch, 2000; Postrel, 2003; Hoyer and Stokburger-Sauer, 2012), and make users feel good and experience pleasure during use (Helander, 2003). Therefore, utilizing the unity principle when creating a product with an aesthetic quality is an essential design topic. By manipulating design elements (e.g. lines, surfaces and colours) and aesthetic principles (e.g. unity, contrast, balance and proportion), designers can create a product which can stimulate
* Corresponding author. E-mail address: [email protected] (W. Tyan-Yu). http://dx.doi.org/10.1016/j.ergon.2017.02.003 0169-8141/© 2017 Elsevier B.V. All rights reserved.
potential users’ visual aesthetic sensations (Lauer, 1979; Coates, 2003). To do so, designers must be familiar with the characteristics of design elements (e.g. the line element) and aesthetic principles (e.g. unity), and how these elements affect aesthetic quality. Despite this apparent relevance to product design features, the unity principle has been addressed to a lesser degree in affective design studies, specifically regarding aesthetics with respect to the role of unity in influencing emotional responses. Unity is defined as ‘congruity among the elements of a design which gives each element the appearance of belonging together; that is to say, there appears to be some visual connection beyond mere chance that has caused these elements to come together’ (Veryzer et al., 1998). Chairs were used as stimuli in our experiment because of the clear composition of each component, and their properties were assessed in an objective manner (Coates, 2003). This study examined the relationship among unity, aesthetics and users' emotional responses. We present a hypothesis, namely that user emotions are evoked naturally when users observe an aesthetic product manufactured with a congruent unity attribute.
W. Tyan-Yu et al. / International Journal of Industrial Ergonomics 59 (2017) 92e99
2. Literature review 2.1. Aesthetics and consumers’ pleasure responses Aesthetics, derived from the Greek word aisthetikos, indicates an aesthetic or something sensitive or sentient (i.e. ‘I perceive, feel and sense’) (Chignell and Costelloe, 2011). The Oxford English Dictionary defines aesthetics as ‘a set of principles concerned with the nature and appreciation of beauty.’ Aesthetic principles which broadly include repetition, gradation, symmetry, harmony, balance, contrast, proportion, rhythm, dynamics, unity and simplicity are frequently used in product design (Graves, 1941; Kim, 2006). By manipulating product elements to assume a unified order, the aesthetic qualities of a product can be enhanced, thereby attracting users' attention (Norman, 2004). For instance, Bell et al. (1991) utilised colour photographs of living room furniture to test the unity of styles, and found that the aesthetic response was correlated with the perceived unity of a style. Lennon (1990) reported that a model with matching clothes and accessories appeared more competent and desirable compared with a model with nonmatching garments and accessories. Prior studies have shown that consumers respond more positively to products exhibiting a high degree of unity compared with those exhibiting a low degree of unity. Therefore, the study of unity is a critical issue in designing a product for aesthetic pleasure. Consumers’ emotions may be also affected by line-style elements (Coates, 2003) which visually comprise a product with an aesthetic quality. Bar and Neta (2006) indicated that people prefer objects defined by curved lines over those with sharp angles. A chair composed of curved lines appears to be more likely to please users compared with one composed of straight or slanted lines. For example, the Pingo chair designed by Hans Jakobsen in 2000 was devised specifically with curved elements in accordance with the shape of the back, seat, and legs, resulting in an elegant and aesthetically pleasing object. On the basis of the prior research, we assume that chair aesthetics is associated with the assembly of component attributes and line-type elements. A chair's back, seat, and legs have a clear relationship. Thus, the basic components (i.e. seat, back and legs) have frequently been used as stimuli in experiments (Jindo et al., 1995; Park and Han, 2004). The line attributes of each component were often classified by previous researcher as square, round, and slanted for the back; the same elements were applied to the seat; and straight, curved, and slanted lines were designated for the legs. In this study, we argue that the unity of a chair may influence its aesthetic quality, which in turn affects users' emotions (Fig. 1). In other words, a higher degree of unity is equal to a higher aesthetic, and in turn evokes a stronger emotional response in the user mainly (M1 & M2). Furthermore, we examined the causal relationships between the aesthetic and its pleasure reactions (M2), and also to test whether unity had a direct effect on pleasure (M3). 2.2. Gestalt laws and unity In perceiving a product's appearance, our sensory systems detect order in chaos or unity in variety (Hekkert, 2006) when
Fig. 1. Conceptual research framework.
93
examining a group of elements. People have a tendency to view objects such as a chair's structural arrangement as close together or feeling as if belonging together, according to Gestalt laws. Gestalt laws are summarised as ‘the whole is more than the sum of its € hler, 1920, pp.17). However, previous studies on Gestalt parts’ (Ko laws have focused on graphical patterns in dynamic graph drawings (Nesbitt and Friedrich, 2002), instructional screen design (Smith-Gratto and Fisher, 1999) and even the haptic and visual grouping of elements (Chang et al., 2007). Specifically, for the present study, we used Gestalt laws to explain chair shapes in terms of a chair's unity (i.e. line attribute arrangement). The principles of proximity, similarity, and continuation were adopted to explain how it is possible to perceive a chair's unity through its compositional arrangement. The proximity principle is defined as elements that are located closest to each other tending to form a group (Wertheimer, 1923). In Fig. 2, Lines 1 and 2 are close together and are viewed as a single group, whereas Lines 3 and 4 are regarded as separate. In this case, viewers may attempt to perceive both the chair's back and seat as one proximity piece (e.g. , , and ) because of the closed array between the two components. According to the similarity principle, similar elements tend to form groups; users regard similar elements (e.g. having the same characteristics) as belonging together, and perceive them to be as such on the basis of the attributes (e.g. line, shape, colour, texture, value, volume, and orientation) (Wertheimer, 1923; Chang et al., 2007). For instance, two groups of heavy black lines and two groups of light lines are arranged in straight lines (Fig. 3). Heavy lines are grouped as one, whereas light lines are grouped separately. Moreover, a chair composed of three types of lines (e.g. , , and ) may appear to be completely inconsistent, thus exhibiting no unity at all. Continuation concerns the eye seeking relationships between shapes. This occurs when the eye follows a line, a curve, or a sequence of shapes, even when it crosses over negative and positive shapes (Graham, 2008). As shown in Fig. 4, the human eye tends to perceive the ‘ ’ as a completed ‘S’. On a verticle plane, the back and legs of a chair (e.g. , , and ) may convey a continuous shape which tends to appear to have a greater degree of unity to the senses. In this study, the chair arrangement of each component (i.e. back, seat, and legs) was controlled and arrayed into every possible combination which could be associated with Gestalt laws. Participants may thus perceive the chair shape by subconsciously applying Gestalt laws.
3. Methods 3.1. Visual stimuli The chair was adopted as a stimulus in this experiment because of its widespread use and special structure array, indicating a clear composition on each component. To control the experiment, we used the chair archetype composed of three distinct components: the back, seat, and four legs. Chair arms were omitted because of research limitations concerning experimental complexity. Stimuli, including square, round, and slanted shapes, were employed as representative elements, resulting in 27 stimulus combinations (Fig. 5). Component shape was used as a factor to facilitate unity variation. The following section details how we varied and combined the components to reflect unity as a variable.
Fig. 2. Illustrated proximity.
94
W. Tyan-Yu et al. / International Journal of Industrial Ergonomics 59 (2017) 92e99
arrangements and (2) the three types of line attributes (straight, curved, and slanted). The following five groups were included in the first category: one group of chairs with three similar attribute components; three groups with at least two similar attribute components; and one group with no similar attribute components (Fig. 5). By contrast, four groups were included in the second category: chairs with three straight lines; two straight lines, and one curved line; one straight line and two curved lines; and three curved lines (Fig. 6).
Fig. 3. Illustrated similarity.
3.3. Emotional assessments The pleasure arousal dominance (PAD) scale, developed by Mehrabian and Russell (1974), is a semantic differential scale that includes the dimensions of pleasure, arousal, and dominance. Pleasure can be defined as a positive affective state that is distinct from feelings (e.g. preference and reinforcement). Arousal is an emotional state progressing from sleepiness to intense excitement. The third dimension, dominance, refers to the extent to which a person feels either positively or negatively towards an environment (Mehrabian and Russell, 1974). A set of bipolar adjective pairs was used for each item question, and it was rated on a 9-point scale. The PAD scale is widely used in environmental psychology as well as in product and interface design. Dominance was omitted for validity purposes (Russell and Pratt, 1980; Russell et al., 1981). Only the pleasure and arousal dimensions were adopted, and the stimuli are presented as 2D images in this study. Six emotional adjectives were used for each dimension. Additional items, namely ‘perceived unity’ and ‘aesthetics’, were included in the questionnaire. A 5point Likert scale was used for each questionnaire item.
Fig. 4. Illustrated continuation.
3.4. Participants
Fig. 5. Chair classification based on component attributes.
Fifty-four students (men ¼ 23, women ¼ 31, average age ¼ 21 years) were recruited from universities in Taiwan for the experiment. Of the participants, 27 majored in furniture design at the National Taipei University of Technology, and the other 27 majored in product design at Chang Gung University. To ensure balanced participation, 27 of the students majored in furniture design had sufficient aesthetic experience and were able to perceive line drawing chairs, whereas the other 27 first-year students had limited experience in aesthetic training.
3.2. Chair unity variable
4. Results and analysis
Chair unity was regarded as the independent variable, whereas ‘perceived unity’ (assessed through questionnaire items), aesthetics, pleasure, and arousal were considered the dependent variables. To test unity, the stimuli were classified according to (1) the three types of component attribute (back, seat, and legs)
A pretest was conducted to examine the validity of the PAD scale. Factor analysis confirmed the two emotional dimensions of pleasure and arousal. The KMO measure of sampling adequacy was 0.885, above the recommended value of 0.5, and Bartlett's test of sphericity indicated significance (p < 0.001). Internal consistency
Fig. 6. Chair classification based on line attributes.
W. Tyan-Yu et al. / International Journal of Industrial Ergonomics 59 (2017) 92e99 Table 1 Relationship between independent and dependent variables. Variables
Perceived unity
Aesthetics
Pleasure
Arousal
Chair unity Perceived unity Aesthetics Pleasure
0.433***
0.205*** 0.492***
0.145*** 0.359*** 0.621***
-0.009 0.053 0.094*** 0.112***
***p < 0.01, **p < 0.05, *p < 0.1.
was tested with Cronbach's alpha (0.925 for pleasure and 0.867 for arousal). The statistical results confirmed the reliability of the PAD scale. 4.1. Independence and dependence tests We regarded chair unity as the independent variable, and perceived chair unity, aesthetics, pleasure, and arousal as the dependent variables. The results of Pearson's correlation test revealed highly positive correlations between perceived unity and chair unity (Table 1). Perceived unity, aesthetics, and pleasure were strongly and positively correlated. However, arousal was weakly correlated to both chair unity and perceived unity. 4.2. Unity on ‘component attributes’ As shown in Table 2, the ANOVA results for a randomised block design revealed significant differences between chair unity and perceived unity, aesthetics and pleasure; however, such significant differences were not found between chair unity and arousal. Furthermore, the Scheffe test results revealed that the participants could sense a high degree of unity when they perceived the chair's back, seat, and legs as consistent (e.g.), but a low degree of unity when they perceived the chair's back, seat, and legs as individually distinct (e.g.). In other words, a chair composed of similar component attributes is likely to convey a higher degree of unity, whereas a chair composed of completely different component attributes is likely to convey a lower degree of unity. The chair composed of two
95
(out of three) similar attribute components exhibited less unity compared with those with three of the same attribute components; the unity ranking for the chairs was as follows: chairs with the same attributes for the back and seat > back and legs > seat and legs. The Scheffe test results consistently showed that the participants perceived a high aesthetic quality when they observed the chair's back, seat, and legs as consistent, but a low aesthetic quality when they observed the chair's back, seat, and legs as individually distinct attributes. Furthermore, the results revealed a weaker aesthetic quality for chairs with two similar components than for those that had similar attributes for all the components. The aesthetic ranking of the chairs was as follows: a chair with the same attributes for the back and seat > back and legs > seat and legs. Similarly, as shown in Table 2, the Scheffe test results also revealed that the participants felt a high degree of pleasure when they perceived the chair's back, seat, and legs as a consistent attribute as well as the back and seat. A lower degree of pleasure was found with a lower degree of unity when they perceived the chair's back, seat, and legs as inconsistent and the seat and legs as consistent. Overall, a greater unity between the chair components led to a greater sense of pleasure evoked in the participants. However, no significant differences were observed in the effect of different degrees of unity on arousal. We suspect that the images used in this experiment, because of their flat and simple lines, were visually insufficient to evoke a strong sense of arousal. 4.3. Unity on ‘line attributes’ Bar and Neta (2006) indicated that line attributes may have considerable influence on consumers' aesthetic perception and emotions. To understand how chair line attributes can affect consumers’ sense of unity, aesthetic sense, and emotions, in the experiment, we further examined the effects of two types of lines, namely straight (including straight and slanted lines) and curved lines (Fig. 6), on unity, aesthetic, pleasure, and arousal. As shown in Table 3, the ANOVA results revealed highly significant differences
Table 2 ANOVA (randomised block design) for ‘component unity’ influencing perceived unity, aesthetic, pleasure and arousal.
Perceived unity Aesthetic Pleasure Arousal
BG WG BG WG BG WG BG WG
Sum of Squares
df
Mean Square
F
Sig.
Ranking (Scheffe test)
403.398 1324.157 57.529 1214.915 24.398 847.825 0.745 832.901
4 1400 4 1400 4 1400 4 1400
100.850 0.946 14.382 0.868 6.099 0.606 0.186 0.595
106.626
0.000
back ¼ seat ¼ leg > back ¼ seat > back ¼ leg > seat ¼ leg > Back s seat s leg
16.573
0.000
back ¼ seat ¼ leg, back ¼ seat > back ¼ leg (seat ¼ leg) > Back s seat s leg
10.072
0.000
back ¼ seat, back ¼ seat ¼ leg > back ¼ leg, seat ¼ leg, Back s seat s leg
0.313
0.869
Note: Between groups ¼ BG, Within groups ¼ WG; ‘ ¼ ‘ means the same attribute.
Table 3 ANOVA (randomised block design) for ‘line unity’ affecting perceived unity, aesthetic, pleasure, and arousal.
Perceived unity Aesthetic Pleasure Arousal
BG WG BG WG BG WG BG WG
Sum of Squares
df
Mean Square
F
32.311 1407.245 36.992 1235.453 38.337 833.885 25.870 807.776
3 1401 3 1401 3 1401 3 1401
106.770 1.004 12.331 0.882 12.779 0.595 8.623 0.607
106.296 0.000 all curved lines > all straight lines > two curved and one straight lines > one curved and two straight lines 13.983 0.000 all curved lines > all straight lines, two curved and one straight lines > one curved and two straight lines 21.470 0.000 all curved lines > two curved and one straight lines, all straight lines, one curved and two straight lines 14.956 0.000 all straight lines > one curved and two straight lines, two curved and one straight lines, all curved lines
Note: Between groups ¼ BG, Within groups ¼ WG.
Sig.
Ranking (Scheffe test)
96
W. Tyan-Yu et al. / International Journal of Industrial Ergonomics 59 (2017) 92e99
between chair unity and perceived unity, aesthetics, pleasure, and arousal. Table 3 shows that the chairs composed of only curved lines (e.g. back, seat, and legs) and those composed of only straight lines led to a high degree of unity. In other words, the participants perceived that the two types of chairs composed of the same line attributes conveyed a significantly high degree of unity. By contrast, chairs composed of either two curved or two straight lines had less unity compared with those composed of three similar attributes. Overall, the chairs composed of more similar line attributes yielded a higher degree of unity compared with those with less similar attributes. Similarly, we found that a chair composed of a greater number of similar line attributes also had a higher likelihood of conveying a higher aesthetic quality compared with that with less similar line attributes. The chairs with two curved lines and one straight line evoked a stronger aesthetic response compared with the chairs with one curved and two straight lines. The result indicated that the participants noted that the chairs with more curved lines can evoke a stronger aesthetic response than can chairs with less curved lines. These results implied that increasing the number of curved lines consistently may enhance aesthetic quality. As mentioned, a chair composed of only curved lines was observed to evoke a strong sense of pleasure. By contrast, a chair composed of only straight lines was found to evoke less pleasure than one with all three components curved. This chair sample was followed by another with three curved features, and afterward, a chair with two curved features which evoked a less intense pleasure response. The results implied that consumers' pleasure response may be affected by the number of curved attributes in a chair. A greater number of curves in the chair appeared to yield a higher likelihood of evoking a stronger pleasure response compared with straight lines. However, Table 3 shows that chairs composed of only straight lines (i.e. a high degree of unity) generated greater arousal compared with those composed of curved lines. This result implied that a chair composed of more straight lines has considerable potential in evoking consumers’ arousal. Conversely, curved lines may reduce the intensity of arousal. 4.4. Unity of aesthetic evoking users’ emotions Path analysis was performed to examine the affective relationships among unity, aesthetic and pleasure reactions (Fig. 1). Path analysis involved three regression models. The first regression model (M1) was used to verify the causal relationships between the unity of a chair and its aesthetic quality. By controlling for participant differences, we found that aesthetic, as the dependent variable, was significantly influenced by the independent variables of unity (Table 4a), with significance at 1%. The second regression model (M2) was used to examine the causal relationships between the aesthetic and its pleasure reactions. Table 4b shows that the relationship was supported at a significance of 1%. The third regression model (M3) was used to test whether unity had a direct effect on pleasure. Table 4c shows that the direct effect was significant at 10%. The overall effect of unity on pleasure was 0.141 (0.406 0.285 þ .025). The result implied that unity also directly Table 4a Regression analysis of the effect of aesthetic on unity (M1) (dependent variance ¼ aesthetic). Variable
Parameter
Std. Error
T
Sig.
Constant Perceived- Unity
1.951*** 0.406***
0.174 0.020
11.207 20.116
0.000 0.000
R2 ¼ 0.387 (R2adj ¼ 0.364); ***p < 0.01, **p < 0.05, *p < 0.1.
Table 4b Regression analysis results of the effect of pleasure on aesthetic (M2) (dependent variance ¼ pleasure). Variable
Parameter
Std. Error
T
Sig.
Constant Aesthetic
2.546*** 0.298***
0.109 0.014
23.359 20.804
0.000 0.000
R2 ¼ 0.409 (R2adj ¼ 0.387); ***p < 0.01, **p < 0.05, *p < 0.1.
Table 4c Regression analysis of the effect of pleasure on aesthetic and unity (M3) (dependent variance ¼ pleasure). Variable
Parameter
Std. Error
T
Sig.
Constant Aesthetic Perceived- Unity
2.511*** 0.285*** 0.025*
0.111 0.016 0.014
22.677 17.502 1.775
0.000 0.000 0.076
R2 ¼ 0.411 (R2adj ¼ 0.388); ***p < 0.01, **p < 0.05, *p < 0.1.
affected consumers' sense of pleasure. In summary, path analysis revealed that both the indirect and direct effects of unity on pleasure were significant. A higher degree of unity resulted in a greater extent of aesthetic responses, and the participants’ sense of pleasure was likely to have increased considerably as well. 4.5. Decision tree analysis on unity The statistical results showed that chair unity had a considerable influence on aesthetics and furthered users' pleasure responses. Thus, using the unity principle as a vehicle for enhancing a product's aesthetic quality can be an effective strategy that benefits designers in the development of a product meant to evoke pleasure. This study proposes a unity-based method which can guide designers in manipulating components affectively and effectively during form development. To devise this systematic method, we examined how our participants perceived the unity of each stimulus (i.e. 27 chairs) by performing decision tree analysis with the MATLAB 8.2 package, which was used to analyse the trends of a small amount of data without meaning connected, in which the decision tree allows the identification of the optimal arrangement of unity with the composition of the three components (i.e. back, seat, and legs). A decision tree is composed of branch and leaf nodes, and each branch node contains an if/then rule. The decision process starts from the top branch node which is called ‘root’ in the decision tree literature. When the ‘if’ condition is satisfied, for example, the leg component is assigned to a straight line, and the process progresses to its consequent node; otherwise, it advances to another node. Fig. 7 displays the results of a decision tree in which three trails derived from straight, curved, and slanted legs were assigned. The bottom degree of each trail displayed optimal as well as poor samples. For instance, with a high value factor (0.35), the right branch under the straight legs suggested four chair samples which enhanced the high degree of unity. The other results appeared under the curved and slanted legs; the curved legs, with the second-highest value factor (0.55), suggested two optimal samples; and the slanted legs, with the third-highest value factor (0.71), suggested three optimal samples. Overall, tree analysis revealed a valuable direction for designing a chair with a significant degree of unity which can enhance the chair's aesthetic quality and further evoke a considerable sense of pleasure in users. For future application, the decision tree method (DTM) can be integrated with the concept of a database system which can potentially allow designers to access the system freely during the form development stage.
W. Tyan-Yu et al. / International Journal of Industrial Ergonomics 59 (2017) 92e99
97
Fig. 7. Result of the decision tree classifier, applied to the chair components.
5. Discussion 5.1. Unity, aesthetic, and emotion effects The empirical results revealed the significant effect on the various extents of unity arrangement. Specifically, a chair with all three attribute components that were similar was likely to convey the highest degree of unity, whereas a chair with all three attribute components that were different was likely to convey the lowest degree of unity (Table 2). Because of the differences in component arrangement, chairs with two similar components may have a different degree of unity and aesthetic responses. The participants' perceptions were likely influenced by how they viewed ‘two component arrangements’ as being together. Accordingly, each assembled component had a special relationship with another, with the Gestalt laws able to provide a reasonable explanation for these visual relationships. For instance, the proximity and continuation principles explain why the participants perceived greater unity in a chair when its back and seat had consistent attributes (Fig. 5, fourth row) compared with the unity of a chair with back and leg consistency (Fig. 5, third row) or seat and leg consistency (Fig. 5, second row). In this seat and leg consistency case, proximity explains the unity sensation when a closed position between the back and seat as a single attribute is perceived. In addition, continuation supports the visual connection between the chair's back and seat as a unit, whereas two components posited next to each other demonstrate similar attributes that are connected. Thus, the chair's back and seat as a single element are supported both by proximity and continuation, which enhances the chair form as a whole, making participants perceive the unit sensation clearly. Moreover, with a chair's back and legs as consistent attributes, proximity explains the unity sensation when someone perceives the back and legs as a respectively upper and lower relationship on a vertical orientation as wholeness. Conversely, the chair with seat and leg consistency exhibited less unity because of the lack of a clear visual orientation or any connection whatsoever. Similarly, the results of the aesthetic and pleasure responses also revealed that the participants' greatest responses on aesthetic and pleasure were evoked when they perceived the three attributes of a chair to be consistent (Table 2); conversely, the chair with three components having all different attributes evoked the lowest degree of aesthetic and pleasure responses. As shown consistently by the unity results, the participants presented a similar ranking of aesthetic and pleasure responses when they perceived a chair to
have two consistent attributes. The results further confirmed that unity is a critical design principle for enhancing a product's aesthetic quality. However, according to previous research, the line attributes of a product may also influence the aesthetic and pleasure responses. The examination of line attributes further indicated that the participants' sense of pleasure was also affected by the types of line elements. For instance, a chair composed of more curves appeared to have a greater likelihood of evoking a stronger pleasure response compared with one composed of straight lines (Table 3). Specifically, the chair whose attributes all have curved lines exhibited a high degree of unity which also consistently evoked the greatest sense of pleasure. The results also revealed that the chair composed of straight lines for all its attributes exhibited a high degree of unity but evoked a lower degree of pleasure compared with the chair with all three curved attributes. Chairs with three curved attributes may evoke the strongest aesthetic and pleasure responses. Conversely, chairs with two straight and one curved line attribute evoked the weakest aesthetic and pleasure responses. Moreover, chairs with two curved and one straight line evoked a stronger pleasure response compared with those with all straight lines, implying that curved-line elements have a greater likelihood of dominating consumers' emotional responses compared with straight lines. Designers should take advantage of curved elements for the enhancement of the pleasure response. These results implied that a product composed of more curved elements may increase the sense of pleasure more than the same product composed of straight lines may. The results confirmed the finding of Bar and Neta (2006) that people prefer objects with curved elements to those with straight lines and sharp angles. This results also confirmed that chairs with curved or rounded lines appear more attractive and perhaps even convey a greater sense of warmth compared with other chairs (Park and Han, 2004), whereas straight lines convey rigidity and sharpness. Overall, these results implied that a product composed of curved lines and with a high sense of unity should have a greater likelihood of evoking consumers’ pleasure responses compared with those with less unity and more straight lines.
5.2. Unity enhances aesthetic and pleasure The results revealed that aesthetic response was strengthened by an increased perception of unity and weakened by a reduced degree of unity. Tables 4aeb shows that the aesthetic responses increased with high-unity forms, and users' sense of pleasure was
98
W. Tyan-Yu et al. / International Journal of Industrial Ergonomics 59 (2017) 92e99
Fig. 8. a) Meditation chair, b) New Ming's chair, c) Devil's chair, d) Trapezoid chair.
evoked following the increase of the aesthetic response. These findings are consistent with research on people's preference for objects that exhibit harmonious traits (Hekkert, 2006; Kumar and Garg, 2010); unity provides a harmonious characteristic. This result also supported the findings of Coates (2003); that is, as a single variable, unity can enhance aesthetic quality. Accordingly, a product with attributes consistency is always implied to be an optimal choice for the enhancement of chair unity for strengthening consumers' aesthetic and pleasure responses.
5.3. Application of the unity principle in design According to the unity experiment and decision tree analysis results, we recommend an effective design process for the development of products with unity which can further enhance a product's aesthetic and pleasure responses. The processes were developed on the basis of four aspects: (1) defining the product image and line style, (2) using DTM to identify the optimal unity of an object, (3) enhancing the aesthetic quality, and (4) rationalising the design. The first step involves ‘defining the product image and line style’ which concerns identifying the type of product image (e.g. line style) required for a particular project or one that satisfies a user's requirements. In actuality, the product images or styles should fit into users' lifestyle or personality. For instance, to create a chair with floral elements intended to satisfy a user who prefers to live in a floral and elegant space. Designers may choose curved elements as a basic component for developing a chair to match the similar interior style. After defining the line style, the second step involves reviewing the data with the DTM and identifying an optimal sample as a reference for ideation. In this step, the data from the decision tree diagram enable designers to review optimal concepts and helps them determine the design direction so that they can create an ideal form quickly. By doing a trial-and-error, designers have the opportunity to test different shapes based on various data from the decision tree diagram and to determine the ideal result. The third step concerns how aesthetic qualities are pleasing to people; to achieve the desired results, designers are also suggested to incorporate aesthetic principles (e.g., balance, contrast, and visual weight) and Gestalt laws when finalizing their product forms. The final step considers the details of engineering and manufacturing applications. Here, designers should rationalise the final designed product in terms of object structure, material strength, and tooling. To test this design process, a 9-week chair workshop was held at Chang Gung University. At the end of the workshop, four chairs (Fig. 8a, b, c, d) were constructed and recommended by five faculty members and three professional designers for display in the Young Designers Exhibition, a well-recognised exhibition at Taipei World Trade Center. Overall, the participants from the workshop agreed that this design process is both effective and affective for the development of aesthetically pleasing chairs.
6. Future research Follow the emerging trend of CAD design, we suggest integrating both the DTM and unity principles within a database in future studies. This system will allow designers to simulate 3D results quickly when developing an aesthetic product that enhances unity. Moreover, the system should encompass three parts: a components databank, unity aesthetic principles, and DTM tools. The component databank should include several chair parts (e.g., legs, backs, and seats), which will allow designers to retrieve and create 3D forms quickly according to their design style. A DTM system will provide further optimising suggestions, by displaying 3D aesthetic forms on the computer screen immediately; 3D forms that prompt users will allow designers to freely review and manipulate the design forms. Finally, the system perhaps should also include more aesthetic functions, such as proportion, visual weight, and colour parameters to enhance aesthetic quality. Notably, however, these concepts must be further tested before they can be used for designing purposes. 7. Conclusion In this study, the unity principle plays an effective and affective role in the enhancement of the product aesthetic and consumers' sense of pleasure. The decision tree method may benefit designers seeking to develop a product with aesthetic effectively, which further evokes consumers’ positive emotions. However, designers should not limit themselves to employing only the unity principle when other principles (e.g. balance, colour, and material) can also have great potential in affecting the final design. Because of the research limitations, this study did not cover all types of products and consumers which required having further test in the future research. References Bar, M., Neta, M., 2006. Humans prefer curved visual objects. Psychol. Sci. 17 (8), 645e648. Bell, S.S., Holbrook, M.B., Solomon, M.R., 1991. Combining esthetic and social value to explain preferences for product styles with the incorporation of personality and ensemble effects. J. Soc. Behav. Personal. 243e273. Blijlevens, J., Creusen, M.E., Schoormans, J.P., 2009. How consumers perceive product appearance: the identification of three product appearance attributes. Int. J. Des. 3 (3), 27e35. Bloch, P.H., Brunel, F.F., Arnold, T.J., 2003. Individual differences in the centrality of visual product aesthetics: concept and measurement. J. Consum. Res. 29 (4), 551e565. Chang, D., Nesbitt, K.V., Wilkins, K., 2007. The Gestalt principles of similarity and proximity apply to both the haptic and visual grouping of elements. In: Proc. Eighth Australasian User Interface Conference (AUIC2007) (Ballarat, Australia). Chang, W., Wu, T.Y., 2007. Exploring types and characteristics of product forms. Int. J. Des. 1 (1), 3e14. Chignell, A., Costelloe, T.M., 2011. A dialogue concerning aesthetics and apolaustics. J. Scott. Philos. 9 (1), 5e16. Coates, D., 2003. Watches Tell More than Time. McGraw-Hill, New York. Dhar, R., Wertenbroch, K., 2000. Consumer choice between hedonic and utilitarian goods. J. Mark. Res. 60e71. Graves, M.E., 1941. The Art of Color and Design. McGraw-Hill, New York. Graham, L., 2008. Gestalt theory in interactive media design. J. Humanit. Soc. Sci. 2 (1).
W. Tyan-Yu et al. / International Journal of Industrial Ergonomics 59 (2017) 92e99 Hekkert, P., 2006. Design aesthetics: principles of pleasure in design. Psychol. Sci. 48 (2), 157e172. Helander, M.G., 2003. Forget about ergonomics in chair design? Focus on aesthetics and comfort! Ergonomics 46 (13e14), 1306e1319. Hirschman, E.C., Holbrook, M.B., 1982. Hedonic consumption: emerging concepts, methods and propositions. J. Mark. 92e101. Hoyer, W.D., Stokburger-Sauer, N.E., 2012. The role of aesthetic taste in consumer behavior. J. Acad. Mark. Sci. 40 (1), 167e180. Jindo, T., Hirasago, K., Nagamachi, M., 1995. Development of a design support system for office chairs using 3-D graphics. Int. J. Ind. Ergon. 15 (1), 49e62. Kim, N., 2006. A history of design theory in art education. J. Aesthet. Educ. 40 (2), 12e28. €hler, W., 1920. Physical Gestalten (Die Physischen Gestalten in Ruhe und im Ko stationaren Zustand). Erlangen: eine naturphilosophische Untersuchung: ix259. In: Ellis, W.D. (Ed.), A Sourcebook of Gestalt Psychology. Routledge & Kegan Paul, London, pp. 17e54. Kumar, M., Garg, N., 2010. Aesthetic principles and cognitive emotion appraisals: how much of the beauty lies in the eye of the beholder? J. Consum. Psychol. 20 (4), 485e494. Lennon, S.J., 1990. Effects of clothing attractiveness on perceptions. Home Econ. Res. J. 18 (4), 303e310. Lauer, A.D., 1979. Design Basics. Holt, Rinehart. New York: Holt, Rinehart, & Winston, New York. McDonagh, D., Bruseberg, A., Haslam, C., 2002. Visual product evaluation: exploring
99
users' emotional relationships with products. Appl. Ergon. 33 (3), 231e240. Mehrabian, A., Russell, J.A., 1974. An Approach to Environmental Psychology. the MIT Press. Nesbitt, K.V., Friedrich, C., 2002. Applying Gestalt principles to animated visualizations of network data. In: Proc. IV 2002. IEEE ComputerSociety Press, London, UK, pp. 10e12. Norman, D.A., 2004. Emotional Design: Why We Love (Or Hate) Everyday Things. Basic Books, New York. Park, J., Han, S.H., 2004. A fuzzy rule-based approach to modeling affective user satisfaction towards office chair design. Int. J. Ind. Ergon. 34 (1), 31e47. Postrel, V., 2003. The Substance of Style. HarperCollons, New York. Russell, J.A., Pratt, G., 1980. Affective quality attributed to environments a factor analytic study. Enviro description of the affective quality attributed to environments. J. Personal. Soc. Psychol. 38 (2), 311. Russell, J.A., Ward, L.M., Pratt, G., 1981. Affective quality attributed to environments: a factor-analytic study. Environ. Behav. 13, 259e288. Smith-Gratto, K., Fisher, M., 1999. Gestalt theory: a foundation for instructional screen design. J. Instr. Technol. Syst. 27 (4), 361e371. Veryzer, J., Robert, W., Hutchinson, J.W., 1998. The influence of unity and prototypicality on aesthetic responses to new product designs. J. Consum. Res. 24 (4), 374e385. Wertheimer, M., 1923. Untersuchungen zur Lehre von der Gestalt. II. Psychol. Res. 4 (1), 301e350.
Contents lists available at ScienceDirect
International Journal of Industrial Ergonomics journal homepage: www.elsevier.com/locate/ergon
Unity enhances product aesthetics and emotion Wu Tyan-Yu a, *, Tsao Chueh-Yung b, Sie Cian-Yu a a b
Department of Industrial Design, Chang Gung University, No. 259, Wenhua 1st Rd, Guishan District, Taoyuan City 333, Taiwan Department of Industrial and Business Management, Chang Gung University, No. 259, Wenhua 1st Rd, Guishan District, Taoyuan City 333, Taiwan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 2 August 2016 Received in revised form 3 January 2017 Accepted 14 February 2017
Although the aesthetic properties of a product are likely associated with users' emotional responses, empirical evidence does not explain how the aesthetic properties of a product evoke an emotional response. This study presents the hypothesis that users' emotions are evoked when they observe an aesthetically pleasing product with unity. The results implied that a product form with more unity had a greater likelihood of affecting users' positive emotions compared with those with less unity, and that the unity aesthetic may act as a mediator in evoking emotion. In addition, the results confirmed that products composed of curvier elements tended to evoke a stronger pleasure response compared with those defined by straight lines. A systematic approach, namely the decision tree method, acts as a unity design guideline for the enhancement of product aesthetics, which may evoke users’ pleasure responses further. Relevance to industry: Our findings imply that a product form with more unity had a greater likelihood of affecting users’ positive emotions compared with those with less unity. In addition, a systematic approach, namely the decision tree method, acts as a unity design guideline for the enhancement of product aesthetics, which may apply for designing products with pleasures. © 2017 Elsevier B.V. All rights reserved.
Keywords: Aesthetics Unity Emotions
1. Introduction Unity, an aesthetic principle, is frequently adopted for enhancing a product's aesthetic features, and aesthetics may play a critical role in consumers' purchase decisions (Hirschman and Holbrook, 1982; McDonagh et al., 2002; Bloch et al., 2003; Blijlevens et al., 2009; Hoyer and Stokburger-Sauer, 2012). Moreover, the perception of a pleasingly designed product can provide users with sensory pleasure in terms of positive emotions (Hekkert, 2006; Chang and Wu, 2007; Kumar and Garg, 2010). For instance, a harmony chair may convey the aesthetic taste of its users (Dhar and Wertenbroch, 2000; Postrel, 2003; Hoyer and Stokburger-Sauer, 2012), and make users feel good and experience pleasure during use (Helander, 2003). Therefore, utilizing the unity principle when creating a product with an aesthetic quality is an essential design topic. By manipulating design elements (e.g. lines, surfaces and colours) and aesthetic principles (e.g. unity, contrast, balance and proportion), designers can create a product which can stimulate
* Corresponding author. E-mail address: [email protected] (W. Tyan-Yu). http://dx.doi.org/10.1016/j.ergon.2017.02.003 0169-8141/© 2017 Elsevier B.V. All rights reserved.
potential users’ visual aesthetic sensations (Lauer, 1979; Coates, 2003). To do so, designers must be familiar with the characteristics of design elements (e.g. the line element) and aesthetic principles (e.g. unity), and how these elements affect aesthetic quality. Despite this apparent relevance to product design features, the unity principle has been addressed to a lesser degree in affective design studies, specifically regarding aesthetics with respect to the role of unity in influencing emotional responses. Unity is defined as ‘congruity among the elements of a design which gives each element the appearance of belonging together; that is to say, there appears to be some visual connection beyond mere chance that has caused these elements to come together’ (Veryzer et al., 1998). Chairs were used as stimuli in our experiment because of the clear composition of each component, and their properties were assessed in an objective manner (Coates, 2003). This study examined the relationship among unity, aesthetics and users' emotional responses. We present a hypothesis, namely that user emotions are evoked naturally when users observe an aesthetic product manufactured with a congruent unity attribute.
W. Tyan-Yu et al. / International Journal of Industrial Ergonomics 59 (2017) 92e99
2. Literature review 2.1. Aesthetics and consumers’ pleasure responses Aesthetics, derived from the Greek word aisthetikos, indicates an aesthetic or something sensitive or sentient (i.e. ‘I perceive, feel and sense’) (Chignell and Costelloe, 2011). The Oxford English Dictionary defines aesthetics as ‘a set of principles concerned with the nature and appreciation of beauty.’ Aesthetic principles which broadly include repetition, gradation, symmetry, harmony, balance, contrast, proportion, rhythm, dynamics, unity and simplicity are frequently used in product design (Graves, 1941; Kim, 2006). By manipulating product elements to assume a unified order, the aesthetic qualities of a product can be enhanced, thereby attracting users' attention (Norman, 2004). For instance, Bell et al. (1991) utilised colour photographs of living room furniture to test the unity of styles, and found that the aesthetic response was correlated with the perceived unity of a style. Lennon (1990) reported that a model with matching clothes and accessories appeared more competent and desirable compared with a model with nonmatching garments and accessories. Prior studies have shown that consumers respond more positively to products exhibiting a high degree of unity compared with those exhibiting a low degree of unity. Therefore, the study of unity is a critical issue in designing a product for aesthetic pleasure. Consumers’ emotions may be also affected by line-style elements (Coates, 2003) which visually comprise a product with an aesthetic quality. Bar and Neta (2006) indicated that people prefer objects defined by curved lines over those with sharp angles. A chair composed of curved lines appears to be more likely to please users compared with one composed of straight or slanted lines. For example, the Pingo chair designed by Hans Jakobsen in 2000 was devised specifically with curved elements in accordance with the shape of the back, seat, and legs, resulting in an elegant and aesthetically pleasing object. On the basis of the prior research, we assume that chair aesthetics is associated with the assembly of component attributes and line-type elements. A chair's back, seat, and legs have a clear relationship. Thus, the basic components (i.e. seat, back and legs) have frequently been used as stimuli in experiments (Jindo et al., 1995; Park and Han, 2004). The line attributes of each component were often classified by previous researcher as square, round, and slanted for the back; the same elements were applied to the seat; and straight, curved, and slanted lines were designated for the legs. In this study, we argue that the unity of a chair may influence its aesthetic quality, which in turn affects users' emotions (Fig. 1). In other words, a higher degree of unity is equal to a higher aesthetic, and in turn evokes a stronger emotional response in the user mainly (M1 & M2). Furthermore, we examined the causal relationships between the aesthetic and its pleasure reactions (M2), and also to test whether unity had a direct effect on pleasure (M3). 2.2. Gestalt laws and unity In perceiving a product's appearance, our sensory systems detect order in chaos or unity in variety (Hekkert, 2006) when
Fig. 1. Conceptual research framework.
93
examining a group of elements. People have a tendency to view objects such as a chair's structural arrangement as close together or feeling as if belonging together, according to Gestalt laws. Gestalt laws are summarised as ‘the whole is more than the sum of its € hler, 1920, pp.17). However, previous studies on Gestalt parts’ (Ko laws have focused on graphical patterns in dynamic graph drawings (Nesbitt and Friedrich, 2002), instructional screen design (Smith-Gratto and Fisher, 1999) and even the haptic and visual grouping of elements (Chang et al., 2007). Specifically, for the present study, we used Gestalt laws to explain chair shapes in terms of a chair's unity (i.e. line attribute arrangement). The principles of proximity, similarity, and continuation were adopted to explain how it is possible to perceive a chair's unity through its compositional arrangement. The proximity principle is defined as elements that are located closest to each other tending to form a group (Wertheimer, 1923). In Fig. 2, Lines 1 and 2 are close together and are viewed as a single group, whereas Lines 3 and 4 are regarded as separate. In this case, viewers may attempt to perceive both the chair's back and seat as one proximity piece (e.g. , , and ) because of the closed array between the two components. According to the similarity principle, similar elements tend to form groups; users regard similar elements (e.g. having the same characteristics) as belonging together, and perceive them to be as such on the basis of the attributes (e.g. line, shape, colour, texture, value, volume, and orientation) (Wertheimer, 1923; Chang et al., 2007). For instance, two groups of heavy black lines and two groups of light lines are arranged in straight lines (Fig. 3). Heavy lines are grouped as one, whereas light lines are grouped separately. Moreover, a chair composed of three types of lines (e.g. , , and ) may appear to be completely inconsistent, thus exhibiting no unity at all. Continuation concerns the eye seeking relationships between shapes. This occurs when the eye follows a line, a curve, or a sequence of shapes, even when it crosses over negative and positive shapes (Graham, 2008). As shown in Fig. 4, the human eye tends to perceive the ‘ ’ as a completed ‘S’. On a verticle plane, the back and legs of a chair (e.g. , , and ) may convey a continuous shape which tends to appear to have a greater degree of unity to the senses. In this study, the chair arrangement of each component (i.e. back, seat, and legs) was controlled and arrayed into every possible combination which could be associated with Gestalt laws. Participants may thus perceive the chair shape by subconsciously applying Gestalt laws.
3. Methods 3.1. Visual stimuli The chair was adopted as a stimulus in this experiment because of its widespread use and special structure array, indicating a clear composition on each component. To control the experiment, we used the chair archetype composed of three distinct components: the back, seat, and four legs. Chair arms were omitted because of research limitations concerning experimental complexity. Stimuli, including square, round, and slanted shapes, were employed as representative elements, resulting in 27 stimulus combinations (Fig. 5). Component shape was used as a factor to facilitate unity variation. The following section details how we varied and combined the components to reflect unity as a variable.
Fig. 2. Illustrated proximity.
94
W. Tyan-Yu et al. / International Journal of Industrial Ergonomics 59 (2017) 92e99
arrangements and (2) the three types of line attributes (straight, curved, and slanted). The following five groups were included in the first category: one group of chairs with three similar attribute components; three groups with at least two similar attribute components; and one group with no similar attribute components (Fig. 5). By contrast, four groups were included in the second category: chairs with three straight lines; two straight lines, and one curved line; one straight line and two curved lines; and three curved lines (Fig. 6).
Fig. 3. Illustrated similarity.
3.3. Emotional assessments The pleasure arousal dominance (PAD) scale, developed by Mehrabian and Russell (1974), is a semantic differential scale that includes the dimensions of pleasure, arousal, and dominance. Pleasure can be defined as a positive affective state that is distinct from feelings (e.g. preference and reinforcement). Arousal is an emotional state progressing from sleepiness to intense excitement. The third dimension, dominance, refers to the extent to which a person feels either positively or negatively towards an environment (Mehrabian and Russell, 1974). A set of bipolar adjective pairs was used for each item question, and it was rated on a 9-point scale. The PAD scale is widely used in environmental psychology as well as in product and interface design. Dominance was omitted for validity purposes (Russell and Pratt, 1980; Russell et al., 1981). Only the pleasure and arousal dimensions were adopted, and the stimuli are presented as 2D images in this study. Six emotional adjectives were used for each dimension. Additional items, namely ‘perceived unity’ and ‘aesthetics’, were included in the questionnaire. A 5point Likert scale was used for each questionnaire item.
Fig. 4. Illustrated continuation.
3.4. Participants
Fig. 5. Chair classification based on component attributes.
Fifty-four students (men ¼ 23, women ¼ 31, average age ¼ 21 years) were recruited from universities in Taiwan for the experiment. Of the participants, 27 majored in furniture design at the National Taipei University of Technology, and the other 27 majored in product design at Chang Gung University. To ensure balanced participation, 27 of the students majored in furniture design had sufficient aesthetic experience and were able to perceive line drawing chairs, whereas the other 27 first-year students had limited experience in aesthetic training.
3.2. Chair unity variable
4. Results and analysis
Chair unity was regarded as the independent variable, whereas ‘perceived unity’ (assessed through questionnaire items), aesthetics, pleasure, and arousal were considered the dependent variables. To test unity, the stimuli were classified according to (1) the three types of component attribute (back, seat, and legs)
A pretest was conducted to examine the validity of the PAD scale. Factor analysis confirmed the two emotional dimensions of pleasure and arousal. The KMO measure of sampling adequacy was 0.885, above the recommended value of 0.5, and Bartlett's test of sphericity indicated significance (p < 0.001). Internal consistency
Fig. 6. Chair classification based on line attributes.
W. Tyan-Yu et al. / International Journal of Industrial Ergonomics 59 (2017) 92e99 Table 1 Relationship between independent and dependent variables. Variables
Perceived unity
Aesthetics
Pleasure
Arousal
Chair unity Perceived unity Aesthetics Pleasure
0.433***
0.205*** 0.492***
0.145*** 0.359*** 0.621***
-0.009 0.053 0.094*** 0.112***
***p < 0.01, **p < 0.05, *p < 0.1.
was tested with Cronbach's alpha (0.925 for pleasure and 0.867 for arousal). The statistical results confirmed the reliability of the PAD scale. 4.1. Independence and dependence tests We regarded chair unity as the independent variable, and perceived chair unity, aesthetics, pleasure, and arousal as the dependent variables. The results of Pearson's correlation test revealed highly positive correlations between perceived unity and chair unity (Table 1). Perceived unity, aesthetics, and pleasure were strongly and positively correlated. However, arousal was weakly correlated to both chair unity and perceived unity. 4.2. Unity on ‘component attributes’ As shown in Table 2, the ANOVA results for a randomised block design revealed significant differences between chair unity and perceived unity, aesthetics and pleasure; however, such significant differences were not found between chair unity and arousal. Furthermore, the Scheffe test results revealed that the participants could sense a high degree of unity when they perceived the chair's back, seat, and legs as consistent (e.g.), but a low degree of unity when they perceived the chair's back, seat, and legs as individually distinct (e.g.). In other words, a chair composed of similar component attributes is likely to convey a higher degree of unity, whereas a chair composed of completely different component attributes is likely to convey a lower degree of unity. The chair composed of two
95
(out of three) similar attribute components exhibited less unity compared with those with three of the same attribute components; the unity ranking for the chairs was as follows: chairs with the same attributes for the back and seat > back and legs > seat and legs. The Scheffe test results consistently showed that the participants perceived a high aesthetic quality when they observed the chair's back, seat, and legs as consistent, but a low aesthetic quality when they observed the chair's back, seat, and legs as individually distinct attributes. Furthermore, the results revealed a weaker aesthetic quality for chairs with two similar components than for those that had similar attributes for all the components. The aesthetic ranking of the chairs was as follows: a chair with the same attributes for the back and seat > back and legs > seat and legs. Similarly, as shown in Table 2, the Scheffe test results also revealed that the participants felt a high degree of pleasure when they perceived the chair's back, seat, and legs as a consistent attribute as well as the back and seat. A lower degree of pleasure was found with a lower degree of unity when they perceived the chair's back, seat, and legs as inconsistent and the seat and legs as consistent. Overall, a greater unity between the chair components led to a greater sense of pleasure evoked in the participants. However, no significant differences were observed in the effect of different degrees of unity on arousal. We suspect that the images used in this experiment, because of their flat and simple lines, were visually insufficient to evoke a strong sense of arousal. 4.3. Unity on ‘line attributes’ Bar and Neta (2006) indicated that line attributes may have considerable influence on consumers' aesthetic perception and emotions. To understand how chair line attributes can affect consumers’ sense of unity, aesthetic sense, and emotions, in the experiment, we further examined the effects of two types of lines, namely straight (including straight and slanted lines) and curved lines (Fig. 6), on unity, aesthetic, pleasure, and arousal. As shown in Table 3, the ANOVA results revealed highly significant differences
Table 2 ANOVA (randomised block design) for ‘component unity’ influencing perceived unity, aesthetic, pleasure and arousal.
Perceived unity Aesthetic Pleasure Arousal
BG WG BG WG BG WG BG WG
Sum of Squares
df
Mean Square
F
Sig.
Ranking (Scheffe test)
403.398 1324.157 57.529 1214.915 24.398 847.825 0.745 832.901
4 1400 4 1400 4 1400 4 1400
100.850 0.946 14.382 0.868 6.099 0.606 0.186 0.595
106.626
0.000
back ¼ seat ¼ leg > back ¼ seat > back ¼ leg > seat ¼ leg > Back s seat s leg
16.573
0.000
back ¼ seat ¼ leg, back ¼ seat > back ¼ leg (seat ¼ leg) > Back s seat s leg
10.072
0.000
back ¼ seat, back ¼ seat ¼ leg > back ¼ leg, seat ¼ leg, Back s seat s leg
0.313
0.869
Note: Between groups ¼ BG, Within groups ¼ WG; ‘ ¼ ‘ means the same attribute.
Table 3 ANOVA (randomised block design) for ‘line unity’ affecting perceived unity, aesthetic, pleasure, and arousal.
Perceived unity Aesthetic Pleasure Arousal
BG WG BG WG BG WG BG WG
Sum of Squares
df
Mean Square
F
32.311 1407.245 36.992 1235.453 38.337 833.885 25.870 807.776
3 1401 3 1401 3 1401 3 1401
106.770 1.004 12.331 0.882 12.779 0.595 8.623 0.607
106.296 0.000 all curved lines > all straight lines > two curved and one straight lines > one curved and two straight lines 13.983 0.000 all curved lines > all straight lines, two curved and one straight lines > one curved and two straight lines 21.470 0.000 all curved lines > two curved and one straight lines, all straight lines, one curved and two straight lines 14.956 0.000 all straight lines > one curved and two straight lines, two curved and one straight lines, all curved lines
Note: Between groups ¼ BG, Within groups ¼ WG.
Sig.
Ranking (Scheffe test)
96
W. Tyan-Yu et al. / International Journal of Industrial Ergonomics 59 (2017) 92e99
between chair unity and perceived unity, aesthetics, pleasure, and arousal. Table 3 shows that the chairs composed of only curved lines (e.g. back, seat, and legs) and those composed of only straight lines led to a high degree of unity. In other words, the participants perceived that the two types of chairs composed of the same line attributes conveyed a significantly high degree of unity. By contrast, chairs composed of either two curved or two straight lines had less unity compared with those composed of three similar attributes. Overall, the chairs composed of more similar line attributes yielded a higher degree of unity compared with those with less similar attributes. Similarly, we found that a chair composed of a greater number of similar line attributes also had a higher likelihood of conveying a higher aesthetic quality compared with that with less similar line attributes. The chairs with two curved lines and one straight line evoked a stronger aesthetic response compared with the chairs with one curved and two straight lines. The result indicated that the participants noted that the chairs with more curved lines can evoke a stronger aesthetic response than can chairs with less curved lines. These results implied that increasing the number of curved lines consistently may enhance aesthetic quality. As mentioned, a chair composed of only curved lines was observed to evoke a strong sense of pleasure. By contrast, a chair composed of only straight lines was found to evoke less pleasure than one with all three components curved. This chair sample was followed by another with three curved features, and afterward, a chair with two curved features which evoked a less intense pleasure response. The results implied that consumers' pleasure response may be affected by the number of curved attributes in a chair. A greater number of curves in the chair appeared to yield a higher likelihood of evoking a stronger pleasure response compared with straight lines. However, Table 3 shows that chairs composed of only straight lines (i.e. a high degree of unity) generated greater arousal compared with those composed of curved lines. This result implied that a chair composed of more straight lines has considerable potential in evoking consumers’ arousal. Conversely, curved lines may reduce the intensity of arousal. 4.4. Unity of aesthetic evoking users’ emotions Path analysis was performed to examine the affective relationships among unity, aesthetic and pleasure reactions (Fig. 1). Path analysis involved three regression models. The first regression model (M1) was used to verify the causal relationships between the unity of a chair and its aesthetic quality. By controlling for participant differences, we found that aesthetic, as the dependent variable, was significantly influenced by the independent variables of unity (Table 4a), with significance at 1%. The second regression model (M2) was used to examine the causal relationships between the aesthetic and its pleasure reactions. Table 4b shows that the relationship was supported at a significance of 1%. The third regression model (M3) was used to test whether unity had a direct effect on pleasure. Table 4c shows that the direct effect was significant at 10%. The overall effect of unity on pleasure was 0.141 (0.406 0.285 þ .025). The result implied that unity also directly Table 4a Regression analysis of the effect of aesthetic on unity (M1) (dependent variance ¼ aesthetic). Variable
Parameter
Std. Error
T
Sig.
Constant Perceived- Unity
1.951*** 0.406***
0.174 0.020
11.207 20.116
0.000 0.000
R2 ¼ 0.387 (R2adj ¼ 0.364); ***p < 0.01, **p < 0.05, *p < 0.1.
Table 4b Regression analysis results of the effect of pleasure on aesthetic (M2) (dependent variance ¼ pleasure). Variable
Parameter
Std. Error
T
Sig.
Constant Aesthetic
2.546*** 0.298***
0.109 0.014
23.359 20.804
0.000 0.000
R2 ¼ 0.409 (R2adj ¼ 0.387); ***p < 0.01, **p < 0.05, *p < 0.1.
Table 4c Regression analysis of the effect of pleasure on aesthetic and unity (M3) (dependent variance ¼ pleasure). Variable
Parameter
Std. Error
T
Sig.
Constant Aesthetic Perceived- Unity
2.511*** 0.285*** 0.025*
0.111 0.016 0.014
22.677 17.502 1.775
0.000 0.000 0.076
R2 ¼ 0.411 (R2adj ¼ 0.388); ***p < 0.01, **p < 0.05, *p < 0.1.
affected consumers' sense of pleasure. In summary, path analysis revealed that both the indirect and direct effects of unity on pleasure were significant. A higher degree of unity resulted in a greater extent of aesthetic responses, and the participants’ sense of pleasure was likely to have increased considerably as well. 4.5. Decision tree analysis on unity The statistical results showed that chair unity had a considerable influence on aesthetics and furthered users' pleasure responses. Thus, using the unity principle as a vehicle for enhancing a product's aesthetic quality can be an effective strategy that benefits designers in the development of a product meant to evoke pleasure. This study proposes a unity-based method which can guide designers in manipulating components affectively and effectively during form development. To devise this systematic method, we examined how our participants perceived the unity of each stimulus (i.e. 27 chairs) by performing decision tree analysis with the MATLAB 8.2 package, which was used to analyse the trends of a small amount of data without meaning connected, in which the decision tree allows the identification of the optimal arrangement of unity with the composition of the three components (i.e. back, seat, and legs). A decision tree is composed of branch and leaf nodes, and each branch node contains an if/then rule. The decision process starts from the top branch node which is called ‘root’ in the decision tree literature. When the ‘if’ condition is satisfied, for example, the leg component is assigned to a straight line, and the process progresses to its consequent node; otherwise, it advances to another node. Fig. 7 displays the results of a decision tree in which three trails derived from straight, curved, and slanted legs were assigned. The bottom degree of each trail displayed optimal as well as poor samples. For instance, with a high value factor (0.35), the right branch under the straight legs suggested four chair samples which enhanced the high degree of unity. The other results appeared under the curved and slanted legs; the curved legs, with the second-highest value factor (0.55), suggested two optimal samples; and the slanted legs, with the third-highest value factor (0.71), suggested three optimal samples. Overall, tree analysis revealed a valuable direction for designing a chair with a significant degree of unity which can enhance the chair's aesthetic quality and further evoke a considerable sense of pleasure in users. For future application, the decision tree method (DTM) can be integrated with the concept of a database system which can potentially allow designers to access the system freely during the form development stage.
W. Tyan-Yu et al. / International Journal of Industrial Ergonomics 59 (2017) 92e99
97
Fig. 7. Result of the decision tree classifier, applied to the chair components.
5. Discussion 5.1. Unity, aesthetic, and emotion effects The empirical results revealed the significant effect on the various extents of unity arrangement. Specifically, a chair with all three attribute components that were similar was likely to convey the highest degree of unity, whereas a chair with all three attribute components that were different was likely to convey the lowest degree of unity (Table 2). Because of the differences in component arrangement, chairs with two similar components may have a different degree of unity and aesthetic responses. The participants' perceptions were likely influenced by how they viewed ‘two component arrangements’ as being together. Accordingly, each assembled component had a special relationship with another, with the Gestalt laws able to provide a reasonable explanation for these visual relationships. For instance, the proximity and continuation principles explain why the participants perceived greater unity in a chair when its back and seat had consistent attributes (Fig. 5, fourth row) compared with the unity of a chair with back and leg consistency (Fig. 5, third row) or seat and leg consistency (Fig. 5, second row). In this seat and leg consistency case, proximity explains the unity sensation when a closed position between the back and seat as a single attribute is perceived. In addition, continuation supports the visual connection between the chair's back and seat as a unit, whereas two components posited next to each other demonstrate similar attributes that are connected. Thus, the chair's back and seat as a single element are supported both by proximity and continuation, which enhances the chair form as a whole, making participants perceive the unit sensation clearly. Moreover, with a chair's back and legs as consistent attributes, proximity explains the unity sensation when someone perceives the back and legs as a respectively upper and lower relationship on a vertical orientation as wholeness. Conversely, the chair with seat and leg consistency exhibited less unity because of the lack of a clear visual orientation or any connection whatsoever. Similarly, the results of the aesthetic and pleasure responses also revealed that the participants' greatest responses on aesthetic and pleasure were evoked when they perceived the three attributes of a chair to be consistent (Table 2); conversely, the chair with three components having all different attributes evoked the lowest degree of aesthetic and pleasure responses. As shown consistently by the unity results, the participants presented a similar ranking of aesthetic and pleasure responses when they perceived a chair to
have two consistent attributes. The results further confirmed that unity is a critical design principle for enhancing a product's aesthetic quality. However, according to previous research, the line attributes of a product may also influence the aesthetic and pleasure responses. The examination of line attributes further indicated that the participants' sense of pleasure was also affected by the types of line elements. For instance, a chair composed of more curves appeared to have a greater likelihood of evoking a stronger pleasure response compared with one composed of straight lines (Table 3). Specifically, the chair whose attributes all have curved lines exhibited a high degree of unity which also consistently evoked the greatest sense of pleasure. The results also revealed that the chair composed of straight lines for all its attributes exhibited a high degree of unity but evoked a lower degree of pleasure compared with the chair with all three curved attributes. Chairs with three curved attributes may evoke the strongest aesthetic and pleasure responses. Conversely, chairs with two straight and one curved line attribute evoked the weakest aesthetic and pleasure responses. Moreover, chairs with two curved and one straight line evoked a stronger pleasure response compared with those with all straight lines, implying that curved-line elements have a greater likelihood of dominating consumers' emotional responses compared with straight lines. Designers should take advantage of curved elements for the enhancement of the pleasure response. These results implied that a product composed of more curved elements may increase the sense of pleasure more than the same product composed of straight lines may. The results confirmed the finding of Bar and Neta (2006) that people prefer objects with curved elements to those with straight lines and sharp angles. This results also confirmed that chairs with curved or rounded lines appear more attractive and perhaps even convey a greater sense of warmth compared with other chairs (Park and Han, 2004), whereas straight lines convey rigidity and sharpness. Overall, these results implied that a product composed of curved lines and with a high sense of unity should have a greater likelihood of evoking consumers’ pleasure responses compared with those with less unity and more straight lines.
5.2. Unity enhances aesthetic and pleasure The results revealed that aesthetic response was strengthened by an increased perception of unity and weakened by a reduced degree of unity. Tables 4aeb shows that the aesthetic responses increased with high-unity forms, and users' sense of pleasure was
98
W. Tyan-Yu et al. / International Journal of Industrial Ergonomics 59 (2017) 92e99
Fig. 8. a) Meditation chair, b) New Ming's chair, c) Devil's chair, d) Trapezoid chair.
evoked following the increase of the aesthetic response. These findings are consistent with research on people's preference for objects that exhibit harmonious traits (Hekkert, 2006; Kumar and Garg, 2010); unity provides a harmonious characteristic. This result also supported the findings of Coates (2003); that is, as a single variable, unity can enhance aesthetic quality. Accordingly, a product with attributes consistency is always implied to be an optimal choice for the enhancement of chair unity for strengthening consumers' aesthetic and pleasure responses.
5.3. Application of the unity principle in design According to the unity experiment and decision tree analysis results, we recommend an effective design process for the development of products with unity which can further enhance a product's aesthetic and pleasure responses. The processes were developed on the basis of four aspects: (1) defining the product image and line style, (2) using DTM to identify the optimal unity of an object, (3) enhancing the aesthetic quality, and (4) rationalising the design. The first step involves ‘defining the product image and line style’ which concerns identifying the type of product image (e.g. line style) required for a particular project or one that satisfies a user's requirements. In actuality, the product images or styles should fit into users' lifestyle or personality. For instance, to create a chair with floral elements intended to satisfy a user who prefers to live in a floral and elegant space. Designers may choose curved elements as a basic component for developing a chair to match the similar interior style. After defining the line style, the second step involves reviewing the data with the DTM and identifying an optimal sample as a reference for ideation. In this step, the data from the decision tree diagram enable designers to review optimal concepts and helps them determine the design direction so that they can create an ideal form quickly. By doing a trial-and-error, designers have the opportunity to test different shapes based on various data from the decision tree diagram and to determine the ideal result. The third step concerns how aesthetic qualities are pleasing to people; to achieve the desired results, designers are also suggested to incorporate aesthetic principles (e.g., balance, contrast, and visual weight) and Gestalt laws when finalizing their product forms. The final step considers the details of engineering and manufacturing applications. Here, designers should rationalise the final designed product in terms of object structure, material strength, and tooling. To test this design process, a 9-week chair workshop was held at Chang Gung University. At the end of the workshop, four chairs (Fig. 8a, b, c, d) were constructed and recommended by five faculty members and three professional designers for display in the Young Designers Exhibition, a well-recognised exhibition at Taipei World Trade Center. Overall, the participants from the workshop agreed that this design process is both effective and affective for the development of aesthetically pleasing chairs.
6. Future research Follow the emerging trend of CAD design, we suggest integrating both the DTM and unity principles within a database in future studies. This system will allow designers to simulate 3D results quickly when developing an aesthetic product that enhances unity. Moreover, the system should encompass three parts: a components databank, unity aesthetic principles, and DTM tools. The component databank should include several chair parts (e.g., legs, backs, and seats), which will allow designers to retrieve and create 3D forms quickly according to their design style. A DTM system will provide further optimising suggestions, by displaying 3D aesthetic forms on the computer screen immediately; 3D forms that prompt users will allow designers to freely review and manipulate the design forms. Finally, the system perhaps should also include more aesthetic functions, such as proportion, visual weight, and colour parameters to enhance aesthetic quality. Notably, however, these concepts must be further tested before they can be used for designing purposes. 7. Conclusion In this study, the unity principle plays an effective and affective role in the enhancement of the product aesthetic and consumers' sense of pleasure. The decision tree method may benefit designers seeking to develop a product with aesthetic effectively, which further evokes consumers’ positive emotions. However, designers should not limit themselves to employing only the unity principle when other principles (e.g. balance, colour, and material) can also have great potential in affecting the final design. Because of the research limitations, this study did not cover all types of products and consumers which required having further test in the future research. References Bar, M., Neta, M., 2006. Humans prefer curved visual objects. Psychol. Sci. 17 (8), 645e648. Bell, S.S., Holbrook, M.B., Solomon, M.R., 1991. Combining esthetic and social value to explain preferences for product styles with the incorporation of personality and ensemble effects. J. Soc. Behav. Personal. 243e273. Blijlevens, J., Creusen, M.E., Schoormans, J.P., 2009. How consumers perceive product appearance: the identification of three product appearance attributes. Int. J. Des. 3 (3), 27e35. Bloch, P.H., Brunel, F.F., Arnold, T.J., 2003. Individual differences in the centrality of visual product aesthetics: concept and measurement. J. Consum. Res. 29 (4), 551e565. Chang, D., Nesbitt, K.V., Wilkins, K., 2007. The Gestalt principles of similarity and proximity apply to both the haptic and visual grouping of elements. In: Proc. Eighth Australasian User Interface Conference (AUIC2007) (Ballarat, Australia). Chang, W., Wu, T.Y., 2007. Exploring types and characteristics of product forms. Int. J. Des. 1 (1), 3e14. Chignell, A., Costelloe, T.M., 2011. A dialogue concerning aesthetics and apolaustics. J. Scott. Philos. 9 (1), 5e16. Coates, D., 2003. Watches Tell More than Time. McGraw-Hill, New York. Dhar, R., Wertenbroch, K., 2000. Consumer choice between hedonic and utilitarian goods. J. Mark. Res. 60e71. Graves, M.E., 1941. The Art of Color and Design. McGraw-Hill, New York. Graham, L., 2008. Gestalt theory in interactive media design. J. Humanit. Soc. Sci. 2 (1).
W. Tyan-Yu et al. / International Journal of Industrial Ergonomics 59 (2017) 92e99 Hekkert, P., 2006. Design aesthetics: principles of pleasure in design. Psychol. Sci. 48 (2), 157e172. Helander, M.G., 2003. Forget about ergonomics in chair design? Focus on aesthetics and comfort! Ergonomics 46 (13e14), 1306e1319. Hirschman, E.C., Holbrook, M.B., 1982. Hedonic consumption: emerging concepts, methods and propositions. J. Mark. 92e101. Hoyer, W.D., Stokburger-Sauer, N.E., 2012. The role of aesthetic taste in consumer behavior. J. Acad. Mark. Sci. 40 (1), 167e180. Jindo, T., Hirasago, K., Nagamachi, M., 1995. Development of a design support system for office chairs using 3-D graphics. Int. J. Ind. Ergon. 15 (1), 49e62. Kim, N., 2006. A history of design theory in art education. J. Aesthet. Educ. 40 (2), 12e28. €hler, W., 1920. Physical Gestalten (Die Physischen Gestalten in Ruhe und im Ko stationaren Zustand). Erlangen: eine naturphilosophische Untersuchung: ix259. In: Ellis, W.D. (Ed.), A Sourcebook of Gestalt Psychology. Routledge & Kegan Paul, London, pp. 17e54. Kumar, M., Garg, N., 2010. Aesthetic principles and cognitive emotion appraisals: how much of the beauty lies in the eye of the beholder? J. Consum. Psychol. 20 (4), 485e494. Lennon, S.J., 1990. Effects of clothing attractiveness on perceptions. Home Econ. Res. J. 18 (4), 303e310. Lauer, A.D., 1979. Design Basics. Holt, Rinehart. New York: Holt, Rinehart, & Winston, New York. McDonagh, D., Bruseberg, A., Haslam, C., 2002. Visual product evaluation: exploring
99
users' emotional relationships with products. Appl. Ergon. 33 (3), 231e240. Mehrabian, A., Russell, J.A., 1974. An Approach to Environmental Psychology. the MIT Press. Nesbitt, K.V., Friedrich, C., 2002. Applying Gestalt principles to animated visualizations of network data. In: Proc. IV 2002. IEEE ComputerSociety Press, London, UK, pp. 10e12. Norman, D.A., 2004. Emotional Design: Why We Love (Or Hate) Everyday Things. Basic Books, New York. Park, J., Han, S.H., 2004. A fuzzy rule-based approach to modeling affective user satisfaction towards office chair design. Int. J. Ind. Ergon. 34 (1), 31e47. Postrel, V., 2003. The Substance of Style. HarperCollons, New York. Russell, J.A., Pratt, G., 1980. Affective quality attributed to environments a factor analytic study. Enviro description of the affective quality attributed to environments. J. Personal. Soc. Psychol. 38 (2), 311. Russell, J.A., Ward, L.M., Pratt, G., 1981. Affective quality attributed to environments: a factor-analytic study. Environ. Behav. 13, 259e288. Smith-Gratto, K., Fisher, M., 1999. Gestalt theory: a foundation for instructional screen design. J. Instr. Technol. Syst. 27 (4), 361e371. Veryzer, J., Robert, W., Hutchinson, J.W., 1998. The influence of unity and prototypicality on aesthetic responses to new product designs. J. Consum. Res. 24 (4), 374e385. Wertheimer, M., 1923. Untersuchungen zur Lehre von der Gestalt. II. Psychol. Res. 4 (1), 301e350.