Materials and Design 31 (2010) 2932–2941

Contents lists available at ScienceDirect

Materials and Design

journal homepage: www.elsevier .com/locate /matdes

A tool for meaning driven materials selection

Elvin Karana *, Paul Hekkert, Prabhu KandacharFaculty of Industrial Design Engineering, Delft University of Technology, Landbergstraat 15, 2628 CE, Delft, The Netherlands

a r t i c l e i n f o

Article history:Received 4 November 2009Accepted 14 December 2009Available online 21 December 2009

Keywords:H. Selection of materialsE. Properties of materialsMeaningExpressive characteristicsEmotional design

0261-3069/$ - see front matter � 2009 Elsevier Ltd. Adoi:10.1016/j.matdes.2009.12.021

* Corresponding author. Tel.: +31 15 27 85726; faxE-mail address: e.karana@tudelft.nl (E. Karana).

a b s t r a c t

There are several tools used in materials selection processes by designers. However, they are mostly engi-neering based tools, which are dominated by numerical (or technical) material data that is mostly of usein embodiment or detailed design phases of new product development. On the other hand, productdesigners consider certain aspects such as product personality, user-interaction, meanings, emotions intheir material decisions. In this regard, existing tools and methods do not fully support designers in theirmaterials selection processes. This paper describes the development of a new materials selection toolholding the idea of [meaning driven materials selection]. In addition, the paper consists of a study con-ducted to create data for a dummy application.

� 2009 Elsevier Ltd. All rights reserved.

1. Introduction

In most of the materials selection sources, an analytical ap-proach is followed [1–3]. In an analytical approach, a set of objec-tives and constraints are defined. Afterwards, the properties of anumber of existing materials are analyzed based on the definedobjectives and constraints. The candidate materials are then se-lected. Ashby and Cebon [4] sum up the materials selection activityin four main steps: (1) translate the design requirements as con-straints and objectives, (2) screen the material world to identifymaterials that cannot do the job, (3) rank the materials that cando the job best, and (4) explore the top rated materials. In thatsense, materials selection is carried out (consciously or not) as adesign activity, involving the phases concept creation (by formulat-ing material objectives and constraints, and arriving at candidatematerials), testing and comparing candidate materials, and makinga detailed selection with technical specifications.

The four steps described by Ashby and Cebon [4] summarize thetraditional materials selection approach promoted in engineeringdesign. Constraints and objectives are mainly determined by tech-nical requirements and materials are selected accordingly. In prod-uct design, however, materials should not only fulfill technicalrequirements but also appeal to the user’s senses and contributeto the intended meaning of a product. These concerns are intro-duced to the domain of design with alternative approaches suchas design for experience [5], pleasure in design [6], design for emotions[7] and multi sensory design [8]. Product designers are responsiblefor taking these concerns into account in order to use materials

ll rights reserved.

: +31 15 27 81839.

efficiently to transfer certain meanings. In other words, materialsare selected for creating certain experiences with their physical en-tity as well as intangible characteristics. Following this notion,materials selection in product design in this paper is defined asthe selection of appropriate material(s) for designed products byconsidering related design criteria such as manufacturing pro-cesses, availability, cost, function, shape, use, as well as meanings,associations, emotions, characteristics of users, cultural aspects.

Designers who intend to create certain meanings through thematerials of their products are confronted with the difficulty thatthere is not a one-to-one relationship between material propertiesand intended meanings [9]. Combinations of different propertiesevoke particular meanings for specific users within specific con-texts. This statement is introduced with the Meanings of Materials(MoM) model (Fig. 1) in Karana’s PhD thesis [10]. The model pre-sents the meaning of a material as a relational concept in whichmaterial, product and user are jointly effective. Furthermore, a setof related aspects are identified and tested in a series of studies,such as sensorial properties, manufacturing processes, shape, func-tion, gender, age, expertise and culture [10]. Following the MoMmodel, this paper describes the development of a tool, namely[Meanings of Materials] tool, which aims to assist designers inmanipulating meaning creation in materials selection.

The following section reports on the two main steps that werefollowed in developing the Meanings of Materials tool: (1) thestructure of the tool (order of actions) and (2) the content of thetool (generating data and presenting the outcome. At the end ofthis section, the proposed tool is summarized. In the third section,a study is conducted in order to generate data for a dummy appli-cation. The paper ends with a comprehensive discussion on the re-sults of the conducted study.

Fig. 1. [Meanings of Materials] Model [10].

E. Karana et al. / Materials and Design 31 (2010) 2932–2941 2933

2. The [Meanings of Materials] tool

When people are asked to describe a certain material, they fre-quently refer to its expressive characteristics and these character-istics are grounded in different aspects of materials (and products).A particular material of a product, for instance, might express pro-fessionalism predominantly through its shiny, robust and smoothproperties and the product’s sharp-edged geometry. Herein, shini-ness, robustness, smoothness and sharp-edge geometry cooperateand jointly contribute to a material’s expressive character. Expres-sive characteristics (or meanings, variously called figurative or ab-stract characteristics, see [11]) are not factually part of a materials’physical entity or appearance (i.e. a material is not literally femi-nine or masculine) [9].

A meaning of a material is evoked by the interactions betweenproduct aspects (such as shape, function) and material properties,with respect to how and in which context the material is used andwho the user is, and can change over time. Thus, a meaning of amaterial cannot be reduced to a single property or a single sensorydomain [10]. Therefore, it is not possible to define simple designrules for a certain material–meaning relationship. Nevertheless,there are some patterns that identify how materials obtain theirmeanings [10]. A material, for instance, may express professional-ism when it is smooth and dark (coloured), when its used in an of-fice environment and when certain technical properties arecombined for enhancing its function (e.g. combining strength andlightness). We assume that a designer who can understand theserelationships (which we may call ‘meaning evoking patterns’) canmore deliberately (or systematically) manipulate meaning creationin materials selection processes. In order to make designers capa-ble of finding these patterns, a tool should first familiarize design-ers with the key aspects (such as shape, user, manufacturingprocesses) that play an important role in attributing meanings tomaterials. The tool should convey the idea that many meaningscan be attributed to many materials dependent on different prod-ucts and contexts.

The three major aims of a proposed [Meanings of Materials] toolare: (1) to familiarize designers with the main components (or factors)of the Meanings of Materials model, (2) to show which aspects (undermain components) play an important role for certain meanings (suchas sensorial properties, gender, culture, shape), and (3) to stimulatedesigners to find the relationships (or patterns) between these aspectsand meanings. In this way, we aim to encourage designers to system-atically involve meaning considerations in their materials selection

processes. This approach is termed meaning driven materialsselection in this research. For the three goals listed above, weaim to provide designers with a collection of material examples(as material samples or materials embodied in products) that havebeen selected by a number of individuals who think that eachmaterial example expresses a certain meaning. In this way, theintention is not to provide designers with explicit design rulesbut rather to encourage designers to make their own conclusionsby analyzing the selected materials.

2.1. Step 1: the structure of the tool

Scholars in the materials and design domain, underlie the needfor a materials selection tool to support designers in their materialsselection activities at early stages of the design process (i.e. con-cept creation) [12–15]. The tool was therefore required to be infor-mative, inspiring and appealing to designers. In order to achievethis, a level of interactivity in the tool was sought: the aim is toconstruct a database derived from a number of people who areasked to select materials expressing certain meanings. They areasked to provide a picture of the materials they selected and to ex-plain why they thought that the material they selected expressedthe given meaning. Then, they are asked to appraise the selectedmaterial in terms of sensorial properties via five point scales.

In approaching the proposed tool, designers ware expected tohave in mind the meaning(s) they would like to create throughthe material(s) of their designs. From this standpoint, it is impor-tant to provide designers with a number of material examples pre-sented alongside explanations made by the individuals who selectthe materials and point out their associated meanings.

In the completed MoM tool, designers can navigate through se-lected materials and explanations. Furthermore, the MoM modeland a list of important sensorial properties of materials are pre-sented in the tool to guide designers in their analysis of the se-lected materials. The main assumption is that: even though eachcase (comprising a single person’s explanation of the meaning theyattribute to a certain material) is unique, designers will be stimu-lated to combine the cases and identify meaning evoking patterns.The materials selection process, aided by the MoM tool, is intendedto finalize with an idea(s) of a material(s) conveying a certainmeaning.

The MoM tool incorporates 76 meanings, which are identifiedas material relevant meanings that designers are likely to wantto convey through the materials of their products [16], in order

2934 E. Karana et al. / Materials and Design 31 (2010) 2932–2941

to guide designers at the beginning of their materials selection pro-cesses. A designer, with an intention in mind, is first encouraged tobrowse through the meanings and see if his/her intention is similarto (or the same as) one of the offered meanings. The designer canthen access the examples and results from previous studies abouta meaning that he/she is interested in, or can require a new study ifthere are no examples or results about the intended meaning (or aclosely related meaning) in the tool. Thus, a new study is con-ducted and the results are added to the database. The designercan also require a new study to expand the data about a particularmeaning already existing in the tool. For instance, a designer whointends to create a feminine product and who wishes to select amaterial(s) that expresses the meaning feminine can find that thetool consists of data about feminine materials selected by a narrowgroup of people. The designer may want an additional study in or-der to see, for instance, what Mediterranean people think aboutmaterials that express femininity. In this way, the MoM tool is con-ceived as a growing database of material meanings obtained fromongoing studies. Fig. 2 summarizes the order of actions followed inthe MoM tool as: (a) a designer’s request, (b) data generation, and(c) outcome evaluation.

2.2. Step 2: the content of the tool

With the MoM tool, we aimed to provide designers with a vari-ety of material (and product) examples, along with data on mate-rial sensorial properties (ranked on the basis of evaluations) andexplanations of the MoM model components. The content of thetool was developed accordingly.

2.2.1. Data generationData are proposed to be generated through the results of differ-

ent studies conducted online with groups of people who are asked

Fig. 2. Order of actions foll

Fig. 3. Process of data ge

to select materials expressing particular meanings. People whoparticipate in the study are given the following three tasks: (1) se-lect a material that you think is ‘X’ (such as sexy, feminine, mod-ern), (2) provide a picture of the material you selected, and (3)explain your choice and evaluate the material on the given senso-rial scales (Fig. 3).

2.2.1.1. TASK 1: select a material. . .. While designing TASK 1 for thetool, a special attempt was made to define a task that would becomprehensible even for non-designers, and which would leadthem to material rather than object evaluations. Accordingly, thetask was defined as: select a material which you think is X (X: ameaning such as sexy, futuristic). An explanation page was pre-pared to provide participants with more detailed informationabout how to select a material expressing the given meaning(Fig. 4). The explanation points out that the selected material couldbe embodied in an object or in a part of an object (such as a handleof a kettle).

Differences in terminology between object and material are alsoexplained in two additional pages with the help of examples(Fig. 5). The participants are reminded that if they do not possessthe material themselves, but instead noticed it in a magazine,Internet, etc., and if they think it is a good example for the givenmeaning, they could still select that material as long as they areable to provide visuals of the material (pictures, photos, etc.) at alater stage.

2.2.1.2. TASK 2: provide a picture of. . .. Combining several picturesinto a visual whole makes it possible for designers to representnot just one object or idea, but something more multifaceted likea mood or a context [17]. Designers usually use visual collages todefine their target group and deduce certain product characteris-tics (such as form, colour) from these collages [18]. In addition,

owed in the MoM tool.

neration for the tool.

Fig. 4. Task 1 and the first explanation page.

Fig. 5. The second explanation page of TASK 1.

Fig. 6. TASK 2 and the explanation page.

E. Karana et al. / Materials and Design 31 (2010) 2932–2941 2935

2936 E. Karana et al. / Materials and Design 31 (2010) 2932–2941

designers prefer to transfer an idea (or a mood, or a context) to aclient (especially if he does not have a design background) using vi-sual representations [15].

On the basis of these considerations, the participants are askedto provide pictures of the materials they selected. On an additional

Fig. 7. Sensorial scales used for

page, they are instructed that the pictures may either be photo-graphs taken by themselves or others, or any type of visual (photo,modeling, etc.) taken from the Internet, magazines, or a similarsource (Fig. 6). They are asked for a supplementary detailed picturein cases where the selected material is not embodied in an object’s

generating data for the tool.

Fig. 8. Evaluation of the outcome of studies undertaken for the MoM tool.

E. Karana et al. / Materials and Design 31 (2010) 2932–2941 2937

whole, but in a part of it. In other words, two pictures are needed ifthe object is made of more than one material: one for the entire ob-ject and one for the part made of the material that expresses thegiven meaning.

2.2.1.3. TASK 3: evaluate the material. . .. Finally, the participants areasked to evaluate their selections. The appraisal of the selectedmaterials is done in two ways: first verbally, by responding to anopen ended question, and second with a list of sensorial propertiespresented with pictograms and five point scales (the sensorialproperties come from a previously conducted study, see [19])(Fig. 7). The aim of the open verbal explanation is to providedesigners with more specific information concerning the reasonsbehind each individual’s selection. After the verbal explanations,the participants are asked to assess the sensorial properties ofthe selected materials, presented with pictograms and five pointscales. The outcome of the tool is presented below (Fig. 8).

2.2.2. Outcome of the tool2.2.2.1. Collages of the materials and objects. Here, the main idea isto support designers in identifying meaning evoking patternsregarding materials of products in an inspirational way. It is pro-posed that displaying the selected materials altogether in a singleimage might enhance the probability of designers detecting thesimilarities and differences among the materials. For this reason,the collected pictures from the participants are presented as acollage, in which designers first see all the materials associatedwith a certain meaning in zoomed views (thumbnails). When azoomed view is clicked, the entire image appears.

2.2.2.2. Individual explanations of the participants. The MoM toolprovides designers with the verbal explanations made by eachindividual who participated in the data generation studies. Whena designer clicks on an image of a material presented in theMoM tool, besides the picture of the entire object, he/she findsthe rationale of the individual who selected the material. Theaim is to communicate to designers the participants’ main motiva-tions in their material–meaning selections.

2.2.2.3. Comparisons of the selected materials based on certain groupdivisions. This information type is provided to allow designers to

limit the results of the MoM tool to only those meanings offeredby a certain target group, such as materials selected by femalesor males, materials selected by Asian people. Designers are ableto compare the results of the sensorial scales and to see significantdifferences between the compared groups’ selections.

2.2.2.4. Quantitative results from the sensorial scales. This part of theMoM tool presents material ratings based on sensorial scales. Theresults are presented with a graph ranking the properties accord-ing to their mean scores. The most important properties for the gi-ven meaning are determined. In this way, designers first becomefamiliar with the properties that may have a general effect onthe meaning of a material, and second develop an appreciation ofthe properties that play a crucial role in attributing the intendedmeaning.

2.2.2.5. A link to a technical materials selection website. Designers areprovided with a link to a technical material selection websitewhere they can find detailed information about the technical prop-erties of materials. The reason for inclusion of the technical link isthe premise that it provides a good way to find out about thosematerials that deliver sensorial properties found to be significantin the overall evaluation.

2.2.2.6. The Meanings of Materials model. The Meanings of Materials(MoM) model provides designers with the key factors (or compo-nents) that play a central role in an individual’s materials deci-sions. These factors (material, product, user and context) arebriefly explained and the key aspects under each factor and thekey variables under each aspect are listed. Manufacturing pro-cesses, for instance, is presented under the product factor. A listof key processes that may affect the material’s aesthetics is offered.We expect the model to be used as a guideline and a checklist bydesigners to formulate their thoughts and ideas, and define themeaning evoking patterns that will lead to materials choices.

2.3. Summary of the MoM tool

An important difference of the MoM tool compared with exist-ing materials selection tools and methods is that it is proposed tobe a growing database, augmented by data generated from contin-

2938 E. Karana et al. / Materials and Design 31 (2010) 2932–2941

ual studies. In addition, the MoM tool is intended to be an inspiringinteractive alternative materials selection resource that has appealto product designers. Summarizing, the purpose of the tool is tosupport designers: (1) to understand the key variables in meaningattribution to materials and (2) to define the patterns behind a partic-ular material–meaning relationship. The tool offers visual and tex-tual inspiration for various types of products and materialsrelated to the intended meanings. Designers are encouraged tonavigate through the material examples and read the explanationsfor each meaning case.

There are four strong points of the tool expected to be founduseful by designers. First, designers need to summarize and docu-ment the findings of fieldwork related to a target group [15]. TheMoM tool is expected to answer this need by offering a collage ofmaterials alongside ratings of the materials against sensorialscales. Together these are expected to lead to easier discussionswith people involved in a design project, such as product designersfrom the project team, material suppliers or clients. Furthermore,the list of sensorial properties is expected to enhance designers’vocabulary about materials, which is expected to be helpful in sum-marizing and documenting a study’s results.

Second, the interpretation of the results is left to designers. Thismeans that although the quantitative data is presented foraddressing the high rated sensorial properties, designers are ex-pected to: (1) be stimulated by the collage of materials, (2) pickup the useful or relevant points for their designs with their ownintuition and creativity, and (3) define the meaning evoking pat-terns in their own way. We expect not only the similarities but alsothe differences between the selected materials to be inspiring fordesigners.

The last point is that the online application of the tool is ex-pected to provide a growing database. The outcome of every studyconducted to enhance the MoM tool is added to the tool’s database.Thus, designers might either navigate through the results of anexisting study, or request a new study to be performed. The nextsection focuses on the realization of the proposed tool. Data is gen-erated for a dummy application.

3. A study: data generation for a dummy application

This study was conducted with the aim of answering threequestions. First, are the tasks involved in using the [MoM] toolcomprehensible for people with different cultural backgroundsand expertise? Second, are people able to practically provide pic-tures of the materials they select? Third, can people explain theirmaterial selections in their own words and evaluate them usingthe sensorial scales? Above all, the goal of the study was to gener-ate data for the dummy application of the tool. Two studies (StudyA and Study B), each focusing on a different material meaning,were conducted simultaneously with 48 participants.

3.1. Method

3.1.1. MeaningsThe two meanings used for the studies were ‘elegant’ and ‘sexy’

(see Karana and Hekkert [20] for a further explanation on the rea-son behind the selected meanings).

3.1.2. ParticipantsTwo studies were conducted simultaneously. In the first (Study

A), 12 Dutch and 12 Turkish academic staff and university studentsfrom various departments of Dutch and Turkish Universities volun-teered to participate. The second study (Study B) was conductedwith 12 female and 12 male participants with different nationali-ties (Dutch, Turkish, Italian, Portuguese, Argentinean, Colombian,

Chinese, Tanzanian, German, South Korean, and Brazilian). Genderwas expected to create differences in attributing the meaning sexyto materials. All of the participants were either academic staff orstudents from various departments of Dutch Universities. The mainreason behind conducting this study within an academic environ-ment was that the (dummy) tool is intended to be universal/globalin its application, comprehensible by people from different nation-alities. For the same reason, English was adopted as the language ofthe tool. The pool of participants for this study consisted of stu-dents and staff with different nationalities and a sufficient levelof English. The studies took approximately half an hour perparticipant.

3.1.3. ProcedureThe studies were performed by electronic mail, with the partic-

ipants asked to perform a total of three tasks: TASK 1 to select amaterial, TASK 2 to document the selected material with a visual(picture), and TASK 3 to evaluate the selected material. In orderto prevent bias or unwarranted manipulation in the participants’selections, the tasks were requested iteratively and only once a re-sponse from the previous task had been received.

In Study A, the participants were asked to select a material thatthey thought was elegant. The participants were informed that thematerial they selected could be embodied in an object or in a partof an object. They were further explained that the objects could bemade of more than one material and that not all of those materialsneed have the same expression (meaning). It was particularlyemphasized that the focus should be on materials. Thus, if the se-lected material was a material of an object, the object as a wholeneed not have the same expression as the material. In addition,the terms ‘object’ and ‘material’ were defined and a few materialand object examples were included in the definition (see Figs. 4and 5).

When the participants informed us that they had found a suit-able material, they were sent the second task, in which they wererequested to provide visual evidence of the material they selected.They were informed that the visual evidence might be either pho-tographs they took or were taken by others, or illustrations (suchas photos, modeling) that they could find on the Internet, in mag-azines or similar sources. It was underlined that if the materialthey selected was a part of an object made of more than one mate-rial, an additional picture (a close-up) of that particular part wasrequired.

The third task focused on the evaluation of the materials thatthe participants had selected. There were two sub-tasks in thistask: TASK 3a (verbal explanation) and TASK 3b (evaluation againstthe sensorial scales). Having received the visuals of the materialsfrom the participants, they were then sent TASK 3a, which askedthem to explain in their own words why they thought that thematerial they selected was elegant. Next, the participants were sentthe final task (TASK 3b), consisting of five-point bipolar scales at-tached to the sensorial properties. The participants were asked toevaluate the sensorial properties of the material they had selectedon the given scales. The procedure for Study B was similar to StudyA, however the participants were asked to select a material thatthey thought was sexy.

3.2. Results

3.2.1. Study A: elegant materials selected by 12 Dutch and 12 Turkishparticipants

In total the participants selected 24 elegant materials (Fig. 9).The Dutch participants mainly focused on hard materials with nat-ural colours and smooth surfaces, such as ceramics and metal. TheTurkish participants selected more glossy and transparent materi-

Fig. 9. Elegant materials selected by 12 Dutch and 12 Turkish participants.

Table 2Results of multiple analysis of variance (MANOVA) for the effect of culture in Study A.

Dependent variable df F Sig.

CultureSoft 1 .071 .792Rough 1 .234 .633Glossy 1 .750 .396Reflective 1 1.641 .214Warm 1 .540 .470Elastic 1 .146 .706Transparent 1 5.453 .029Ductile 1 .122 .730Weak 1 .015 .902Heavy 1 .080 .780

E. Karana et al. / Materials and Design 31 (2010) 2932–2941 2939

als such as glass. Two flashy coloured (red and purple) materialsappeared in the Turkish participants’ selections.

The sensorial scales filled by the participants were analyzed sta-tistically both to see the most significant properties in attributingthe meaning elegant to materials in an overall evaluation and alsoreveal the main differences between the sensorial properties ofmaterials selected by Dutch and Turkish participants. A One Samplet-test was executed to compute the importance of the properties.The overall mean score for 10 items (M = 2.25) was taken as thetest value for the One Sample t-test. Bold items in Table 1 showthe properties that received scores significantly above or belowthe overall mean score. The properties that received scores signif-icantly below the overall mean score are presented with a minussign (�) in the table. The implication is that the opposite pole ofthese poorly scoring properties is significant for the given meaning.For instance, roughness (1.42) was significantly below (–) the over-all mean, therefore smoothness of a material would appear to beone of the most important properties in attributing the meaningelegant to materials. Elasticity (1.67) was also rated significantly

Table 1Results of the One Sample t-test for elegant materials.

Test value = 2.25 t df Sig. (2-tailed) Mean

Soft �.544 23 .592 2.08Rough (�) �4.920 23 .000 1.42Glossy 3.607 23 .001 3.46Reflective 2.106 23 .046 2.88Warm .742 23 .465 2.42Elastic (�) �2.723 23 .012 1.67Transparent �.377 23 .709 2.13Ductile �.119 23 .907 2.21Weak 1.141 23 .266 2.63Heavy .578 23 .569 2.42

below the overall mean score, which shows that the selected mate-rials were significantly rated as not-elastic. Glossiness (3.46) andreflectiveness (2.88) were rated significantly above the overallmean score (i.e. the selected materials were commonly glossyand reflective).

The effect of cultural differences on attributing the meaning ele-gant to materials was analyzed by a multiple analysis of variance(MANOVA) with the 10 sensorial properties as dependent variablesand culture as the fixed factor. One significant main effect of cul-ture (p < .05) was obtained for transparency (Table 2). Elegantmaterials selected by Turkish participants were more transparentthan the materials selected by Dutch participants (2.83 vs. 1.42).

3.2.2. Study B: sexy materials selected by 12 female and 12 maleparticipants

For the female participants, particularly the soft and velvet likefeelings of a material played an important role in attributing themeaning sexy. Three of the female participants mentioned the

Fig. 10. Sexy materials selected by 12 female and 12 male participants.

Table 3Results of the One Sample t-test for sexy materials.

Test value = 2.56 t df Sig. (2-tailed) Mean

Soft �.052 23 .959 2.54Rough (�) �6.226 23 .000 1.50Glossy (+) 3.408 23 .002 3.50Reflective .728 23 .474 2.79Warm �.922 23 .366 2.33Elastic (�) �3.098 23 .005 1.79Transparent �1.125 23 .272 2.21Ductile �.053 23 .958 2.54Weak .498 23 .623 2.71Heavy .072 23 .943 2.58

Table 4Results of multiple analysis variance for the effect of gender in Study Bb.

Dependent variable df F Sig.

GenderSoft 1 2.543 .125Rough 1 .232 .635Glossy 1 .815 .376Reflective 1 2.178 .154Warm 1 3.143 .090Elastic 1 .246 .625Transparent 1 1.468 .239Ductile 1 .703 .411Weak 1 9.757 .005Heavy 1 2.547 .125

2940 E. Karana et al. / Materials and Design 31 (2010) 2932–2941

importance of transparency (or semi-transparency) on sexiness. Incontrast, hard and strong materials dominated the male partici-pants’ selections. They hardly mentioned the importance of a softtactual property of a material for conveying the meaning sexy (ex-cept for one participant, who selected silk as a sexy material)(Fig. 10).

The same methods (One Sample t-test and MANOVA) were usedfor the statistical analysis of the sensorial scales filled by partici-pants. The results of the One Sample t-test, with bold items depict-ing those properties receiving scores significantly above or belowthe overall mean score (M = 2.56), are presented in Table 3. Rough-ness (1.50) and elasticity (1.79) were rated significantly below theoverall mean score, which reveals that the selected materials werecommonly rated as smooth and not-elastic. Glossiness (3.50) wasrated significantly above the overall mean score.

The results of the multiple analysis of variance (MANOVA), withgender as the fixed factor, revealed one significant main effect ofgender (p < .05) for weakness of materials (Table 4). Sexy materialsselected by the female participants were weaker than the materialsselected by the male participants (3.50 vs. 1.92).

4. Discussion

Designers select materials not only for physical benefits but alsoto convey their ideas and give character to their products. Productdesigners may have several questions regarding the intangible as-pects of materials that arise during materials selection, such as‘‘Does the selected material support the intended meaning of theproduct?’’, ‘‘Does it fit the target user group?’’, or ‘‘what kinds ofassociations can it evoke?’’. In order to convey their intentionsproperly, designers should understand how a material possessesits meaning in different products. In existing literature on materi-als and design, the significance of this intangible side of materialsis mentioned in various ways, such as: the second and third ordermaterials characteristics [21], emotive-stage materials characteris-tics, softer criteria of materials’ considerations, invisible character-istics [22], less tangible issues of materials [23], qualitative

properties [24,25], non-active or passive functions of materials[26], non-technical issues of materials [27], material image, meta-physical aspects of materials, non-physical properties of materials[28], material personality, personal dimensions of materials [29],intrinsic cultural Meanings Of Materials [30], expressiveness ofmaterials [31], subjective dimensions, essential and indicativecharacter of materials [32], and perceived characteristics of mate-rials [33]. Even though the importance of these aspects is empha-sized, only a few researchers [21,27,28,32,33] have conductedstudies on the subject and proposed ways of linking intangiblecharacteristics to designers’ materials selection processes. How-ever, there is currently no materials selection source that focuseson understanding and applying the intangible side of materialswithin product design [15,21–23,26,34–36].

This paper describes the development of a new materials selec-tion tool holding the idea of [meaning driven materials selection]to support designers to involve meaning considerations into theirmaterials selection processes. A study was conducted to createdata for a dummy application and to explore if the principle ofthe tool is well-understood by the people who provide data forthe tool.

Three main questions were addressed: first, are the tasks com-prehensible for people with different cultural backgrounds andexpertise? Almost all participants (44 out of 48) found the tasksclear and accomplished the tasks without questioning. Four partic-ipants found the first task particularly complicated. Two of themreplied to the first e-mail expressing their confusion and statedthat even though the main assignment (select a material that youthink is elegant/sexy) was clear, the explanation of the task confusedthem. They recommended splitting the explanation part and themain assignment, and send them as separate documents. Two par-ticipants found the main assignment too broad. They could notunderstand that they should select a specific material, until theyreceived the second task. Instead, they thought about a materialfamily that they found to be elegant/sexy, such as glass is sexy ormetal is elegant.

The second question was: are people able to provide pictures ofthe materials they select? All of the participants were able to do so.Thirteen participants selected materials that were embodied in apart of an object. However, only six of these sent an additional pic-ture showing the selected material in detail. All of the participantsemphasized the selected materials by name (such as ‘glass in thisobject’) or by referring to the relevant part of the object (such as‘the material of the handle’) in their verbal explanations (Task 3a).Those participants who did not send the detailed picture men-tioned that they found the pictures on the Internet and that itwas difficult and time consuming for them to edit the pictures. In-stead of taking their own photographs, most of the participantspreferred to surf the Internet for sourcing visuals of the materials.Nine participants (out of 48) sent photographs they had takenthemselves. Therefore, on the basis of these findings, in a futureapplication of the [MoM] tool the request for an additional photowill be left out. Instead, a more detailed explanation about the se-lected material (if it is a part of an object) will be required in Task3a. In general, the quality of pictures (pixel resolution) was suffi-cient for use in the [MoM] tool. Only four participants sent pictureswith low resolution.

The third question was: can people explain their selections intheir own words and make an evaluation using the sensorialscales? The participants were able to explain their selections intheir own words. They were willing to talk about their selectionsand their appraisals were comprehensible. Two participants inter-preted the sensorial scales incorrectly, using them to evaluate acertain material family rather than the specific material that theyhad selected. For this reason, instead of filling in the scales by con-sidering the selected material (such as a particular metal type

E. Karana et al. / Materials and Design 31 (2010) 2932–2941 2941

embodied in a vase) they made their ratings based on the materialfamily (‘metal’ in general). These two participants filled out thescales one more time after they were instructed about the purposeof the task.

For both of the meanings elegant and sexy, certain propertieswere found more effective than others for attributing the meaning.However, the variety of the selected materials provided an inspir-ing collage of properties for incorporation into further applicationsof the [MoM] tool. An increased number of participants would en-hance the chance of building an extensive collage, but would re-duce the number of common properties amongst the selectedmaterials. In certain cases, both common points and differencesin the properties of selected materials by a particular target groupcan be valuable for the designer. This statement will be tested inthe next study.

In general, the results of this study confirm that people with dif-ferent cultural backgrounds and expertise are able to fulfill thematerial–meaning related tasks. They can explain their choicesand understand the given sensorial scales supported visually bypictograms. Not surprisingly, it was noticed that the participantswith design backgrounds (such as industrial designers, architects,graphical designers) accomplished the tasks without difficulties.All questions arising from confusion about the first task came fromfour non-designers. The two people who declined to participate inthe study and who explained that they could not evaluate materi-als as ‘sexy’ or elegant’ were also non-designers. This could suggestthat the study would yield different results if it was conducted onlywith designers.

The main goal of the study was to provide sufficient data aboutthe meanings elegant and sexy for creating an example ‘outcome’part of the tool. Only one property for each meaning (transparencyfor elegant and weakness for sexy) appeared as the differences be-tween the specified groups (i.e. Dutch vs. Turkish, females vs.males). On the other hand, we could see some patterns (eventhough they were not that lucid) which evoke the given meaningsfor the specified groups. The materials that the participants se-lected, the selected user aspects (gender and culture) and the over-all results were satisfactory for creating a dummy application.

5. Conclusion

In this paper, we developed a tool that embraced [meaning dri-ven materials selection]. The main idea was that a designer whocan understand the relationships (or meaning evoking patterns)between materials and meanings can then systematically manipu-late meaning creation in materials selection processes. Accord-ingly, the Meanings of Materials tool was developed: (1) tofamiliarize designers with the main components (or aspects) ofthe Meanings of Materials model, (2) to show which aspects (undermain components) play an important role for certain meanings(such as sensorial properties, gender, culture, shape), and (3) tostimulate designers to find relationships (or patterns) betweenthese variables for given meanings. After developing the tool, weconducted a study for the aim of generating data for the dummyapplication, as well as testing the comprehensibility of the tasks in-volved in the tool. In a next study, we plan to test the usefulness ofthe MoM tool with the dummy application.

References

[1] Ashby MF. Materials selection in mechanical design. 3rded. Oxford: Butterworth-Heinemann; 2005.

[2] Cornish EH. Materials and the designer. New York: Cambridge UniversityPress; 1987.

[3] Farag MM. Selection of materials and manufacturing processes for engineeringdesign. UK: Prentice Hall; 1989.

[4] Ashby MF, Cebon D. The CES EduPack. In: Teaching engineeringmaterials. UK: Engineering Department, Cambridge University; 2007.

[5] Hekkert P, Schifferstein HNJ. Introducing product experience. In: SchiffersteinHNJ, Hekkert P, editors. Product experience. Amsterdam: Elsevier; 2008. p.1–8.

[6] Jordan PW. Designing pleasurable products: an introduction to the new humanfactors. London (UK): Taylor & Francis; 2000.

[7] Desmet P. Designing emotions. In: Industrial design engineering. Delft (TheNetherlands): Delft University of Technology; 2002.

[8] Schifferstein HNJ, Spence C. Multisensory product experience. In: SchiffersteinHNJ, Hekkert P, editors. Product experience. Elsevier: Amsterdam; 2008.

[9] Karana E, Hekkert P. Attributing meanings to materials. In: 6th internationalconference on design and emotion, Hong Kong: School of Design, The HongKong Polytechnic University; 2008.

[10] Karana E. Meanings of materials, PhD thesis. Design Engineering, Delft; 2009.ISBN 9789051550559: TU Delft.

[11] Blank P et al. Perceiving what painting express. In: Crozier WR, Chapman AJ,editors. Cognitive process in the perception of art. North Holland: Amsterdam;1984. p. 127–43.

[12] Ashby M, Johnson K. Materials and design: the art and science of materialselection in product design. Oxford: Butterworth-Heinemann; 2002.

[13] Karana E, Hekkert P, Kandachar P. Material considerations in product design: asurvey on crucial material aspects used by product designers. Mater Des2008;29:1081–9.

[14] Sonneveld M. Aesthetics of tactile experiences. Delft (The Netherlands): DelftUniversity of Technology; 2007.

[15] Van Kesteren I. Selecting materials in product design. Delft (TheNetherlands): Delft University of Technology; 2008.

[16] Karana E, Hekkert P, Kandachar P. Sensorial properties of materials for creatingexpressive meanings. In: Kansei engineering and emotion research conference,Sapporo, Japan; 2007.

[17] Govers PCM, Schoormans JPL. Intangible product attributes of watches: anexploratory study into the consensus of consumers’ reactions in 29th EMAC:marketing in the new millennium. Rotterdam: Erasmus UniversiteitRotterdam; 2002.

[18] Muller W. Order and meaning in design. Utrecht (The Netherlands): Lemma;2001.

[19] Karana E, Hekkert P, Kandachar P. Meanings of materials through sensorialproperties and manufacturing processes. Mater Des 2009(30):2778–84.

[20] Karana E, Hekkert P. User–material–product interrelationships in attributingmeanings. Int J Des, in press.

[21] Hodgson SNB, Harper JF. Effective use of materials in the design process–morethan a selection problem. In: International engineering and product designeducation conference, Delft, The Netherlands; 2004.

[22] Lefteri C. The branding of plastics – how important is the branding of amaterial and how far do plastics go in helping to define brands. In:International conference on the art of plastics design, Berlin, Germany; 2005.

[23] Conran S. Creating value – keynote address. In: International conference on theart of plastics design, Berlin, Germany; 2005.

[24] Edwards KL. Towards more strategic product design for manufacture andassembly: priorities for concurrent engineering. Mater Des 2002;23:651–6.

[25] Jee DH, Kang KJ. A method for optimal material selection aided with decisionmaking theory. Mater Des 2001;21:199–206.

[26] Deng YM, Edwards KL. The role of materials identification and selection inengineering design. Mater Des 2007;28:131–9.

[27] Ferrante M, Santos SF, de Castro JFR. Materials selection as an interdisciplinarytechnical activity: basic methodology and case studies. Mater Res2000;3(2):1–9.

[28] Ljungberg LY, Edwards KL. Design, materials selection and marketing ofsuccessful products. Mater Des 2003;24:519–29.

[29] Ashby M, Johnson K. The art of materials selection. Mater Today2003:24–35.

[30] Manzini E. The material of invention. London: Design Council; 1986.[31] Rognoli V, Levi M. Emotions in design through materials: an expressive-

sensorial atlas as a project tool for design of materials Turkey. In: 4thinternational conference on design and emotion, Ankara, Turkey; 2004.

[32] Rognoli V, Levi M. How, what and where is possible to learn design materials?In: International engineering and product design education conference, Delft,The Netherlands; 2004.

[33] Zuo H, Jones M, Hope T. Material texture perception in product design. In:International conference on the art of plastics design, Berlin, Germany;2005.

[34] Karana E. Guidelines for a materials selection source for industrial designactivity: a survey on the expectations of Turkish designers. In: Industrialdesign. Ankara (Turkey): Middle East Technical University; 2004.

[35] Lovatt AM, Shercliff HR. Manufacturing process selection in engineeringdesign, part 1: the role of process selection. Mater Des 1998;19:205–15.

[36] Sapuan SM. A knowledge-based system for materials selection in mechanicalengineering design. Mater Des 2001;22:687–95.