Content uploaded by Jan P.L. Schoormans
Author content
All content in this area was uploaded by Jan P.L. Schoormans on Jan 08, 2019
Content may be subject to copyright.
The development and testing of a product
personality scale
Ruth Mugge
1
,Faculty of Industrial Design Engineering, Delft University of
Technology, Landbergstraat 15, 2628 CE Delft, The Netherlands
Pascalle C. M. Govers
1
,Heineken Brewery, PO Box 530, 2380 BD
Zoeterwoude, The Netherlands
Jan P. L. Schoormans, Faculty of Industrial Design Engineering,
Delft University of Technology, Landbergstraat 15, 2628 CE Delft,
The Netherlands
Product personality refers to the set of personality characteristics that people
use to describe a specific product. Product personality can affect users’
interaction with and evaluation of a product. Accordingly, it may be desirable to
design products with a predetermined personality. In this research, the
authors develop a 20-item scale that can be used to validly and reliably assess
product personality based on product appearance during the design process. The
paper describes the subsequent steps that were taken during the process of
developing the scale.
Ó2008 Elsevier Ltd. All rights reserved.
Keywords: aesthetics, design tools, product design, product personality,
user behaviour
Product personality refers to the set of personality characteristics that
people use to describe a specific product and to discriminate it from
other products (Govers, 2004). For example, a Volkswagen Beetle
has a cheerful and friendly personality, whereas a Volkswagen Touareg is
dominant and tough. Thus, different products, even belonging to the same
brand, can be described to have diverse personalities. For designers, it is inter-
esting to study product personality for two reasons. First, Janlert and Stolter-
man (1997) discussed that product personality can have consequences for the
user’s interaction with the product. The product’s human-like characteristics
serve as an analogy for their behaviour and capabilities (Janlert and Stolter-
man, 1997; Aggarwal and McGill, 2007). Product personality can thus help
users to anticipate how to interact with a product. For example, a user will
handle a product with a delicate and sensitive personality with a great deal
of care and consideration. Second, product personality may affect user prefer-
ence. People prefer products with a personality that is similar to their own,
probably, because these products help them to confirm and express their
self-concept (Govers and Mugge, 2004; Govers and Schoormans, 2005).
Corresponding author:
Ruth Mugge
r.mugge@tudelft.nl
www.elsevier.com/locate/destud
0142-694X $ - see front matter Design Studies 30 (2009) 287e302
doi:10.1016/j.destud.2008.10.002 287
Ó2008 Elsevier Ltd. All rights reserved.
Past research showed that designers can purposely facilitate a particular per-
sonality in the product design based on their expertise (Govers et al., 2002).
Furthermore, people tend to agree on the personality characteristics of a par-
ticular product (Govers et al., 2002). However, a potential problem for the use
of product personality is that designers and users may perceive the meaning of
products differently (Hsu et al., 2000). Such a mismatch can negatively affect
users’ product evaluation. To ensure an intended product personality, it is
essential that the particular product personality that designers aim for during
the design process is correctly understood by users. A possible means to
achieve this is by systematically assessing the way users perceive the personal-
ity of a new product during the design process. The objective of the present
research is to develop a product personality scale that can be used for this
purpose.
1Personality and product design
Research shows that physical appearance features of humans (e.g., facial char-
acteristics, hair colour, beardedness, clothing, and wearing of glasses) have
a strong impact on the perception of a person’s personality (Jones, 1990; Bor-
kenau and Lieber, 1992a, b, 1995). In analogy, the appearance of a product is
a major determinant in the perception of product personality (Govers et al.,
2002; Jordan, 2002; Brunel and Kumar, 2007). Brunel and Kumar (2007) in-
deed found evidence that visual aesthetic characteristics, such as simplicity,
harmony, balance, unity, dynamics, timeliness/fashion, and novelty are linked
to perceptions of product personality.
Whereas product appearance is important for a product’s personality, other
aspects of the product design, such as sound, texture, and smell may be rele-
vant as well (Janlert and Stolterman, 1997). A car that has a tough appearance,
such as a Land Rover, may also sound tough while starting the engine or clos-
ing the door, and feel tough when holding the steering wheel. Personeproduct
interaction (e.g., forces, movements) may influence and enhance the personal-
ity of a product as well (Desmet et al., 2008). The same Land Rover may bring
about a tough personeproduct interaction if opening its door requires a rela-
tively great deal of force. In addition, computers can have personalities (e.g.,
dominant, submissive) as a result of variations in their communication style
with the user (Nass et al., 1995; Moon, 2002).
2Assessing product personality
To assess product personality, several scales have been used in the past. First
of all, during the product development process companies have used
a-theoretical, ad hoc scales, consisting of a set of personality characteristics
that are intuitively selected for the purpose of a particular study. Accordingly,
key personality characteristics may be missing in these scales, resulting in an
incomplete and unsatisfactory impression of the product’s personality. Several
scholars have used standardized human personality scales to assess the person-
288 Design Studies Vol 30 No. 3 May 2009
ality of products and brands (for an overview, Kassarjian, 1971; Sirgy, 1982;
Malhotra, 1988). This approach has been criticized, because these tests have
been developed and validated for measuring the personality of humans.
Though some personality characteristics may be mirrored in products, others
might not. As a result, the validity of such personality scales to study product
personality is questionable. In this respect, Kassarjian (1971) argued that ‘con-
sumer behaviour researchers must develop their own definitions and design
their own instruments to measure the personality variables that go into the
purchase situation’. Wells et al. (1957) and Malhotra (1981) both described
scales that should measure the personality of a product. However, both scales
are only tested on brands and have not been validated in the context of phys-
ical products. Furthermore, the stimuli in these studies are drawn from only
one product category (i.e., cars), which limits the scale’s generalizability across
product categories. As a result, these scales may not be appropriate to assess
the personality of all products. Jordan (2002) also presented a scale for study-
ing product personality. The main goal of this study was to show that product
personality is a meaningful concept for designers rather than scale develop-
ment. As a result, no quantitative tests were performed during the selection
process of the personality characteristics, which casts doubts on the scale’s re-
liability and validity.
In this paper, we present the development of a generalizable, comprehensive,
reliable, valid, and easy to use product personality scale to assess the person-
ality characteristics of individual products. In developing this scale, we ap-
plied the following principles. Firstly, the scale needs to comprise the full
scope of relevant personality characteristics. Secondly, the scale should allow
us to efficiently assess how people evaluate the product personality of prod-
ucts that belong to different categories of durable products. In the next par-
agraphs, we describe how we developed and tested the product personality
scale.
3Developing a product personality scale
To establish validity, several steps were taken during the development of the
scale to ensure that the final set of personality characteristics in the product
personality scale is comprehensive and representative.
3.1 Step 1: collecting an extensive pool of personality
descriptions
Our first step in developing the product personality scale was to collect an
extensive pool of personality descriptions. Existing scales that measure person-
ality associations in psychology, marketing, and design literature provided 348
descriptions to the pool (Asch, 1946; Wells et al., 1957; Norman, 1963; Rosen-
berg and Sedlak, 1972; Cattell et al., 1977; Malhotra, 1981; Anderson and
Klatzky, 1987; Zebrowitz, 1990; Carver and Scheier, 1996; Aaker, 1997; Ey-
senck, 1998; Jordan, 2002). To ensure that the pool of descriptions was
Development and testing of a product personality scale 289
complete, we also collected personality descriptions in a qualitative user study.
For this study (n¼48, 50% male, mean age ¼46), respondents were randomly
selected from a consumer household panel. Accordingly, this allowed us to
gain an overview of all personality characteristics that laymen use to talk
about products. All respondents were asked to describe 12 products belonging
to four product categories (three bottles of wine, three screwdrivers, three soap
dispensers, and three coffee makers) ‘as if they were a person’. To enhance the
generalizability of the resulting scale, we selected a variety of product cate-
gories that have different value to consumers. Accordingly, the product cate-
gories were chosen based on Ratchford’s (1987) think-feel dimensions and
have either symbolic (bottles of wine), utilitarian (screwdrivers), or both sym-
bolic and utilitarian value (soap dispensers and coffee makers) for consumers.
Furthermore, all product categories are mundane products that many people
use frequently. To generate substantial differences in product personality, two
experts (scholars at a faculty of industrial design engineering) selected those
product variants within each product category that represented the greatest
variety in appearances. In total, 953 descriptions were collected and added
to the pool. This resulted in a comprehensive pool of 1142 descriptions that
were all candidates for the final set of personality characteristics in the product
personality scale. Because this is a substantial number of descriptions, we as-
sume that it represents the full spectrum of personality characteristics. For
comparison, Aaker’s (1997) brand personality scale was based on an initial
sample of 309 descriptions.
All 1142 items were formulated in single adjectives. Adjectives are traditionally
seen as the primary terms of personality and are frequently used as items in
personality scales (Hofstee, 1990). Single adjectives were preferred over adjec-
tive pairs representing two opposites, because when using the latter the mean-
ing of the continuum may differ depending on the combination of the
adjectives. For example, the meaning of an item scale ranging from ‘friendly’
to ‘reserved’ differs from the meaning of an item scale ranging from ‘friendly’
to ‘hostile’. This issue may become a problem with respect to personality char-
acteristics, because not every personality characteristic has a clear negative
counterpart (Hofstee, 1990). In addition, we decided to use single adjectives
(‘friendly’) rather than unipolar adjective pairs (‘friendly’ vs. ‘not friendly’).
Although unipolar adjective pairs assume to represent the absence or presence
of one personality characteristic, ‘not friendly’ does not represent the absence
of ‘friendly’. In fact, it represents its negative counterpart: the presence of ‘un-
friendliness’. In conclusion, we believe that the use of single adjectives that can
be rated as either ‘descriptive’ or ‘not descriptive’ for a target product is least
open to bias.
3.2 Step 2: reducing the number of personality descriptions
During the second step, we needed to reduce the 1142 descriptions to a man-
ageable number. In this reduction process, we used four procedures that
290 Design Studies Vol 30 No. 3 May 2009
succeeded each other. As recommended in the literature, we used expert judge-
ment in the early stages to reduce the number of items, and lay judgement in
the final stages to make the definitive selection (Hofstee, 1990; Devellis, 1991).
Firstly, six judges independently evaluated all the descriptions and selected
only those that were considered without any doubt to be human personality
characteristics. All judges were scholars in marketing and consumer behaviour
at a faculty of industrial design engineering. Consequently, these experts pos-
sessed the required knowledge of psychology and design to complete this task
well. Hundred and fifty-four descriptions were not categorized by any of the
judges as a personality characteristic. Because redundancy was considered to
be an asset rather than a liability in the current stage, it was decided to use
the most conservative elimination strategy possible. Consequently, we only
eliminated these 159 descriptions. Secondly, the remaining pool of 988 descrip-
tions was reviewed by the same judges (excluding one judge) with respect to
similarity of meaning. Descriptions that described the same personality char-
acteristic were grouped together. For every group, one item was selected that
represented the personality descriptions of this group best. In doing so, we
kept the comprehensive character of the pool. This second reduction resulted
in a pool of as many as 458 descriptions.
The pool was further reduced by isolating the most relevant personality de-
scriptions with respect to both people and products. In this third procedure, re-
spondents (n¼156, 51% male, mean age ¼21) were asked to indicate for
a subset of 76 descriptions whether they did or did not use a description to ex-
plain the personality of a person or a product. Because we wanted to reduce the
chance of falsely excluding personality characteristics that are used to describe
products, all respondents were students at the faculty of industrial design engi-
neering. Design students are trained to put their impression of products into
words and have developed a much broader vocabulary for describing products
than ordinary people. The chance that a personality characteristic was wrong-
fully identified as irrelevant for describing products was thus minimized.
For each description, we calculated the percentage of respondents that used
this description to describe either people or products. Different cut-off percent-
ages were set for people and products, because these criteria differ in impor-
tance. Including descriptions that are not used to describe the personality of
people would undermine the validity of the product personality scale. The
measure would not measure what was intended: the personality of products
described with personality characteristics that are also used to describe people.
The inclusion of items that are not used to describe the personality of products
does not affect the scale’s validity (because these items would be rated as ‘not
descriptive’ in the final scale), but would make the scale needlessly long. Be-
cause the first criterion is more important than the second, the cut-off percent-
ages were set at 85% and 60%, respectively. The resulting selection of
descriptions meets the defining component of product personality in that
Development and testing of a product personality scale 291
they are used with respect to people and products. However, these criteria do
not rule out the possibility that descriptions, though used with respect to both
people and products, do not measure the personality of a product. Some de-
scriptions can be interpreted literally when used with respect to products.
For example, ‘strong’ is a personality characteristic that can be used to express
that a person is not easily upset. Yet, its primary interpretation becomes ‘not
easily broken’ when it is applied to a product. Three judges (scholars at a fac-
ulty of industrial design engineering) checked whether these descriptions could
be interpreted literally. This resulted in the elimination of seven descriptions.
Our item reduction process resulted in 78 personality characteristics.
3.3 Step 3: defining and testing the product personality scale
In most cases, the product personality scale will be used to test the personality
of several products simultaneously. Comparing multiple products helps de-
signers to understand why certain products are perceived as having a particular
personality and enable them to select the best alternative. However, this will
constrain the number of descriptions that can be included in the scale. Rating
a number of products on 78 items will take respondents substantial amounts of
time and effort. For this reason, other personality scales that are constructed
to measure the personality of products or brands consist of fewer items. For
example, Malhotra’s (1981) personality scale consists of 15 items, Jordan’s
scale to measure the personality of products consists of 17 items, and Aaker’s
brand personality scale consists of 42 items. Accordingly, it would be worth-
while to perform another reduction procedure.
Past research on human personality showed that the presence of certain per-
sonality characteristics almost always go together with the presence of other
personality characteristics (Zebrowitz, 1990). For example, a person that is
perceived as ‘cheerful’ is often also perceived as ‘happy’. If these personality
characteristics are always perceived together, then ‘happy’ does not provide
much extra information. In a sense, the two characteristics describe more or
less the same thing. The same might hold for describing the personality of
products. This suggests that it is possible to further reduce the set of character-
istics without loosing valuable information. To decrease the number of de-
scriptions, we explored to what degree certain descriptions co-occur in the
perception of product personality. Based on the groupings of descriptions,
the final set of personality characteristics can be determined. An extensive
quantitative study was developed to achieve this goal. Furthermore, we used
this study to check the scale’s reliability and validity. The following para-
graphs describe this next study in detail.
3.3.1 Stimuli and procedure
To enhance scale generalizability, the stimuli for this quantitative study were
selected from two product categories with either high or low symbolic value.
To choose these product categories, a pre-test was performed in which
292 Design Studies Vol 30 No. 3 May 2009
students (n¼14) were asked to specify several publicly consumed products
with high symbolic value and several privately consumed products with
low symbolic value. Cars were listed most frequently as a good representative
of a highly symbolic, publicly consumed product (93%). Scissors (43%), sta-
plers (43%), and vacuum cleaners (29%) were listed most often as examples
of privately consumed products with low symbolic value. We selected cars
and vacuum cleaners, because these products can be bought in a great variety
of appearances, which is essential for a reliable scale development. The stim-
uli were selected from the assortment of cars and vacuum cleaners at a mid-
range price in the market. Two experts (scholars at a faculty of industrial
design engineering) selected those cars and vacuum cleaners that represented
the greatest variety in appearances, while still providing a realistic picture of
the alternatives within the price range. This variety was necessary to ensure
that our scale can be applied to different products. Furthermore, lack of di-
versity in appearance could cause items to be paired together for the wrong
reasons, because they are consistently rated as ‘not descriptive’. In total,
eight cars and eight vacuum cleaners were selected. All stimuli were illus-
trated with a high-quality colour picture of the product: the pictures of the
cars showed the front and side of the cars and the pictures of the vacuum
cleaners showed the whole vacuum cleaner, including the tube and nozzle.
Because rating each of the 16 stimuli on 78 personality characteristics would
be too time-consuming, the stimuli were divided into four sets, each contain-
ing four stimuli belonging to the same product category. Respondents rated
the stimuli of one set. Within each set, the stimuli were presented in four bal-
anced orders (Maxwell and Delaney, 1990). Students from 38 different BSc
and MSc programs (n¼125, 34% male, mean age ¼24) filled out the ques-
tionnaire, resulting in a total of 28e33 respondents for each of the sets. Be-
cause lay judgement is preferred for the last step in scale development,
students of industrial design engineering were excluded from participation.
The respondents were asked to rate the extent to which the 78 personality
characteristics describe the four stimuli (products) in their set using a five-
point scale (1 ¼‘not descriptive’, 5 ¼‘descriptive’). This implies that each re-
spondent had to make 312 ratings, which took approximately 30 min to com-
plete. To control for possible fatigue effects, multivariate analyses of variance
(MANOVA) were performed for each of the four sets using the 78 personal-
ity characteristics as dependent variables and stimuli and order as the inde-
pendent variables. No significant interaction effects between stimuli and
order were found (all ps>0.05), suggesting that the order in which the dif-
ferent products were presented did not affect the rating on the personality
characteristics.
3.3.2 Results
3.3.2.1 Defining the product personality scale. The data of this quantita-
tive study was analyzed using Ward’s hierarchical clustering method (Hair
et al., 1998). Cluster analysis classifies objects into clusters so that each object
Development and testing of a product personality scale 293
is very similar to the other objects in that cluster. If certain sets of descriptions
co-occur in the descriptions of different products and by different respondents,
they have more or less the same meaning for the respondents. Therefore, these
descriptions can be replaced by one description out of such a set without losing
valuable information.
In order to assure that the cluster solution is representative of the general
population, and is thus generalizable to other objects (e.g., other products
or respondents), the cluster solution needs to be validated. A common ap-
proach in this respect is to randomly split the sample into two groups and
analyze these groups separately (Hair et al., 1998, p. 501). Instead of a ran-
dom split in two groups, we decided to validate our cluster solution by com-
paring four specific subgroups (cars, vacuum cleaners, men, and women).
These specific groups are more likely to differ than two randomly selected
groups, which strengthens the generalizability of our findings. Specifically,
we used gender as a selection criterion, because men and women may differ
in their evaluation of personality characteristics that are more masculine
(e.g., aggressive, dominant, tough) or more feminine (e.g., cute, romantic,
sensitive). After deleting only three descriptions (i.e., vulnerable, intelligent,
and annoying), all cluster solutions showed a similar pattern of six large
groups of descriptions. The three descriptions were deleted, because their po-
sition changed between the four cluster solutions, indicating that their mean-
ing is not similar for all product categories and respondents. Table 1 provides
a summary of the personality characteristics belonging to the six groups.
Cronbach’s alphas of these groups all exceed 0.80, which strongly supports
internal consistency.
An analysis of the grouping suggested that reducing the pool to only six de-
scriptions would not do justice to the broad scope of meaning represented
by the descriptions in each group. For example, group 1 includes the person-
ality characteristics ‘cheerful’, ‘cute’, and ‘easy-going’. Although these person-
ality characteristics may sometimes go together in a product, it is possible to
think of product designs that are perceived as ‘cute’, without being ‘cheerful’
or ‘easy-going’ at the same time. Accordingly, we analyzed the six groups
again, separately, using Ward’s hierarchical clustering. The results showed
that the six general groups of descriptions could be divided into 20 subgroups,
containing two to seven descriptions each. The descriptions in all subgroups
have related meanings. For example, the descriptions ‘provocative’, ‘wild’,
‘exuberant’, and ‘eccentric’ form one subgroup.
From each of these 20 subgroups, a description was chosen. The selection of
these representing descriptions was based on two criteria. Firstly, the descrip-
tions had to cover the meaning of the other descriptions in the subgroup. Sec-
ondly, the descriptions should have a high item-to-total correlation, which
offers statistical support for the representation of the total subgroup (Hair
294 Design Studies Vol 30 No. 3 May 2009
et al., 1998). Table 2 presents the 20 descriptions that were selected as items for
our final product personality scale.
To check whether all 20 descriptions are important to differentiate products
from each other, a MANOVA was performed with the personality descrip-
tions as the dependent variables and the stimuli as the independent variable.
A significant effect of the stimuli was found (F(315, 7065) ¼3.80, p<0.001).
Furthermore, univariate tests revealed significant differences among the eight
stimuli for all 20 descriptions (all ps<0.001), providing support that the final
product personality scale only consists of personality descriptions that may
play a role in products.
The final set of 20 personality descriptions includes two descriptions that were
exclusively taken from the literature, nine descriptions that were originally
taken from the literature, but were confirmed by the qualitative user study
in step 1, and nine descriptions that were only found in the qualitative user
study. A similar proportion was found for the 78 personality descriptions, re-
sulting from step 2. Furthermore, this proportion corresponds to that of the
comprehensive pool of 1142 descriptions and shows that both the literature
study and the qualitative user study were necessary to collect a comprehensive
pool of descriptions.
Table 1 Summary of the six groups of personality characteristics
Attractive Aggressive Bourgeois Careless Conspicuous Aloof
Casual Dominant Businesslike Chaotic Eccentric Boring
Charming Excessive Calm Childish Exuberant Cheerless
Cheerful Obtrusive Consistent Corny Funny Insular
Cute Showy Decent Creepy Idiosyncratic Masculine
Easy-going Honest Immature Interesting Old-fashioned
Energetic Inconspicuous Odd Lively Reticent
Feminine Mature Pathetic Provocative Strict
Flexible Modest Silly Tough Unattractive
Friendly Precise Unreliable Wild Uninteresting
Happy Predictable Untidy
Informal Reliable
Nice Sensible
Open Serious
Pleasant Well-groomed
Popular
Pretty
Relaxed
Romantic
Sensitive
Sweet
Terrific
Young
a¼0.96 a¼0.83 a¼0.88 a¼0.84 a¼0.91 a¼0.89
Development and testing of a product personality scale 295
3.3.2.2 Testing the reliability and validity of the 20-item product personal-
ity scale. The value of a measurement scale highlydepends on the reliability and
validity of the scale. To be reliable, respondents need to agree about the applica-
bility of a personality characteristic with respect to different products. We
checked the reliability of the 20-item scale by calculating Cronbach’s alpha for
every selected description over the different products (see Table 2). The results
showed that all descriptions had an adequate to a very high reliability score.
In addition to our statistical analysis on the reliability of the scale, we checked
the scale’s face validity. Face validity indicates to what degree a scale appears
to be a good measure of the concept one intends to measure. We checked face
validity by developing the personality profiles of the products that were in-
cluded in the study. A personality profile represents how respondents per-
ceived a product on the 20 personality characteristics. If the product
personality scale has face validity, one should be able to relate the personality
profile of a product to its appearance in a logical way. Figure 1 displays the
personality profiles of two cars based on the data of our study. These two pro-
files clearly indicate how the two products differ from each other with respect
to specific personality characteristics. It is possible to relate the specific profiles
of these two cars to their appearance and explain the similarities and differ-
ences between the two profiles with similarities and differences between the ap-
pearances of both products in a reasonable way. For example, car 1 is
perceived as ‘serious’, ‘aloof’, and ‘boring’, whereas car 2 is not. This is, among
Table 2 Reliability of the selected items (n[33)
Cronbach’s alpha
Cheerful 0.96
Open 0.93
Relaxed 0.89
Pretty 0.93
Easy-going 0.94
Cute 0.95
Dominant 0.89
Obtrusive 0.90
Silly 0.69
Childish 0.91
Untidy 0.78
Idiosyncratic 0.94
Interesting 0.88
Lively 0.91
Provocative 0.92
Modest 0.87
Honest 0.67
Serious 0.95
Aloof 0.91
Boring 0.95
296 Design Studies Vol 30 No. 3 May 2009
other things, the result of car 1’s basic, robust form and its grey colour. In con-
trast, car 2 is perceived as ‘cheerful’, ‘idiosyncratic’, and ‘pretty’, whereas car 1
is not. This is, among other things, the result of car 2’s compact, spherical
shape, the curved grill and the round shape of its head lights. In addition,
from the analyses of the personality profiles of all the products in the study
one can see that products that are described with the same personality charac-
teristics share certain appearance characteristics, even when these products be-
long to different product categories. For example, the car and vacuum cleaner
that were rated as ‘serious’ are both grey and have basic, robust forms
(see Figure 2). Taken together, these results indicate that our scale has face
validity.
4Discussion
The objective of this research was to develop a reliable, valid, and generaliz-
able scale for assessing product personality. We started with a comprehensive
set of 1142 personality descriptions, which was reduced to a collection of 78
descriptions. A series of cluster analyses were used to come to a more manage-
able number of items. The results of these cluster analyses indicated that peo-
ple can comprehensively describe product personality with 20 product
personality descriptions. Next, we showed that a 20-item product personality
scale that is based on these descriptions is reliable and has face validity. Con-
sequently, we conclude that our 20-item product personality scale assesses
Figure 1 Personality profiles
of two cars
Development and testing of a product personality scale 297
product personality based on product appearance in a comprehensive, reli-
able, valid and efficient way. The product personality scale thus offers de-
signers and marketers of consumer products an alternative to the present ad
hoc scales.
The product personality scale can be used to compare either different products
within a specific product category or products from different product cate-
gories. Comparing different products within a category is valuable to evaluate
a product in relationship to the competitors or to select the alternative that
profoundly communicates a desired personality. A comparison among differ-
ent product categories can help designers to create an entire line of products
(e.g., line of kitchen appliances) that should fit together with respect to their
product personality. As discussed, products with a similar personality profile
share appearance characteristics, suggesting that it is possible to create a con-
sistent product line by taking into account product personality. Furthermore,
comparing products from different product categories may enable designers to
uncover benchmark products in different product categories and, conse-
quently, gain a deeper understanding of the appearance characteristics that
result in particular product personalities. Such a database of benchmark prod-
ucts with particular personalities may serve as a source of inspiration in the de-
sign process.
Designers should keep in mind that all personality characteristics can play
a role in determining a product’s personality. Only by using the complete
20-item product personality scale, one can gain insight in the full profile of per-
sonality characteristics for a specific product. However, this does not imply
that all personality characteristics are equally important in all situations.
Figure 2 A car and vacuum cleaner with a serious personality
298 Design Studies Vol 30 No. 3 May 2009
The importance of particular personality characteristics will depend on the
characteristic’s centrality to the person’s self (Markus, 1977). If a person
sees himself/herself first and foremost as a serious person, (s)he will focus pri-
marily on products that communicate the personality characteristic ‘serious’.
In order to achieve a proper balance between the product personality scale’s
comprehensiveness and its efficiency, we reduced the scale to a set of 20 items.
Of course, there is a potential risk in reducing the number of items, because
this may negatively affect the richness of the total product personality scale.
However, to develop a scale that is easy to administer and that can be used
to rate different stimuli in one study without loosing interest, it is important
that redundant items are deleted. In our research, the reduction of items was
performed with great care by only taking out those personality characteristics
that are unimportant for products or that overlap with other personality char-
acteristics. As a result, the width of personality characteristics that are impor-
tant for products is preserved. In conclusion, we believe that the product
personality scale presents a complete overview of the personality characteris-
tics that can play a role in products.
Some other limitations of our research should be mentioned. First, the scale
was developed in the Netherlands. Therefore, the items gathered from litera-
ture had to be translated from English into Dutch. In order to make sure
that the Dutch translation had the same meaning as the English original, we
followed a two-way translation procedure suggested by Mullen (1995). Al-
though this procedure is best practice, chances exist that our product person-
ality scale is sensitive to language differences.
Another limitation of our study is that we limited our research to product ap-
pearance. Janlert and Stolterman (1997) argued that designers should pay at-
tention to the consistency and coherence of all facets of the product design,
when designing a product with a certain personality. As a result, future re-
search should verify whether the product personality scale can be used to as-
sess the personality of product interaction or the sound and texture of the
product. In this respect, Desmet et al. (2008) found that the personality char-
acteristics elegant and gentle may be particularly valuable to describe the per-
sonality of a product’s physical interaction style. In addition, more knowledge
is needed on the combined effect of visual and other design facets to uncover
their potential interactive effects.
Although Brunel and Kumar (2007) provided a first step in linking visual
product design aesthetics to product personality perceptions, more research
is also needed to fully understand the effect of specific product characteristics
on the perception of a product’s personality. A systematic use of the product
personality scale over time can be used to gain knowledge about the relation
Development and testing of a product personality scale 299
between specific product characteristics and the user’s perception of product
personality.
The product personality scale can help designers in different ways. The prod-
uct personality scale may serve as inspiration for designers, because it offers
them a complete overview of the personality characteristics that can play
a role in products. Nevertheless, this does not imply that the creation of a spe-
cific personality characteristic in a product is straightforward and that the
product personality scale may serve as a recipe for creating products with pre-
determined personalities. In fact, designing products with a predetermined
personality is a complex and creative design task. In addition to clearly recog-
nizable product aspects (as discussed in the validation section), small and in-
conspicuous details of the product design can play an important role as
well. Therefore, designers need to design all relevant product aspects in such
a manner that the whole product is perceived as having the desired personality.
In addition, some personality characteristics, such as relaxed and honest, may
be difficult to identify in products. For those personality characteristics, a dis-
crepancy between designers and consumers in their perception of the product
is likely to occur (Hsu et al., 2000). As such, designers may also use the scale to
verify their expertise on product personality during the design process by test-
ing whether the intended personality characteristics are indeed recognized by
consumers.
References
Aaker, J (1997) Dimensions of brand personality, Journal of Marketing Research
Vol 34 No 3 pp 347e356
Aggarwal, P and McGill, A L (2007) Is that car smiling at me? Schema congruity
as a basis for evaluating anthropomorphized products, Journal of Consumer
Research Vol 37 No 4 pp 468e479
Anderson, N H and Klatzky, R L (1987) Traits and social stereotypes: levels of
categorization in person perception, Journal of Personality and Social Psychology
Vol 53 pp 235e246
Asch, S E (1946) Forming impressions of personality, Journal of Abnormal and
Social Psychology Vol 41 pp 258e290
Borkenau, P and Lieber, A (1992a) The cross-modal consistency of personality:
inferring strangers traits from visual or acoustic information, Journal of Research
in Personality Vol 26 pp 128e204
Borkenau, P and Lieber, A (1992b) Trait inferences: sources of validity at zero
acquaintance, Journal of Personality and Social Psychology Vol 62 pp 645e657
Borkenau, P and Lieber, A (1995) Observable attributes as manifestations and
cues of personality and intelligence, Journal of Personality Vol 63 pp 1e25
Brunel, F F and Kumar, R (2007) Design and the Big Five: linking visual product
aesthetics to product personality in G Fitzsimons and V Morwitz (eds) Advances in
consumer research, Association for Consumer Research Vol 34 pp 238e239
Carver, C S and Scheier, M F (1996) Perspectives on personality Allyn and Bacon,
Needham Heights
Cattell, R B, Eber, H W and Tatsuoka, W (1977) Handbook for the personality
factor questionnaire IPAT, Champaign
300 Design Studies Vol 30 No. 3 May 2009
Desmet, P, Ortiz Nicolas, J C and Schoormans, J P L (2008) Product personality
in physical interaction, Design Studies Vol 29 No 5 pp 458e477
Devellis, R F (1991) Scale development: theory and applications Sage, Newbury
Park
Eysenck, H J (1998) Dimensions of personality Transaction Publishers, New York
Govers, P C M, Hekkert, P and Schoormans, J P L (2002) Happy, cute and tough:
can designers create a product personality that consumers understand? in
D McDonagh,P Hekkert,J Van Erp and D Gyi (eds) Design and emotion: the
experience of everyday things, Taylor and Francis, London pp 345e349
Govers, P C M (2004) Product personality Delft University of Technology, Delft
Govers, P C M and Mugge, R (2004) ‘I love my jeep, because its tough like me’: the
effect of productepersonality congruence on product attachment in A Kurtgo
¨zu
¨
(ed) Proceedings of the Fourth International Conference on Design and Emotion,
Ankara, Turkey
Govers, P C M and Schoormans, J P L (2005) Product personality and its influence
on consumer preference, Journal of Consumer Marketing Vol 22 No 4 pp 189e197
Hair, J F, Anderson, R E, Tatham, R L and Black, W C (1998) Multivariate data
analysis (5
th
edn) Prentice Hall, Upper Saddle River, NJ
Hofstee, W K B (1990) The use of everyday language for scientific purposes,
European Journal of Personality Vol 4 pp 77e88
Hsu, S H, Chuang, M C and Chang, C C (2000) A semantic differential study of
designers’ and users’ product form perception, International Journal of Industrial
Ergonomics Vol 25 No 4 pp 375e391
Janlert, L E and Stolterman, E (1997) The character of things, Design Studies Vol
18 No 3 pp 297e314
Jones, E E (1990) Interpersonal perception W.H. Freeman and Company, New
York
Jordan, P W (2002) The personalities of products in W S Green and P W Jordan
(eds) Pleasure with products: beyond usability, Taylor and Francis, London pp
19e47
Kassarjian, H H (1971) Personality and consumer behavior: a review, Journal of
Marketing Research Vol 8 No 4 pp 409e418
Malhotra, N K (1981) A scale to measure self-concepts, person concepts, and
product concepts, Journal of Marketing Research Vol 18 No 4 pp 456e464
Malhotra, N K (1988) Self-concept and product choice: an integrated perspective,
Journal of Economic Psychology Vol 9 No 1 pp 1e28
Markus, H (1977) Self-schemata and processing information about the self, Jour-
nal of Personality and Social Psychology Vol 35 No 2 pp 63e78
Maxwell, S E and Delaney, H D (1990) Designing experiments and analyzing data
Wadsworth, Belmont, CA
Moon, Y (2002) Personalization and personality: some effects of customizing mes-
sage style based on consumer personality, Journal of Consumer Psychology Vol 12
No 4 pp 313e326
Mullen, M R (1995) Diagnosing measurement equivalence in cross-national
research, Journal of International Business Studies Vol 26 pp 573e596
Nass, C I, Moon, Y, Fogg, B J, Reeves, B and Dryer, D C (1995) Can computer
personalities be human personalities, International Journal of HumaneComputer
Studies Vol 43 No 2 pp 223e239
Norman, W T (1963) Toward an adequate taxonomy of personality attributes:
replicated factor structure in peer nomination personality ratings, Journal of
Abnormal and Social Psychology Vol 66 pp 574e583
Ratchford, B T (1987) New insights about the FCB grid, Journal of Advertising
Research Vol Aug/Sept pp 24e38
Development and testing of a product personality scale 301
Rosenberg, S and Sedlak, A (1972) Structural representatives of perceived
personality trait relationships in A K Romney,R N Shepard and S B Nerlove
(eds) Multidimensional scaling: theory and applications in the behavioral sciences,
Seminar Press, New York pp 134e162
Sirgy, M J (1982) Self-concept in consumer behavior: a critical review, Journal of
Consumer Research Vol 9 No 3 pp 287e300
Wells, W D, Andriuli, F J, Goi, F J and Seader, S (1957) An adjective check list for
the study of ‘‘product personality’’, Journal of Applied Psychology Vol 41 No 5 pp
317e319
Zebrowitz, L A (1990) Social perception Brooks Cole, Pacific Grove, CA
1. Both authors contributed equally to the research and are listed in inverse alphabetical
order.
302 Design Studies Vol 30 No. 3 May 2009
