The purpose of the free sorting method is not to uncover underlying cognitive processes, but is a means to discover factors that are likely to result from the list of traits from a consumer’s perspective without any contamination from the researcher’s preconceptions (Rosenberg and Kim, 1975). In essence, the basic idea behind the sorting task is to ask a small sample of consumers to sort the 45 traits into groups; the groups are then chosen and named by respondents. Therefore, the sorting task helps identify relevant categories by investigating commonality and differences between consumers in the use of that categorization. Furthermore, it is used to eliminate traits that commonly cannot be categorized in any groups the consumers formulated.
4.6.1 Respondents
Nine consumers agreed to participate in this study which roughly had an equal number of males (N=5) and females (N=4). Participants that took part in this task ranged from secretaries, undergraduate students to PhD holders. Respondents were instructed to categorize the cards by creating mutually exclusive piles comprised of conceptually similar statements. Thus, statements in the same pile were more conceptually similar to each other compared with those that made up the other piles. Participants were also encouraged to bind the cards with paper clips to ensure accurate recording of traits in each pile.
4.6.2 Sorting Task Procedure
The card sorting task conducted in this study is grounded in Kelly’s personal construct theory that utilizes different types of objects or stimuli (for example, pictures, personality traits and colors) (Rosenberg, Nelson, and Vivekananthan, 1968; Rosenberg and Kim, 1975; Johnston, 1995; Green and Manzi, 2002; Fincher and Tenenberg, 2005; Rugg and McGeorge, 2005). According to Rosenberg and Kim (1975: 489), “[i]n a typical application of the sorting method, the respondent is asked to partition a set of inter-related objects or terms into different groups on the basis of their ‘similarity,’
‘relatedness,’ or ‘co-occurrence’ depending on the particular ‘application’”. In line with the application of the card sorting task, stimuli in the form of traits and definition of the traits were presented in a card format to respondents. The respondents were asked to partition the cards (which had the traits and definitions) into groups they felt the traits could be categorized into. No predefined categories or number of categories were given to consumers; instead, they were
~ 144 ~
encouraged to formulate as many categories as they felt were necessary. This approach is consistent with what Giguere (2006) named as a free card sorting format. Further, instructions were given to ensure each category was homogenous and coherent. Once categories were formulated, the respondents were encouraged to name the category. Traits which respondents were unable to categorize provided an indication of irrelevant Negative Brand Personality traits.
4.6.3 Results of the Sorting Task
After consumers freely categorized traits into piles (factors), it was apparent that five categories had emerged based on what traits consumers put together under a single pile. However, the name of each pile differed from respondent to respondent. Table 4.6 displays the category names that respondents came up with. As a result, the traits in each category were first collated together to list the traits in which each of the 9 respondents had 50% or more agreement. The name of the piles were categorized together to assess the similarity of the category group name to reflect the respondents’ group labels.
~ 145 ~
Table 4.6: Details of the five main categories that emerged from the Sorting Task alongside the names of each of the groups identified by respondents.
Name of Categories Traits
GROUP ONE (Egotistical)
High Self Opinion Pompous Pretentious Vain Snobby Stubborn
Egotistical Pompous Brash Vain Judgmental Flaunt Tyrant
Resentment Pompous Immoral Unethical Antisocial Snobby Tyrant
Selfish Pretentious Vain Arrogant Flaunt Fake
Aloof Coarse Stubborn
Self-centered Pompous Selfish Vain Judgmental Snobby Tyrant
Flaunt brash Pretentious Superficial Manipulative Intimidating Stubborn Arrogant
Superior Stubborn Selfish Vain Arrogant Snobby Tyrant
Pretentious Judgmental Manipulative Intimidating
Self-important Pompous Selfish Vain Arrogant Snobby Aloof
Sad Dull Boring Monotonous Superficial Mischievous Cheap
Unpredictable Dull Boring Monotonous Superficial Predictable Cheap
GROUP THREE (Socially Irresponsible)
Low Minded Deceiving Unethical Immoral Rebellious Snobby Repulsive
Selfish Arrogant Stubborn Mischievous
Bad Faith Deceiving Unethical Immoral Fake
Deviant Manipulative Anti-Social Mischievous GROUP FOUR (Critical)
Anti- establishment Barbarian Rebellious Deviant Tyrant Antisocial Judgmental Selfish Repulsive Rebellious Mischievous Predictable Cheap Coarse
Confused Judgmental
Low minded Repulsive Rebellious Mischievous Stubborn Arrogant Snobby Immoral Selfish Judgmental
Unclear Immoral Rebellious Selfish Mischievous Vain Weird
Lonely Confused Unstable Naive Aloof Judgmental
Forceful Stubborn rebellious Tyrant Judgmental
Envious Repulsive Selfish Superficial Unstable Pretentious Predictable
Vain Mischievous Weird Judgmental
Critical Repulsive Stubborn Rebellious Judgmental
GROUP FIVE (Lacking Logic)
Irrational/Lacking Logic Weird Delusional Unstable Absurd Naive Superficial
Unreal Fake Delusional Predictable Superficial
Different Weird Delusional Unstable Absurd Deviant
Ingenuous Lonely Delusional Naïve Superficial
Unusual Weird Rebellious Mischievous Absurd Deviant Superficial
Flimsy
Shallow Mindedness Weird Delusional Unstable Superficial Naive Pretentious Confused Lonely Monotonous
Different Brash Coarse Naïve Absurd
~ 146 ~
Table 4.6 illustrates the commonalities of traits under each factor; although some traits are commonly categorized in each factor, there still remains some variance in some of the factors. For example, in Group One, nearly all the factors identified by consumers placed the trait ‘Pompous’ as an important trait to be classified within the same pile as other similar traits such as ‘Vain’. However, not all consumers agreed that ‘Selfish’ should be classified within the same pile as ‘Pompous’ and
‘Vain’. As a result, a further data cleaning method was conducted by adopting a fixed sorting method (Giguere, 2006). Giguere’s (2006) fixed card sorting method is similar to the free associated task, except that a restricted number of groups are generated during the card sorting task.
The group labels identified by consumers were collated together to form an overall group name by summing up what consumers initially labeled each group. The researcher grouped each of the five factors as follows: Group One ‘Self-centered’; Group Two ‘Boring’; Group Three ‘Operating outside established/agreed code of conduct’; Group Four ‘Selfish’; and Group Five ‘Irrational’. Three independent expert judges reviewed the overall category names in light of each category label identified by the consumers. One of the expert judges was a PhD holder in personality research and the other two were carrying out their PhD in either English Language or Marketing. All three independent expert judges discussed the factors with each other and raised concerns on three of the factors (‘Self-Centered’, ‘Selfish’ and ‘Operating outside established/agreed codes of conduct’) as they are categorized as similar factors. All three expert judges agreed on the overall category factors as: Group One ‘Egotistical’; Group Two ‘Boring’; Group Three ‘Socially Irresponsible’; Group Four
‘Critical’; and Group Five ‘Lacking Logic’. These categories were cross validated against Aaker’s (1997) five Brand Personality factors to ensure the categories are not the mere opposite of Aaker’s brand personality framework. Table 4.7 details the antonyms of five brand personality factors obtained from Collins English Dictionary (2006).
Table 4.7: Direct Antonyms of Aaker’s (2007) five brand personality factors compared with the Negative Brand Personality factors obtained from the Sorting Task.
Aaker’s Five Factors of Brand
Sincerity Dishonesty Egotistical
Excitement Indifferent Boring
Competence Inept / inability Socially Irresponsible
Sophistication Simplicity Critical
Ruggedness Refined Lacking Logic
Once the five Negative Brand Personality factors had been established, six additional respondents were requested to group all 43 traits into the 5 established groups to assess consistency in traits within each group. Respondents were also given the opportunity to either create a new category or
~ 147 ~
to omit traits if they felt a trait did not fall into the category or could be seen in light of the branding.
All other instructions were the same as the free card sorting task detailed above.
Six participants voluntarily agreed to conduct the fixed card sorting task, of which 4 were female and 2 were male. The card sorting task data was analyzed by visually assessing the frequency of traits occurring in each factor14. Traits that achieved 80% or more in frequency by respondents were shortlisted to reflect the common traits amongst respondents. Overall, the results show some consistency with respondents’ classifications of traits within each of the five groups. Table 4.8 details the common traits consumers grouped in each category.
Table 4.8: Details the traits respondents assigned from the Fixed Sorting Task.
The results indicate that some consistency in Negative Brand Personality traits emerged. However, due to the small number of participants involved in this study, it is difficult to conclude that the traits identified would be the generally accepted traits identified by the wider consumer population.
However, the traits in Table 4.8 were compared to the results of the free card sorting method displayed in Table 4.6. The results show consistent traits within each of the five factors. Furthermore, the results in Table 4.8 show the traits that are reflected in each of the groups in Table 4.6. No single factor has been included in the fixed sorting task that has not already been anticipated from the free sorting task. For example, the high frequency traits from the free sorting task are ‘Pompous’,
‘Snobby’, ‘Brash’, ‘Vain’, ‘Arrogant’, ‘Stubborn’, ‘Pretentious’ and ‘Flaunt’. Comparing these results to the Group 1 traits in Table 4.8, it can be seen that all the traits were assigned within the Group 1 categories: ‘Pompous’, ‘Arrogant’ and ‘Flaunt’ were assigned in four of the six categories in Group 1;
‘Snobby’ and ‘Vain’ were assigned in all six categories in Group 1; ‘Brash’ was assigned in two of the six categories in Group 1; ‘Pretentious’ was assigned in three of the six categories in Group 1 and
14 Multidimensional scaling is one statistical technique that historically has been used to analyze card sort tasks.
However, the focus of this research is on identifying common Negative Brand Personality traits and potential dimensions; therefore, a visual frequency of traits occurrence technique was applied to analyze the data.
Name of Factor Traits
Egotistical Pompous Snobby Brash Vain Arrogant Pretentious Flaunt Stubborn
Boring Boring Monotonous Dull Lonely Anti-Social Cheap
Socially Irresponsible
Immoral Unethical Deceiving Deviant Fake Manipulative
Critical Confused Mischievous Rebellious Selfish Barbaric Judgmental
Lacking Logic Delusional Weird Unstable Naive Superficial
~ 148 ~
‘Stubborn’ was assigned in five out of the six categories in Group 1. The consistency in results suggests that there is some degree of reliability in the Negative Brand Personality factors.
Overall, the results of the free and fixed card sorting methods provided an indication of Negative Brand Personality factors and traits that are likely to result from the Principle Component Analysis. A total of 11 items were eliminated as respondents were unable to categorize the traits in either of the free or fixed factors and thus were concluded to be irrelevant traits to brand personality. The traits are ‘Fickle’, ‘Intimidating’, ‘Angry’, ‘Absurd’ and ‘Tyrant’. However, it is worth noting that six of the eleven traits were categorized by two of the six respondents in the fixed sorting task. Due to the low response rate, a further investigation was required of the following six traits: ‘Flimsy’, ‘Vanity’,
‘Aloof’, ‘Repulsive’, ‘Unstable’ and ‘Coarse’. It was therefore decided by the researcher to include the six items in the subsequent analysis. As a result, the subsequent study (Content Validity Assessment) was set up to help purify Negative Brand Personality traits pertaining to each factor.