A Comparison of Consumer Attitudes Within Hong Kong Society

A COMPARISON OF CONSUMER ATTITUDES WITHIN HONG KONG SOCIETY

Clint. R. Laurent, Hong Kong University

ABSTRACT -

Understanding the attitudes that underlie consumption behavior is important to product positioning and identifying market opportunities and threats. This paper examines the structure of attitudes that exist within Hong Kong's society and considers how these vary across the society.

INTRODUCTION

The manner in which individuals spend their time and money is generally described as their lifestyle (Engel and Blackwell 1982) and within any one society there is generally quite a variety in terms of the patterns of life-style that exist. For example, at one extreme there might be the home orientated person who lives for and with their family, and consume a relatively limited range of low profile items. At the other extreme there is the outward going individual engaging in conspicuous consumption and attempting to maximize their experiences in life.

However, what people consume and do is simply manifestations of more subtle aspects of the consumer - that is their underlying beliefs or attitudes. The generally held hypothesis is that a person's predisposition to act is a function of their beliefs or attitudes (Fishbein 1967) where a belief is an enduring structure of values held by the individual (Rokeach 1960). from a marketing point of view this raises three important questiuns:

1: To what extent do sub-groups of attitudes/belief structures exist in the society.

2: what type of attitudes or beliefs are most significant in influencing consumptionpatterns.

3 : what are the diversity of attitudes within the society and how might they change over the next few years/decade such as to create new market or product opportunities.

The first two of these questions has been dealt with in a separate paper (Laurent 1985) which discusses the general life style constructs that exist for the overall Hong Kong society and how these general constructs have changed over a two year period and their implications for the consumption of different products.

The objective of this paper is to make comparison between different life-style segments of the society on key attitudinal constructs.

METHODOLOGY

The methodology of this study involve the following steps:

i: determine the beliefs which should be measured in Hong Kong in order to ultimately identify the key attitudinal constructs underlying this study.

ii: develop a questionnaire which permitted reliable measurement of peoples' beliefs.

iii: draw a sample which was representative of the population as at August 1984 and large enough to permit adequate analysis of sub-group behavior.

iv: identify from the resulting measures the key attitudinal constructs underlying this society.

v: Group the population into segments according to their similarity of attitudes.

Developing The Attitude Measures.

A particular difficulty faced in this research study is identification of the precise scales or variables to be used for measuring attitudes in the Hong Kong context.

Being the first such study in Hong Kong, no prior measurement instrument existed and, to import a questionnaire from another society (most probably the United States or United Kingdom), would be quite wrong. It would not measure the attitudes of Hong Kong's population reliably. It would miss key dimensions unique to Hong Kong and over emphasize others. Accordingly, the first part of this research had to be the development of the instrument.

The methodology employed to develop the attitude scales was as follows:

a: development of an initial set of attitudinal constructs. These were based on observation of this society, conclusions of studies done in a variety of cultures and relevant literature on consumer attitudes associated with consumer behavior (for example, Leo Burnett 1972).

b: using these constructs as a basis for qualita tive research. This established their relevance and dimensions and also whether any constructs were missing and the dimensions of these.

Focus groups were used as it was considered that a group situation tends to be more dynamic in nature and would potentially engender a greater depth of discussion than would a one-to-one depth interview. in total 63 focus groups were conducted covering all segments of the population in Hong Kong.

To get people to discuss life styles openly projective techniques, where they discuss a third party rather than themselves, were used. 'This was necessary because people in Hong Kong have a particular unwillingness to openly discuss themselves as this may result in a loss of face (personal pride) which is still a dominant influence on behavior in Asia.

In each of these groups the discussion covered each of the hypothesized attitudinal constructs and the group members were asked to explain what they considered this 'third party's' opinions might be. The exact list of constructs varied as the research progressed, with those discovered to be redundant in Hong Kong being deleted, and other constructs added to fill identified gaps. This ensured that the society itself determined the manner in which it should be measured. in total the 63 groups generated 429 unique statements split across 21 hypothesized attitudinal constructs.

c: Selection of indicators for each construct.

These were drawn from the qualitative discussions and involved two stages. The first was determining the manner in which the statements would be expressed individually, and the second was selecting a smaller subset of statements from the overall list to be used in the final questionnaire.

As there were obvious advantages to be gained from using actual statements similar to those employed in overseas studies (validation and comparison, in particular), a bank of statements was made of questions used in a number of published overseas studies and this was then compared With the statements generated by the group discussions. In many cases the statements were identical and in most others there was a very similar statement such that the local one could be modified marginally without any loss of meaning. Accordingly, the actual statement set employed in this study has a degree of universality about it.

As mentioned, 429 unique statements were generated by the group discussions. However, prior experience has shown that the maximum number that a respondent could be expected to handle was about 200 and it was necessary to reduce the list accordingly'. To achieve this reduction the researcher used his judgement as to which individual statements should be retained under each of the hypothesized constructs in order to best measure the intended meaning and minimize duplication. As the statements were already grouped under the hypothetical constructs by the 63 focus groups, the task was more an issue of establishing redundancies within the meaning of the construct rather than changing the meaning of the construct itself

The Questionnaire

This consisted of 5 sections as follows:

Demographics - age, sex, housing type, occupation, marital status education, household and personal income.

Media - exposure to television, newspaper, magazines, radio, cinema and mass transport media.

Product usage- (usage, frequency, main brand) for coffee, beer, brandy, supermarkets, fast food, banking services, soft-drinks, tonic drinks, cigarettes, shampoo and skin care products.

Activities - frequency of doing each of 31 activities popular in Hong Kong.

Attitude statements - as discussed in the previous section. Each of the statements was measured on a 7 point agree-disagree scale.

Once the entire questionnaire had been constructed in Chinese,. it was tested for both length and communication and after some modification a final form was adopted.

The Sample

In the context of this research the sampling objective was to draw a random sample in August-September 1984, which was representative of the Hong Kong population within the age parameters of 16 and 64 inclusive.

In order to maintain the integrity of the achieved sample it was necessary to minimize the incidence of non-response. Accordingly, contact with each household selected from the sample frame had to be attempted three times. Only then was a non-contact recorded.

Because of the length of the questionnaire, interviewee co-operation was ensured by offering an incentive (to be delivered after the interview was completed) in return for his/her co-operation. Overall the refusal rate was not as high as expected. The norm in Hong Kong for household interviews is 30% refusal and this study experienced a 22% refusal rate.

The target sample size was set at 2,000 respondents and 2050 usable questionnaires were obtained. While the resulting sample profile was considered to be satisfactorily close to that of the population, it was decided that it would be desirable to improve this representation further by means of a minimal weighting scheme using age, sex and education.

Identifying the Key Attitudinal Constructs

Principal Component Analysis was used to identify if any underlying structure existed within the response pattern to the 197 attitudinal variables such that they could be summarized into this smaller structure while retaining the essence of the original variables.

To give some measure of the validity of the ultimate solution obtained, the principal component analysis was done on two separate halves of the sample (every alternate case), identifying in the most appropriate solution and then comparing the two solutions for consistency and robustness.

The Scree test, suggested by Cattell (1966), was used to select the number of factors and 16 was found to be the best number. Varimax rotation was then used to improve the interpretation of the factors.

Determining the Life-Style Segments

Having established the sixteen factors that summarize the key attitudinal constructs that exist within the society, and the components thereof, the next step was to cluster the respondents into segments, with each segment being maximally different from the others while at the same time maximizing internal consistency.

Again, the sample was split into two halves in order to provide an internal validity check on the cluster solutions. In addition, it was necessary to experiment with the inclusion or not of the demographic characteristics of the sample. This posed some initial problems in that cluster analysis requires that the variables being used be as independent of each other as possible. 'The attitudinal constructs (the factors) satisfy this criteria but there is a high degree of intercorrelation between the key demographic variables. In particular, in Hong Kong, age and education, education and income, income and housing type. To alleviate this problem a factor analysis was applied to the demographic variables (after being converted as beat possible to ordinal scales) and this identified that the use of age, sex and housing type would minimize inter-correlation within the demographic variable set and maximize explanation of a person's demographic profile. In addition, these three variables, while having some correlation with the factors, it was not excessive - with the possible exception of age.

The clustering process identified a seven solution as being the one which seemed most appropriate. That is, maximizing differences between clusters while minimizing difference within the groups as measured by the root mean square. 'This was also supported by the 'interpretability' of the resulting clusters.

ATTITUDINAL SEGMENTS

Figure 1 is a diagrammatic presentation of how the seven clusters relate to each other. The segments have been given the names shown to describe their overall character. The subsequent discussion will examine in greater depth how these segments differ from each other on some of the key attitudinal constructs and the implications of this for future consumption habits and sociological trends.

'The Immature Teenager

These persons are aged mainly between 15 and 24 (61%), and a significant proportion are students. There is a marginal bias to females. Their attitudes suggest a questioning outlook and they tend to aspire to foreign values. There is a degree of alienation (including from the family) and peer groups are a significant influence on behavior.

FIGURE 1

THESTRUCTURE OF THE SEVEN LIFE-STYLE SEGMENTS

Young Independents

This is also a relatively young segment (15 to 34 years) and is exclusively female. They are of lower socio-economic profile and a significant proportion (31%) are housewives and a further 17% are students.. They are very fashion conscious, and are very forward looking (modern) in their tastes. They lack confidence and are not particularly ambitious. There is some indication that they are not particularly flexible in their thinking.

Home Makers

This is another totally female group, but the average age is higher than that of the preceding group. This segment is above average socioeconomically and 44% are housewives. They are fashionable, and quite cosmopolitan in their outlook and the most self confident of the female segments. However, they are very family orientated.

Assertive Youth

This segment is exclusively male aged between 20 and 35, well educated for the population (about average for their age group) and is essentially middle class. They are quite self centered, have a high opinion of themselves and are quite aggressive and ambitious. Peer groups are more important than the family in terms of orientation and values. They have adopted the traditional values in terms of desire to succeed in business, make money and been seen to be successful. As such their overall behavior and dress is quite conservative to fit those norms of a conventional successful businessman. They are, however, quite permissive.

Self Indulgents

This does not have a particularly distinct demographic profile, and has only a slight male bias. As a segment they are generally unconcerned about their environment and have little family involvement- Collectively they appear as a person who simply takes matter as they come and would never really Lake a stand on anything or strive for an objective.

Male Conservatives

This is an older male segment and because of the age bias they tend to be less well educated than the average for the population. They are also biased to blue collar occupations. They have a high level of anxiety and an above average desire to-succeed, but they do have a degree of reformist attitudes and are not as traditional as one might expect. 'They are relatively content with their political/living environment.

Traditional Matriarchs

This is the older female segment with a low level of educational attainment. They appear to have withdrawn from society and their primary interest is their family. They are not social, have a low community orientation and worry a lot about health and money.

ATTITUDE DIFFERENCES BETWEEN SEGMENTS

In this final section the differences between segments on some of the key attitudinal constructs will be examined. The constructs selected here are, family, money, innovativeness, permissiveness and ambition.

Family

Overall the population is biased towards having a family orientation. On a seven point scale from 1 being definitely disagree to 7 being definitely agree, the average rating on the family orientation statements is 4.84. For any one statement, never more than 31% disagree with the statement.

Looking across the segments, it is no surprise to find-that the segments which either have a strong female bias, or are older have a much higher family orientation than the other segments. However, even the self indulgent segment (which is middle aged and equally male and female) are positive in terms of family orientation.

Money

The statements within this attitudinal construct can be divided into two subgroups. Those that relate to the desire to acquire money, and those related to the use of cash and credit.

Dealing first with the extent to which the society is money driven it is surprising that on average the population is not attempting to maximize its monetary wealth. A typical statement in this category is "in a job, security is more important than money'' to which 77% agreed. In addition 69% felt that there should be less emphasis on being rich and making money. However, it is worth noting one anomaly and that is that when given the choice of either longer work hours and more money, or less hours and less money, the majority (69%) opted for the longer hours. This may well be more a reflection of the work ethic in the Hong Kong society than a desire for money.

Between segments, the two which are most concerned with acquiring wealth were those of 'Assertive Youth' and ' Self Indulgent'. The 'Home Maker' and 'Male Conservative' were the ones least concerned with having money. , This reflects their more settled nature and subsequent desire for security.

The other aspect of money is the willingness to use credit. Traditionally credit was not encouraged in the Chinese society of Hong Kong. Attitudes are changing in this respect and currently the population is almost neutral in terms of the use of credit. The Immature teenager is particularly willing to use credit, with the next closest being the assertive youth segment which is essentially aggressive young males wanting to make an impression on others and hence the credit facility increases their apparent wealth. However, for day to day transactions the majority of the population prefers to use cash (81%). The one segment out of line with this is the self indulgent where 45% prefer to pay with cash/carry enough cash for their daily requirements. This same segment also differs significantly from the others in terms of agreement with the state "here today, gone tomorrow is my philosophy with money'. 29% disagreed, compared with an average of 56% for all the other segments.

Innovativeness

Overall Hong Kong citizens are biased towards innovativeness, a generally willingness to try new things. 'The most innovative segment is the Assertive Youth segment (young well educated males) followed by the Immature Teenager Segment. However, it is worth noting that both the younger female segments are on the mid-point of the innovativeness scale.

Moving from the aggregate index of innovativeness to some of the individual scales there are some variations which are of interest. In particular, while the Assertive Youth is the most innovative, change seeking segment it none-the-less the segment least likely to use less well known brands. (that is, they prefer to use the brand leader). This relates to another aspect of this society which is 'face'. This is a concern about loss of pride. This can happen as a result of such things as being proven wrong in a public situation, lacking prosperity, or using a low/poor quality brand. The assertive Youth is very concerned about maintaining face (this relates to their status/success consciousness) and hence are reluctant to use brands which may subject them to ridicule/criticism.

TABLE 1

AVERAGE SCORE OF EACH SEGMENT ON THE STATEMENTS INCLUDED IN EACH ATTITUDINAL CONSTRUCT

Permissiveness

The overall population is negative towards sexual permissiveness. The least permissive segments are the Traditional Matriarch (Older Female segment) and the Immature Teenager. Conversely the most permissive segments are the Self Indulgent and Assertive Youths, but even they are marginally negative towards permissiveness.

Looking at some of the individual statements within this construct there are some interesting variations across the segments. The Young Independent segment is the one most tolerant of a girl getting pregnant before marriage, .whereas the traditional Matriarch is the least tolerant.

In respect of liberalizing social laws (e.g film censorship and abortion) the immature teenager is the least agreeable to this. 'The Assertive Youth is the segment most in favor of such liberalization. The exception to this pattern is in respect of liberalizing homosexual laws. In Hong Kong homosexual activities are a criminal offense and all but the Self Indulgents are strongly opposed to any relaxation of the law. The Self Indulgents are indifferent about this issue.

Ambition

Of the seven life-style segments only one admits to being ambitious and that is the Assertive Youth segment. The next highest is the Self Indulgent segment which is indifferent to success.

CONCLUSIONS

In this paper it has been possible to only give a brief look at the structure and diversity of attitudes within the Hong Kong population. Being the first life style study done in Asia outside Japan it is useful to note that many of the attitudinal constructs and variations are not inconsistent with those found in European Societies, particularly the United States of America and the United Kingdom. Such a conclusion would suggest that provided the more subtle nuances of the culture are observed there is a basis for global marketing strategies.

REFERENCES

Cattell, R.B., (1966) "The Scree Test for the Number of Factors", Multivariate Behavioural Research, April.

Engel , James F. & Roger D. Blackwell (1982) Consumer Behaviour, 4th ed. Japan, Holt Saunders.

Fishbein, Martin. (1975) ''Attitude, Attitude Change, and Behavior: A Theoretical Overview'', in Philip Levine (ed.), Attitude Research Bridges the Atlantic. Chicago, American Marketing Association.

Laurent, Clint. R. "The Changing Nature of Consumer Attitudes",   Proceedings Management in Asia, Hong Kong University, 1985.

Leo Burnett PTE (1972),UK Life Styles, London

Rokeach, Milton. (1960) The Open and Closed Mind: Investigation into the Nature of Belief Systems and Personality Systems. New York. Basic Books.