The Chinese Personality Assessment Inventory (CPAI)
2.7.1 The Chinese Personality Assessment Inventory
With special attention given to the interpersonal relations domain, Cheung et al. (1996) developed the Chinese Personality Assessment Inventory (CPAI). To construct the CPAI the research team returned to first principles, with the development of a methodology that mirrored, to some degree, that adopted by Cattell (1943) in his ground-breaking research into the sphere of human personality. In particular they initially identified personality constructs by way of:
• A review of contemporary Chinese novels (n = “about 15”).
• A review of Chinese proverbs.
• An informal street survey conducted by students which revealed “about 300” statements of self-descriptions gathered from “about 50” people from a variety of backgrounds in Hong Kong.
• A survey (n = 433) of adjectives used to describe friends, colleagues and co-students with data gathered from professionals in Hong Kong (n = 215) and the PRC (n = 218).
• A literature review of specific constructs that have been investigated among Chinese samples.
Adapted from Cheung et al. (1996, p.184)
Subsequently, 150 distinct personality characteristics were obtained. These were further reduced using a committee/subject matter expert approach, wherein selection of constructs that were believed by the committee to be important aspects of personality and psychopathology among the Chinese people reduced the constructs to 26 normal personality and 12 clinical scales. From these 38 scales, the authors identified 10 personality characteristics which they labelled as indigenous scales. They claimed that such indigenous characteristics are not measured at either the broad- or narrow-band level in Western personality measures . Nine of the ten indigenous scales were considered normal personality traits and one, Somatisation, was seen to be a clinical scale. The identified indigenous scales were:
• Harmony: Measures one’s inner peace of mind, contentment and interpersonal harmony. The avoidance of conflict and maintenance of the equilibrium are considered virtues in the Chinese culture.
• Ren Qing (relationship orientation): Measures the individual’s adherence to cultural norms of social interaction such as courteous rituals, exchange of resources, maintaining and utilising useful ties, and nepotism.
• Traditionalism-Modernity: Measures the degree of individual modernisation as an indication of one’s responses to societal modernisation. This scale covers attitudes toward traditional Chinese beliefs and values in the areas of family relationships, filial piety, rituals, and chastity.
• Thrift-Extravagance: Measures the tendency to save rather than waste, and carefulness in spending. Thrift is one of the basic traditional Confucian Chinese values, and the characteristic of thrift versus extravagance is an indicator of the social response to rapid economic development and increasing materialism.
• Defensiveness/Ah-Q Mentality: Measures the defence mechanisms, such as self-protective rationalisation, externalisation of blame, self-enhancement, and belittling others’ achievements. A mild degree of defensiveness is accepted as a protective mechanism against defeat and disappointment.
• Graciousness-Meanness: Measures how broad-minded and kind people are in their dealings with others.
• Veraciousness-Slickness: Measures the trustworthiness and reliability of an individual.
• Face: Measures the concern for maintaining face and social behaviours that enhance one’s own face and that avoid losing one’s face. Face is a dominant concept in interpreting and regulating social behaviour in the Chinese culture.
• Family Orientation: Measures the extent to which individuals have a strong sense of family solidarity. These family ties provide emotional and economic security and support.
• Somatisation (a clinical scale): The expression of personal and social distress by way of physical symptomatology.
Adapted from Cheung et al. (1996) and Cheung, Leung, Song and Zhang (2001)
Trial items were then developed, and subjected to statistical analyses on a number of Chinese samples. Internal consistency reliabilities ranged considerably (indigenous and non-indigenous scales: a=.91 to a=.55 with a median of a=.74 for the Hong Kong sample and a=.80 to a=.58 with a median of a=.70 for the PRC sample: Cheung et al., 1996, p.189). Separate reliabilities for each scale of the CPAI were not initially reported although there has been some recent psychometric data published (scale alphas ranging from a=.66 to a=.77; no median provided) in Kwong and Cheung (2004). More recently, and in a revised CPAI-2, Cheung, Cheung, Zhang et al. (2004) report scale reliabilities ranging from a=.49 to a=.80 with a median alpha of a=.62. The lower alphas witnessed here may be due to the shorter scale length found within the CPAI-2. Alternatively, they may occur as a result of non-homogenous CPAI scales. In the latter case, test-retest data would be more suitable for establishing reliability. There are however no published test-retest statistics for the CPAI-2. Kwong and Cheung (2003) do report such statistics for the original CPAI, noting that correlations range from a=.54 to a=.94 with a mean of a=.70.
Cheung et al. (1996) argued that the CPAI was “…a pioneering attempt to develop an omnibus personality inventory in a non-Western culture…(and) has expanded the domain of cross-cultural personality assessment beyond that of translation and adaptation…(which)… with the advancement of psychology to non-Western cultures will become the trend in cross-cultural personality assessment.” (p.197). They also claimed that the ‘indigenous’ facets measured by the CPAI might not be unique to China, but might also exist in non-Chinese samples and that: “…the absence of Interpersonal Relatedness in a Western instrument may point to a blind spot in Western personality theories…” (Cheung, Cheung, Leung, Ward & Leong, 2003; p.450; and also supported in Cheung et al., 1996 and Lin, 2003).