This Article Abstract Full Text (PDF) Services Download to citation manager PubMed PubMed Citation

Sekentei and Family Caregiving of Elders Among the Japanese: Development and Psychometric Evaluation of the Sekentei Scale

¹ Department of Sociology and Human Welfare, Meisei University, Tokyo, Japan.
² Social Science Research Institute, University of Hawaii at Manoa.

Address correspondence to Masayuki O. Asai, Department of Sociology and Human Welfare, Meisei University, 2-1-1 Hodokubo, Hino-shi, Tokyo 191-8506 Japan. E-Mail: moa{at}soci.meisei-u.ac.jp

	Abstract

TOP Abstract Methods Results Discussion References

 
Objectives. This study describes the development and psychometric evaluation of a measure for assessing sekentei, a construct rooted in the Japanese culture. Sekentei is a social construct that causes a person to worry about others' evaluations of his or her behavior. The concept of sekentei may be an important factor related to family caregiving of elders in Japan.

Methods. We used a multiphased process to develop, refine, and evaluate the reliability and validity of the Sekentei Scale. Following the development and refinement of the measure, we administered the Sekentei Scale and other measures for assessing validity to 150 Japanese National and 116 Caucasian American family caregivers of elders. We used Cronbach's alpha to estimate internal consistency reliability. We examined validity by (a) correlating scores on the Sekentei and Shame Scales and (b) testing the Sekentei Scale's ability to yield hypothesized differences between the Japanese and the Americans.

Results. The results provided adequate support for the reliability and validity of the Sekentei Scale.

Discussion. The findings provide useful data regarding the construct and measurement of sekentei and provide preliminary psychometric support for the measure's use in future research.

SEKENTEI is a construct rooted in the culture of Japan (Inoue, 1977) that causes an individual to worry about other people's observations and evaluations of his or her behavior (Asai & Kameoka, 2005). Sekentei commonly refers to an individual's concerns about behaving in a socially acceptable manner as judged by others (e.g., a parent worrying about sekentei due to his or her son not achieving good grades in school). According to Miyake and Yamazaki (1995), sekentei is the "eyes" of other people who observe an individual's behavior. The Japanese have learned what is shameful by internalizing this concept. Although sekentei is a significant concept underlying norms that regulate social behaviors among Japanese people (Inoue, 1977), there are underlying aspects of sekentei such as social acceptability and shame that, in varying degrees, commonly shape social behavior across other cultures as well.

In relation to caregiving, sekentei refers to people watching how well families care for elder relatives, which can lead to feelings of shame when family caregivers utilize formal caregiving services (Momose & Asahara, 1996). In Japan, institutional care, as opposed to hospital care, is regarded by the vast majority as sinful and evidence of an unsuccessful family life (Yamamoto & Wallhagen, 1998). The Japanese believe that sending an elderly parent to a hospital for health care is an expected and acceptable behavior and, as such, it is considered "good for sekentei." However, many Japanese believe that sending an elderly parent to a nursing home is undutiful. Thus, the concept of sekentei may be an important factor influencing family caregiving of elders. Despite its significance for understanding family caregiving among the Japanese and possibly in other cultures, researchers have paid this concept little attention in the empirical literature. Given the need for empirical studies on sekentei, the purpose of this study was to develop a measure of sekentei and to evaluate its psychometric characteristics.

Although not focused on family caregivers per se, Momose and Asahara (1996) examined the relationship between sekentei and attitudes toward use of social services by administering a questionnaire composed of questions concerning sekentei, avoidance of social services, and actual use of social services among elders taking classes at a university in Japan. Twelve of the questionnaire items pertained to sekentei. Our study sought to improve upon Momose and Asahara's measure by systematically developing and evaluating the psychometric adequacy of a more comprehensive measure of sekentei.

	METHODS

TOP Abstract Methods Results Discussion References

 
Phase 1: Item Generation
The purpose of Phase 1 was to identify the content domain underlying the construct of sekentei and to generate items that represent this domain. To define the content domain underlying the construct, we searched the available Japanese literature on sekentei, shame, and family caregiving at various libraries and gerontological institutions in Japan. This literature review supplemented our review of the Western literature on shame and related constructs. Our literature review suggested that sekentei is best conceptualized in terms of avoidance of shameful behavior, concerns about how one appears to others, fear of social criticism, self-restraint in deference to others, and privacy of family (Asai & Kameoka, 2005). We generated a Japanese-language scale with 53 items representing this concept domain.

Phase 2: Item Selection and Refinement
The purpose of Phase 2 was to examine the clarity of the items and to refine the scale based on feedback from seven bilingual cultural experts. We asked these experts to comment on all aspects of the 53 sekentei items, including instructions and demographic questions. This procedure resulted in the modification of three items and the deletion of one item.

Phase 3: Item Translation
The purpose of Phase 3 was to develop an English-language version of the revised 52-item Sekentei Scale. Three bilingual Japanese speakers translated either a Japanese or English version of the scale. We used back translation and decentering procedures (Brislin, 1980) to achieve linguistic equivalence. The first bilingual speaker translated the original Japanese version into English. The second bilingual speaker independently translated the scale back into Japanese. The third bilingual speaker then independently translated the revised Japanese scale into English. We compared the two English translations, evaluated the quality of the translations, and selected or modified the translated items to achieve grammatical and colloquial appropriateness.

Phase 4: Pilot Study
The purpose of Phase 4 was to pilot test the translated English and Japanese versions of the 52-item Sekentei Scale and to obtain feedback regarding clarity of the measures. Participants were five Japanese National and three Caucasian American family caregivers living in Hawaii. Participants completed the Sekentei Scale and identified any aspect that was ambiguous or unclear. Participant feedback resulted in the deletion of nine ambiguous items and the modification of nine items in the Japanese version and five items in the English version. The result was a 43-item Sekentei Scale.

Phase 5: Psychometric Evaluation of the Sekentei Scale
The purpose of Phase 5 was to evaluate the reliability and validity of the 43-item Sekentei Scale. Participants were 150 Japanese National and 116 Caucasian American family caregivers of elders. We included a sample of Americans to provide comparison data for evaluating divergent validity (Nunnally & Bernstein, 1994) of the scale. We hypothesized that the Americans would be relatively less conscious of sekentei than the Japanese, because Western individualist cultures tend to emphasize a person's independence and autonomy in contrast to non-Western collectivist cultures, in which interdependence is valued and the group's interests tend to override the goals of the individual (Triandis, 1995).

Participants
We obtained our Japanese sample with the assistance of participating social service agencies in Tokyo and Ibaraki Prefecture located near Tokyo. Similarly, we obtained the American sample with the assistance of social service agencies in the San Francisco Bay area and in Honolulu. Participants were family caregivers of elders who were receiving social services from the participating agency.

Of the 150 Japanese participants, 39% resided in Tokyo and 61% resided in Ibaraki Prefecture. Of the 116 American participants, 91% resided in the San Francisco Bay area and 9% lived in Hawaii. The mean age of the Japanese participants was 58.63 years (SD = 12.03), and the mean age of the American participants was 62.19 years (SD = 12.95). Caregivers in both groups were predominantly female (78% of Japanese and 74% of Americans) and married (83% of Japanese and 69% of Americans) with an average of two children. The mean age of Japanese care recipients was 81.36 years (SD = 8.08), and the mean age of American care recipients was 80.69 years (SD = 8.44). The majority of care recipients were female (70% of Japanese and 69% of Americans) living with caregivers (80% of Japanese and 63% of Americans). The social services most frequently used by the Japanese as well as the American caregivers were adult day care centers (87.90% and 70.43%, respectively), followed by in-home services (47.58% and 41.74%, respectively).

The sociodemographic characteristics of the Japanese caregivers and care recipients appeared to be typical of those reported in prior studies in Japan (e.g., Li, 2004; Oshima, Kodama, Gotoh, Adachi, & Miyake, 2005; Suyama, Kawano, & Kawano, 2004). In these studies, as well as in a recent national survey of caregivers in Japan (Ministry of Health, Labour, and Welfare, 2006), the majority of caregivers were women in their late fifties or early sixties who provided care for their elder relative aged 80 to 90, characteristics comparable to those of our Japanese sample. Also, a national survey of caregivers in the United States (Wolff & Kasper, 2006) showed sociodemographic characteristics similar to those of our American sample. Nationally, average ages of U.S. caregivers and care recipients were 62.5 and 79.7 years, respectively (62.19 and 80.69 years in our study); 41.3% of those providing care were children (41.7% in our study); and 66.6% of care recipients were women (74% in our study). Thus, the available data suggest that participants in our study tended to resemble family caregivers studied in Japan and the United States.

Measures
In addition to our Sekentei Scale, we administered two measures for evaluating construct validity of the Sekentei Scale: (a) the Shame Scale and (b) Momose and Asahara's (1996) 12 sekentei items.

We developed the Shame Scale based on two shame-related scales: the revised Dimensions of Conscience Questionnaire (DCQ; Johnson et al., 1987) and the Other as Shamer Scale (OAS; Goss, Gilbert, & Allan, 1994). The DCQ contains 28 items that measure feelings of shame and guilt in situations likely to occur in Asian cultures. The measure has been shown to have high internal consistency and adequate construct validity in samples from Hawaii, Korea, and Taiwan (Johnson et al., 1987). The OAS is an 18-item scale that measures perceptions of how others judge one's self. Goss and colleagues (1994) reported that the OAS has satisfactory reliability and concurrent validity. We developed our Shame Scale by selecting five DCQ items and four OAS items based on their relevance to the concept of shame in the Japanese culture and by creating one new item. We designed our 10-item Shame Scale to measure degree of shame in specific situations. Each item was rated on a 4-point scale ranging from 1 ("I would not feel ashamed at all") to 4 ("I would feel very ashamed").

Along with the Sekentei and Shame Scales, we included Momose and Asahara's (1996) 12 items assessing sekentei. The authors reported that the internal consistency of the 12 items was.72. We modified the original 5-point rating scale to a 4-point scale so that we could incorporate the 12 items into our Sekentei Scale.

The Shame Scale was translated into Japanese, and Momose and Asahara's sekentei items were translated into English. We used back translation and decentering procedures described in "Phase 3: Item Translation" to achieve linguistic equivalence.

We asked participating social service agencies to distribute questionnaires containing the measures to participants. We provided participants with a stamped envelope for returning the questionnaire.

	RESULTS

TOP Abstract Methods Results Discussion References

 
Reliability
We estimated internal consistency reliability of the scales with Cronbach's alpha and by examining item–total correlations. As shown in Table 1, coefficient alphas for the original 43-item Sekentei Scale were.93 for the total sample,.90 for the Japanese sample, and.87 for the American sample. Item–total correlations for 10 items were below.30. Further item analyses showed higher reliability estimates when these 10 items were eliminated. Because these results indicated that the 10 items were not highly correlated with other items, we revised the original measure by eliminating these items. Also, we deleted an eleventh item because it was similar to another item in the scale. Thus, the final Sekentei Scale consisted of the 32 items shown in the Appendix. Despite the reduction in the number of items (32 vs 43), internal consistency estimates improved for the revised 32-item Sekentei Scale (see Table 1). Coefficient alphas for the 32-item scale were.95 for the total sample,.91 for the Japanese sample, and.89 for the American sample.

Validity
We examined the convergent validity of the Sekentei Scale by correlating the Sekentei and Shame Scales. Convergent validity would be demonstrated by significant correlations between the two measures because shame is an underlying feature of sekentei. As expected, higher scores on the Sekentei Scale were significantly associated with higher levels of shame, r(264) =.76, p <.01, for the total sample; r(148) =.43, p <.01, for the Japanese; and r(114) =.57, p <.01, for the Americans. Also, scores on the Sekentei Scale were significantly related to scores on Momose and Asahara's sekentei items, r(264) =.63, p <.01, for the total sample; r(148) =.73, p <.01, for the Japanese; and r(114) =.66, p <.01, for the Americans.

We evaluated divergent validity of the Sekentei Scale by testing the scale's ability to yield hypothesized differences between the Japanese and the Americans in sekentei scores. Table 2 presents means and standard deviations for the 32-item Sekentei Scale, Momose and Asahara's sekentei items, and the Shame Scale. Results of multivariate analysis of variance and subsequent univariate t tests supported the hypothesized differences between the two groups, Wilks' Lambda =.020, F(7, 258) = 1848.98; p <.001. Specifically, Japanese participants had significantly higher scores on the Sekentei Scale, the Shame Scale, and Momose and Asahara's sekentei items than did American participants.

	DISCUSSION

TOP Abstract Methods Results Discussion References

We evaluated validity of the Sekentei Scale by examining convergent and divergent validity. Convergent validity evidence showed that higher scores on the Sekentei Scale were significantly related to higher scores on both the Shame Scale and Momose and Asahara's (1996) sekentei items. These findings suggest that the concept of sekentei is associated with feelings of shame and perseverance to avoid that feeling. Divergent validity evidence showed that Japanese participants scored significantly higher on the Sekentei Scale than did the Americans. Inspection of each item score showed that Japanese participants scored markedly higher than the Americans on nine items. Of these nine items, three pertained to shame and embarrassment, two concerned the importance of avoiding being laughed at, and two referred to the importance of obtaining approval from others. The remaining two items related to the importance of not drawing attention from others and difficulty behaving independently. These concerns are fundamental aspects of sekentei, and the findings indicated that these aspects of sekentei differentiated our samples of Japanese and American family caregivers. These results, along with differences between the groups on the Shame Scale, are consistent with the literature that characterizes the Japanese as particularly worried about failing or making mistakes in front of others, which, in turn, lead to public embarrassment (Miyake & Yamazaki, 1995).

Although our findings suggest that the Sekentei Scale is psychometrically adequate, the results of this psychometric evaluation should be interpreted within the limitations of our study. First, our samples consisted of family caregivers who were using social services in several areas in the United States and Japan; issues regarding the representativeness of the samples in both countries may limit the generalizability of the findings. As described previously, however, the sociodemographic characteristics of our study's participants resembled those reported in national surveys of family caregivers of elders in Japan and the United States. Nonetheless, further studies are needed to evaluate the psychometric properties of the Sekentei Scale in other samples, including family caregivers who do not use social services. A second limitation of our study is the lack of evidence regarding the discriminant validity of the measure. Although our results provided support for the measure's convergent and divergent validity, further research on the scale's discriminant validity would enhance scholars' understanding of the nomological net (Cronbach & Meehl, 1955) that defines the construct. Despite these limitations, this study offers useful information regarding the measurement of sekentei and provides preliminary psychometric support for use of the Sekentei Scale in future research.

	Footnotes

 
Decision Editor: Kenneth F. Ferraro, PhD

Received for publication January 25, 2006. Accepted for publication February 6, 2007.

	References

TOP Abstract Methods Results Discussion References

 

• Asai, M. O., Kameoka, V. A. (2005). The influence of sekentei on family caregiving and underutilization of social services among Japanese caregivers. Social Work, 50,111-118.[Medline]
• Brislin, R. W. (1980). Translation and content analysis of oral and written material. In H. C. Triandis & J. W. Berry (Eds.), Handbook of cross-cultural psychology: Methodology (Vol. 2, pp. 389–444). Boston: Allyn & Bacon.
• Cronbach, L. J., Meehl, P. E. (1955). Construct validity of psychological tests. Psychological Bulletin, 52,281-302.[Medline]
• Goss, K., Gilbert, P., Allan, S. (1994). An exploration of shame measures: The Other As Shamer Scale. Personality and Individual Differences, 17,713-717.
• Inoue, T. (1977). Sekentei no kozo. [The structure of sekentei]. Tokyo: Nihon Hoso Shuppan Kyokai.
• Johnson, R. C., Danko, G. P., Huang, Y., Park, J. Y., Johnson, S. B., Nagoshi, C. T. (1987). Guilt, shame, and adjustment in three cultures. Personality and Individual Differences, 8,357-364.
• Li, W. (2004). Factors related to hope of continuation of caring for the frail elderly at home. Japanese Journal of Gerontology, 25,471-481.
• Ministry of Health, Labour, and Welfare. (2006). Abridged National Life Basic Survey, 2004. Retrieved November 20, 2006, from http://www.mhlw.go.jp/toukei/saikin/hw/k-tyosa/k-tyosa04/4-3.html.
• Miyake, K., Yamazaki, K. (1995). Self-conscious emotions, child rearing, and child psychopathology in Japanese culture. In J. P. Tangney & K. W. Fischer (Eds.), Self-conscious emotions: The psychology of shame, guilt, embarrassment, and pride (pp. 488–504). New York: Guilford Press.
• Momose, Y., Asahara, K. (1996). Relationship of "Sekentei" to utilization of health, social and nursing services by the elderly. Nihon Koshu Eisei Zasshi, 43,209-219.
• Nunnally, J. C., Bernstein, I. H. (1994). Psychometric theory. New York: McGraw-Hill.
• Oshima, C., Kodama, K., Gotoh, T., Adachi, K., Miyake, Y. (2005). A study of home environment modifications for elderly people with dementia. Journal of Japanese Society for Dementia Care, 4,475-487.
• Suyama, K., Kawano, R., Kawano, Y. (2004). Factor analysis in the formation of positive appraisal of family caregivers. Japanese Journal of Gerontology, 25,461-470.
• Triandis, H. C. (1995). Individualism and collectivism. Boulder, CO: Westview Press.
• Wolff, J. L., Kasper, J. D. (2006). Caregivers of frail elders: Updating a national profile. The Gerontologist, 46,344-356.[Abstract/Free Full Text]
• Yamamoto, N., Wallhagen, M. I. (1998). Service use by family caregivers in Japan. Social Science and Medicine, 47,677-691.

This Article Abstract Full Text (PDF) Services Download to citation manager PubMed PubMed Citation