- Email: [email protected]
A preliminary reliability and validity study of the Chinese version of the self-efficacy for exercise scale for older adults
Available online at www.sciencedirect.com
International Journal of Nursing Studies 46 (2009) 230–238 www.elsevier.com/ijns
A preliminary reliability and validity study of the Chinese version of the self-efficacy for exercise scale for older adults Ling-Ling Lee a,*, Shoa-Jen Perng a, Chin-Chih Ho b, Hsiang-Ming Hsu c, Shu-Chuan Lau d, Antony Arthur e a
Department of Nursing, Tzu Chi College of Technology, and School of Nursing, Tzu Chi University, Hualien, Taiwan b Department of Health Administration, Tzu-Chi College of Technology, Hualien, Taiwan c Department of Public Health, Tzu Chi University, Hualien, Taiwan d Medical School, Fu Jen Catholic University, Taipei, Taiwan e School of Nursing, University of Nottingham, Nottingham, UK Received 28 February 2008; received in revised form 16 August 2008; accepted 4 September 2008
Abstract Background: Population ageing affects most undeveloped, developing and developed countries. Less than a quarter of older people undertake a level of physical activity worldwide that is sufficient to lead to health benefits. Understanding older people’s confidence regarding engaging in exercise helps to structure physical activity interventions that motivate them to initiate and adhere to regular exercise. Estimates of the reliability and validity of the English version of the self-efficacy for exercise (SEE) scale has been widely tested and shown to be valid for use in various settings and among older people. The reliability and validity of a Chinese version of the SEE for older adults has not been tested. Aim: To undertake a preliminary assessment of the reliability and validity of the SEE scale when applied to the older Chinese adults. Methods: The Chinese version of the self-efficacy for exercise (SEE-C) scale was tested on a sample of 192 older people from Taiwan with a mean age of 71.2 years recruited between October and December 2003. Results: There was acceptable internal consistency of the SEE-C scale. The fit of the measurement model to the data for the SEE-C scale was acceptable. There was evidence of validity of the measure based on hypothesis testing: health status predicted exercise self-efficacy, and exercise self-efficacy predicted physical activity. Those who exercised regularly (n = 102) had a higher mean SEE-C score (5.3) than those who did not (2.9, n = 90). Conclusion: This preliminary validation study provided evidence for the reliability and validity of the Chinese version of the SEE scale. Future testing of the SEE-C scale needs to be carried out to see whether these results are generalisable to older Chinese people living in urban areas and with different characteristics. A test of the scale among younger adults for a wider use of the instrument is also warranted. # 2008 Elsevier Ltd. All rights reserved. Keywords: Chinese self-efficacy for exercise scale (SEE-C); Older people; Confidence; Validity; Reliability
What is already known about the topic? * Corresponding author at: Tzu Chi College of Technology, Department of Nursing, Room 429, No. 880, Chien-Kuo Road, Section 2, Hualien 970, Taiwan. E-mail address: [email protected] (L.-L. Lee).
Self-efficacy plays a role in the maintenance of physical activity beyond programme involvement, as well as in the early stages of exercise adoption.
0020-7489/$ – see front matter # 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijnurstu.2008.09.003
L.-L. Lee et al. / International Journal of Nursing Studies 46 (2009) 230–238
Understanding older people’s confidence regarding engaging in exercise will help to structure physical activity interventions that motivate them to initiate and adhere to regular exercise programmes. English version of the self-efficacy for exercise (SEE) scale has been shown to be valid for use in various settings and particularly among older people. What this paper adds The Chinese version of the SEE scale is likely to be internally consistent across its nine items although individual items explain relatively little of the overall variance. This study provided evidence for good criterion-related validity of the SEE-C but not for its construct validity. 1. Background Population ageing affects most undeveloped, developing and developed countries (The Department of Statistics Ministry of the Interior ROC, 2007). Despite the well-established health benefits of physical activity, surveys suggest that over 60% of older people do not participate in regular exercise, and only about 25% of this age group report regular participation in physical activity (Hawkins et al., 2004; Health Education Authority UK, 1999). In Taiwan, up to 27% of men and 39% of women aged over 65 years do not participate in exercise activity (Health Bureau, 2002). Understanding older people’s confidence regarding engaging in exercise will help to structure physical activity interventions that motivate them to initiate and adhere to regular exercise programmes. Self-efficacy theory is one of the most widely applied theories in predicting health behaviour. Self-efficacy, a central concept of Bandura’s social learning theory (Bandura, 1977), attempts to predict and explain human behaviour by referring to an individual’s assessment of their competence or effectiveness to perform a specific behaviour successfully. The theory of self-efficacy suggests that the stronger the individual’s self-efficacy and outcome expectations, the more likely it is that he or she will initiate and persist with a given activity. The theory of self-efficacy incorporates self-efficacy expectations, which are an individual’s beliefs in their capabilities to perform a course of action to attain a desired outcome; and outcome expectations, which are the beliefs that a certain consequence will be produced by personal action (Bandura, 1986a; Gecas, 1989; Schunk and Carbonari, 1984). There is an evidence from empirical studies of a relationship between self-efficacy and a variety of health-related behaviour, such as healthy eating (Long and Stevens, 2004) and weight loss (Roach et al., 2003). Regarding exercise behaviour, self-efficacy plays a role in the maintenance of physical activity beyond programme involvement, as well as in the early stages of exercise adoption (Aittasalo et al.,
231
2007; Garcia and King, 1991; McAuley, 1992; Sallis et al., 1990). The relationship between self-efficacy expectations and level of exercise is shown in various studies, in spite of the diverse samples, measurement instruments, and measurement timing (Clark, 1996; Gortner and Jenkins, 1990; Stutts, 2002). People with higher self-efficacy expectations were found to maintain a sense of energy during exercise, perceive less effort being expended during exercise, and report both more positive affect during and after exercise and feeling more revitalised (Bozoian et al., 1994; McAuley and Courneya, 1992). To improve exercise activity in older adults, it is suggested that self-efficacy expectations related to exercise are considered, because these beliefs influence motivation to exercise and actual exercise activity (Resnick and Jenkins, 2000). Several generic self-efficacy measurements have been developed (Judge et al., 2002; Muris, 2001; Tipton, 1984) as self-efficacy does enjoy certain generic features concerning the judgments about one’s self-efficacy (Bandura, 1997). However, the usefulness of the generic self-efficacy measurements has been questioned (Bandura, 1986b; Eysenck, 1978). An individual’s confidence in carrying out different behaviours may vary across behaviours and contexts. Bandura (1997) argues that there is no universal measure of perceived self-efficacy and that any scale should be tailored to the particular domain of interest. Without reference to the specific behaviour and situation, it is inappropriate to label an individual as having ‘high’ or ‘low’ self-efficacy. Therefore, the task of the researcher is to measure self-efficacy within a specific focus, such as self-efficacy in drug abstention, quitting smoking, or selfefficacy in doing exercise. An English version of the SEE scale was designed to test people’s confidence to continue exercising in the face of barriers to exercise (Resnick and Jenkins, 2000). Estimates of the reliability and validity of the nine item SEE scale have been widely tested and shown to be valid for use in various settings, particularly among older people with good internal consistency and validity. For example, there is evidence of internal consistency (alpha = 0.93) and validity with efficacy expectations significantly related to exercise activity, and factor loadings all greater than 0.50 (Resnick, 1999,2002; Resnick and Jenkins, 2000; Resnick et al., 2000). The theoretical basis, clinical feasibility, and some acceptable psychometric test features of the SEE scale have made it a valuable tool for the evaluation of SEE among North American populations. However, whether or not these estimates hold true for older people from different social and ethnic backgrounds needs to be examined. Little data exist on the use of the SEE scale among older Chinese people living in the community and how cultural differences might affect the administration of the SEE scale has not been investigated. Before the SEE scale can be applied more broadly, a cross-cultural evaluation is warranted. The purpose of the present study is to evaluate if there is evidence for the reliability and validity of
232
L.-L. Lee et al. / International Journal of Nursing Studies 46 (2009) 230–238
the SEE scale when used with older Chinese adults in Taiwan.
2. Methods The participants were obtained from the baseline measures of a randomised controlled trial carried out in Taiwan (n = 202) (Lee et al., 2007a,b) and recruited between October and December 2003. The trial was designed to examine a 6-month community-based walking intervention using selfefficacy theory to enhance confidence in regular walking as exercise and to examine the effect on blood pressure control. Data were collected concerning the participants’ age (over 60 years), systolic blood pressure (between 140 mm and 179 mmHg), and able to carry out regular walking as exercise. The original English version of the SEE scale was translated into Chinese. As with the original English version, the translated instrument contained nine items measuring the participants’ perceived ability to exercise under various circumstances (such as bad weather and tiredness). Two professionals in the fields of gerontology and public health in Taiwan were consulted for the wording and phrasing of each item of the initial Chinese version of the scale. Two items of the translated SEE scale were raised as being of concern with regard to the Chinese words employed to describe ‘‘enjoy’’ (item 5 ‘‘You did not enjoy it’’) and ‘‘depressed’’ (item 9 ‘‘You felt depressed’’). A modified Chinese version was developed after a series of meetings with members of the study team. This procedure was intended to ensure the Chinese version of the SEE (SEE-C) scale to be linguistically appropriate (Bracken and Barona, 1991; Bradley, 1994; Chang et al., 1999). Twenty-two older people living in another community were recruited to pilot test the translated measure for readability and feasibility of completion. The data of the SEE-C scale were collected by a set of questionnaires and administered by a nurse researcher to a group of older people living in the community via face-toface interviews. The participants were instructed to listen to the statement and then choose an option from 0 (not confident) to 10 (very confident) that represented their perception of confidence regarding engaging in regular exercise for 20 min three times per week. In prior research, older adults were able to respond to the 0–10 scale questionnaire (Resnick, 1998a; Resnick and Jenkins, 2000). Therefore, the original format of the 10-point scale was retained in the Chinese version. In addition, an answer card was designed with larger words stating all possible choice of responses and used to facilitate participants to rate their perceptions of confidence about doing exercise. The scale was scored by summing the numerical ratings for each response and dividing the total by the number of non-missing responses. The mean scores for the self-efficacy of exercise ranged from 0 to 10, with the higher scores representing greater exercise selfefficacy.
Several other measures were used to test the validity of the SEE-C scale. These measures included demographic information, physical activity status, and perceived health. The demographic information included age, gender, and years of education. Physical activity was measured by asking how often the participant exercised (e.g. playing table tennis, cycling, hiking, and swimming) and walked for at least 20 min in one go. The participants rated their physical activity as one of: never; less than once a week; once a week; two to three times a week; or more than three times a week. Regular exercisers were defined as adopting any form of physical activity three times a week and at least 20 min in one go and maintaining the physical activity for 3 months. Perceived health was measured by three items used in others’ work (Mason-Hawkes and Holm, 1993) that are summed to create a score of between 3 and 10 with higher scores indicating better perceived health. The participants rated their health as ‘very good’, ‘good’, ‘poor’, or ‘very poor’. They then compared their health to others similar to in age to themselves and rated it as ‘better than average’, ‘average’, or ‘worse than average’ (Mason-Hawkes and Holm, 1993). Finally, participants rated the extent to which their health interfered with their desired activities as ‘not at all’, ‘somewhat’, or ‘a great deal’. The criterion validity of this measure of perceived health has been supported by findings that suggest positive correlations with physician visits (r = 0.25) (Mason-Hawkes and Holm, 1993), number of reported symptoms (r = 0.38) (Mason-Hawkes and Holm, 1993), number of chronic illnesses (r = 0.46) (Conn, 1998) and number of prescriptive medications (r = 0.40) (Conn, 1998). The internal consistency of the perceived health scale, as measured by Cronbach’s alpha was 0.67 from the pilot testing (n = 22), consistent with an alpha of 0.76 reported in a previous study (alpha of 0.76) (Conn, 1998). 2.1. Statistical analysis Descriptive analyses of the data were performed with all study variables to describe the study participants. Of the 202 participants recruited, 10 were not used because of missing responses to at least three questions in the SEE-C scale resulting in a sample of 192. 2.1.1. Reliability testing Estimates of internal consistency were carried out using Cronbach’s alpha. In the present study, an alpha coefficient of 0.7 or greater was considered as evidence of the internal consistency of the SEE-C scale (Nunnally and Bernstein, 1994). This indicates that 70% of the variance of the observed score is systematic, and 30% is due to random errors. However, internal consistency cannot be used in a single item measure. An alternative estimate of reliability using a structural equation modelling, a squared multiple correlation coefficient (R2) was used (Bollen, 1989), as R2 estimates the systematic variance in the observed score that can be explained by each item in the measurement model
L.-L. Lee et al. / International Journal of Nursing Studies 46 (2009) 230–238
(Bollen, 1989; Jagodzinski and Kuhnel, 1987). For example, if item 1 R2 = 0.68, 68% of the variance in item 1 is accounted for by the variance in exercise self-efficacy. The remaining 32% of the variance cannot be accounted for and is attributed to error. Ideally, the R2 for each item, or the amount of variance accounted for, should be at least 0.5 (Bagozzi and Yi, 1988). 2.1.2. Validity testing Validity testing for the SEE-C scale was based on a confirmatory factor analysis using structural equation modelling. The chi-square statistic, the Normed Fit Index (NFI), and Steigers’ root mean square error of approximation (RMSEA) were used to estimate model fit. The larger the probability associated with the chi-square, the better the fit of the model to the data (Loehlin, 2004). Because the chisquare statistic is influenced by sample size, the chi-square divided by degrees of freedom was used to evaluate the model fit. Ideally, this ratio should be below 3 (Arbuckle, 1997). The NFI tests the hypothesised model against a reasonable baseline model and should yield a value of 1.0. The RMSEA, unlike the chi-square statistic, gives an estimate of model fit that is not influenced by sample size. An RMSEA of <0.10 is considered good, and <0.05 is very good (Loehlin, 2004). Validity testing for the SEE-C scale was also evaluated by examining its construct validity and concurrent validity. Construct validity of the SEE-C scale was tested using an evidence supported hypothesis that higher self-efficacy expectations are more likely to be observed among individuals with better health status (Grembowski et al., 1993; Rodin, 1989) and operationalised by observing whether there was a positive correlation between perceived health and SEE-C score. Concurrent validity was used to examine criterion-related validity. As criterion validity refers to the ability of an instrument to correlate well with a criterion that is related to the construct measured by the instrument (Cronbach, 1990; Nunnally and Bernstein, 1994), it is assumed that self-efficacy expectations had a directly significant effect on exercise behaviour (Ronen et al., 2003) in the present study. This had been operationalised by observing if there would be a positive
233
correlation between SEE-C score and level of physical activity frequency. The correlations were assessed using Pearson’s correlation coefficient (r). Multiple regression was used for the hypothesis testing. Four variables (age, gender, education year, and perceived health) were included as predictors of self-efficacy expectations for exercise. Multiple regression was also used for testing criterion-related validity with age, gender and education year adjusted for. Additionally, the score for self-efficacy expectations were included as a predictor of physical activity. The original trial was reviewed by the University of Nottingham Medical School Ethics Committee and ethical approval was granted in July 2003 (Study reference number: D/6/2003). Signed informed consent was obtained prior to the data collection.
3. Results There were more male subjects than females in the study (57.8% vs. 42.2%), with a mean age of 71.2 and mean education year of 4.2 (n = 192). The mean scores of the study participants on the SEE-C scale, their perceived health, and levels of physical activity frequencies are presented in Table 1. The mean SEE-C score was 4.2 on a scale of 0– 10 and the mean perceived health score was 5.7 on a scale of 3–10. Slightly more than half (102/192, 53.1%) of the participants took part in regular physical activity. 3.1. Reliability and validity The Cronbach’s alpha coefficient was 0.75 (n = 192), which was lower than the original English version, which was 0.92 (n = 187) (Resnick and Jenkins, 2000) but acceptable for the internal consistency of the SEE-C measure. The squared multiple correlation coefficients (R2) were all less than 0.50 and ranged from 0.15 (item 6) to 0.36 (item 4). There was support for the confirmatory factor analysis of the Chinese version of the SEE scale. Both the chi-square (x2 = 45, d.f. = 27, p = 0.016) and x2 divided by degrees of
Table 1 Demographic data and mean scores for the study measures (n = 192).
Age (years) Educational (years) Resting systolic blood pressure (mmHg) Resting diastolic blood pressure (mmHg) Chinese version of self-efficacy for exercise (SEE-C) scale Perceived health Men/women Participants doing regular physical activity S.D. = standard deviation.
Mean (S.D.)
Range
71.2 4.2 151.9 82.3 4.2 5.7
60–91 0–15 140–180 60–112 0–10 3–10
(6.0) (4.1) (10.6) (9.8) (2.6) (1.7)
n (%)
111(57.8)/81(42.2) 102 (53.1)
234
L.-L. Lee et al. / International Journal of Nursing Studies 46 (2009) 230–238
Table 2 Bivariate Correlations of the major study variables (n = 192). SEE-C SEE-C Physical activity Perceived health Outcome expectations for exercise
1.00 0.46** 0.17* 0.36**
Physical activity 1.00 0.19* 0.23**
Perceived health
Outcome expectations for exercise
1.00 0.13
1.00
Note: SEE-C = Chinese version of self-efficacy for exercise scale. *p < 0.05, **p < 0.01 (Pearson’s correlation coefficient).
freedom (x2/d.f. ratio = 1.67) suggested a fair model fit (Arbuckle, 1997). Moreover, the NFI was 0.90, and RMSEA was less than 0.10 (0.059). In total, the nine items explained 45% of the variance in exercise self-efficacy. The evidence for the criterion-related validity of the SEE-C scale was supported by the positive correlation between SEE-C score and level of physical activity (r = 0.46, p < 0.0001) (Table 2). After adjusting for age and gender, the SEE-C scores predicted exercise activity (F = 16.83, p < 0.0001), accounting for 44.6% of the variance in exercise activity. Evidence of criterion-related validity of the SEE-C scale was further supported by a statistically significant difference between SEE-C scores of those exercising regularly (at least 20 min three times per week) and those who did not (t = 6.88, p < 0.0001). Those who exercised regularly had a mean SEE-C score of 5.3 (S.D. = 2.4), whereas those who did not had a mean score of 2.9 (S.D. = 2.3). The evidence for the construct validity of SEE-C scale is weak. Although the Pearson’s correlation revealed a statistically significant correlation between perceived health and SEE-C score (r = 0.17, p = 0.019) (Table 2), the association between perceived heath and exercise self-efficacy was weak. When controlled for age, gender and education year, the scores of perceived health subscale significantly predicted the SEE-C scores (F = 3.43, p = 0.04), but accounted for only 15.1% of the variance in exercise self-efficacy. However, there was evidence for the construct validity of the SEE-C scale based on a statistically significant correlation between exercise self-efficacy and exercise outcome expectations (r = 0.36, p < 0.0001) (Table 2).
4. Discussion The present study provides preliminary evidence for the reliability and validity of the SEE-C scale. The participants in this study were community-based, from a rural area of Taiwan, with at least one chronic health problem (mild to moderate hypertension). The diversity of the participants contributes to the generalisability of the use of the scale and the findings can be used to strengthen the measure of exercise self-efficacy among older Chinese people for future use. There was acceptable evidence of internal consistency, tested by Cronbach’s alpha. However, there was limited
evidence of the reliability, based on squared multiple correlation coefficients (R2). Only 36% of the variance in item 4 was explained by the model. There was evidence for the construct validity of the measure based on statistically significant relationship between perceived health and exercise self-efficacy and between exercise outcome expectations and exercise selfefficacy. The evidence of the criterion-related validity of the measure was supported by a statistically significant correlation between exercise self-efficacy and physical activity behaviour. Although these relationships were statistically significant, it is recognised that these correlations were weak. This may be related to the fact that many factors influence exercise and other forms of physical activity behaviours, particularly among older people. Confirmatory factor analysis indicated that the fit of the full model to the data was acceptable but relatively lower than that of the original English version. Further testing of this scale with older Chinese people living in urban areas is needed to evaluate the reliability of the SEE-C scale and explore the reasons for the low R2 values. The differences in the participants’ characteristics between the present study and the previous investigations of the SEE scale (Resnick and Jenkins, 2000) include race (Chinese vs. American), education (fairly educated vs. well educated), gender (more male vs. more female), the mean age of the participants (71.2 6.0 years vs. 85 6.2 years), and marital status (mainly married vs. unmarried). There were more participants involved in regular physical activity in the present study than in the previous one (53% vs. 38%) (Resnick and Jenkins, 2000), whilst the exercise self-efficacy scores were higher in Resnick et al.’s study (2000) (4.2 2.6 vs. 5.5 3). These differences may be explained by the sources of the population – communitydwelling older residents vs. a continuing care retirement community who had an ongoing walking programme, exercise room, and a safe indoor environment. As a result, those individuals were exposed to many sources of efficacy information to reinforce a generally high level of efficacy expectations. The above differences may cause the variation in the results of reliability and validity testing and need further study. The cultural relevance of the SEE scale is another area warranting further examination. Although acceptable levels of reliability and validity of the SEE-C scale for older
L.-L. Lee et al. / International Journal of Nursing Studies 46 (2009) 230–238
Chinese people in Taiwan were attained, cultural factors appeared to play a role in the applicability of some of the items on the scale. For example, the reason why item 2 (you were bored by the programme or activity) and item 6 (you were too busy with other activities) had the lowest squared multiple correlation coefficients (R2) may relate to the intention to adopt and maintain exercise activity and the time management of daily living among older people in Taiwan. As older people’s intention to exercise is mainly for the sake of their health to prevent becoming a burden (Lee et al., 2007a,b) and participants would usually exercise very early in the morning, those two items may not be the factors that influence their confidence to ‘‘exercise three times per week for 20 min’’. Perhaps then items reflecting the typical cultural behaviour of older Chinese adults (such as the responsibility for looking after their grandchildren) should be added to obtain a more reliable and valid measure. Several studies have also illustrated that self-efficacy might be influenced by some culturally specific elements, such as language (Scholz et al., 2002) and cultural beliefs (Shellman and Shellman, 2006). The present preliminary validation study of the SEE-C scale is part of a randomised controlled trial carried out among a group of mild to moderate hypertensive older people. These older participants with high blood pressure may limit the generalisability of the findings although previous study has found that there were about 80% of older adults having at least one chronic condition (Miller, 1999). Future studies are recommended to examine the reliability and validity of the SEE-C scale among different Chinese populations with different characteristics. There are a number of possible reasons for the issue of a slight male majority in our sample. Participants were recruited from each village’s community activity centre. The village’s community activity centres were usually used by older male people as they were spaces that encouraged ‘male’ activities, such as playing billiards, cards, etc. Recruitment to the study took place during the day when many older women were looking after their grandchildren. Another limitation of the present study is that the exercise behaviour was based on self-report. Prior research has found that the reliance on self-reported walking and physical activity may limit the validity of the data, as participants over 65 may tend to overestimate the amount of their physical activity (Sims et al., 1999). More objective measures of exercise, such as including
235
an accelerometer, are needed to better establish the relationship between exercise self-efficacy and exercise behaviour.
5. Conclusion To our knowledge, the SEE-C scale is the first Chinese scale for measuring SEE among older people, and this study is preliminarily systematic examination of its reliability and validity. The results obtained in the present study provided acceptable levels of reliability and validity for the scale when used with older people in Taiwan. This indicates that the Chinese version of the SEE scale can be employed to measure exercise self-efficacy among older people with at least one chronic health condition. Among the older Chinese adults, the findings can be used to facilitate the measure for assessing confidence regarding carrying out exercise on different occasions. Data obtained from the scale may also provide information for the design of community-based physical activity interventions. Although the instrument showed moderate reliability and validity, further improvements are needed. The strengths and weaknesses of the scale will be identified clearly through its future use in clinical practice, including in the community, and in research studies.
Conflict of interest There are no potential conflicts of interest known to any of the contributing authors.
Acknowledgments The authors would like to express their sincere gratitude to all of the participants in this study, for the kind assistance of the Public Health Nurses at JiAnn Health Station in Hualien, Taiwan, and to the leaders and volunteers of the nine villages’ community activity centres for providing study accommodation; and advisory personnel Prof. Mark Avis and Dr. Aisha Holloway; the financial support from The Tzu-Chi College of Technology for Ling-Ling Lee’s PhD study. Deep appreciation is expressed to Dr. Barbara Resnick who kindly offered her guidance throughout the course of this research project.
236
L.-L. Lee et al. / International Journal of Nursing Studies 46 (2009) 230–238
Appendix A. Chinese self-efficacy for exercise scale (SEE-C)
References Aittasalo, M., Miilunpalo, S., Stahl, T., Kukkonen-Harjula, K., 2007. From innovation to practice: initiation, implementation and evaluation of a physician-based physical activity promotion programme in Finland. Health Promotion International 22 (1), 19–27. Arbuckle, J., 1997. Amos Users’ Guide Version 3.6. Small Waters Corporation, Chicago. Bagozzi, R.P., Yi, Y., 1988. On the evaluation of structural equation models. Journal of Academy of Marketing Science 16 (1), 74–94. Bandura, A., 1986a. Cognitive regulators. In: Bandura, A. (Ed.), Social Foundations of Thought and Action: A Social Cognitive Theory. Prentice-Hall, Englewood Cliffs, NJ, pp. 467–480.
Bandura, A., 1986b. Self-efficacy. In: Bandura, A. (Ed.), Social Foundations of Thought and Action: A Social Cognitive Theory. Prentice-Hall, Englewood Cliffs, NJ, pp. 399–409. Bandura, A., 1997. Self-efficacy: The Exercise of Control. W.H. Freeman, New York. Bandura, A., 1977. Self-efficacy: toward a unifying theory of behavioral change. Psychological Review 84 (2), 191–215 (comment). Bollen, K.A., 1989. Structural Equations With Latent Variables. Wiley, New York. Bozoian, S., Rejeski, W.J., McAuley, E., 1994. Self-efficacy influences feeling states associated with acute exercise. Journal of Sport and Exercise Psychology 16 (3), 326–333. Bracken, B.A., Barona, A., 1991. State of the art procedures for translating, validating and using psychoeducational tests in
L.-L. Lee et al. / International Journal of Nursing Studies 46 (2009) 230–238 cross-cultural assessment. School Psychology International 12 (1–2), 119–132. Bradley, C., 1994. Translation of questionnaires for use in different languages and cultures. In: Bradley, C. (Ed.), Handbook of Psychology: A Guide to Psychological Measurement in Diabetes Research and Management. Harwood Academic Publishers, Singapore, pp. 43–56. Chang, A.M., Chau, J.P., Holroyd, E., 1999. Translation of questionnaires and issues of equivalence. Journal of Advanced Nursing 29 (2), 316–322. Clark, D.O., 1996. Age, socioeconomic status, and exercise selfefficacy. Gerontologist 36 (2), 157–164. Conn, V.S., 1998. Older adults and exercise: path analysis of selfefficacy related constructs. Nursing Research 47 (3), 180–189. Cronbach, L.J., 1990. Essentials of Psychological Testing. Harper & Row, New York. Eysenck, H.J., 1978. Expectations as causal elements in behavioural change. Advances in Behaviour Research and Therapy 1, 171– 175. Garcia, A.W., King, A.C., 1991. Predicting long-term adherence to aerobic exercise: a comparison of two models. Journal of Sport and Exercise Psychology 13 (4), 394–410. Gecas, V., 1989. The social psychology of self-efficacy. Annual Review of Sociology 15, 291–316. Gortner, S.R., Jenkins, L.S., 1990. Self-efficacy and activity level following cardiac surgery. Journal of Advanced Nursing 15 (10), 1132–1138. Grembowski, D., Patrick, D., Diehr, P., Durham, M., Beresford, S., Kay, E., Hecht, J., 1993. Self-efficacy and health behavior among older adults. Journal of Health & Social Behavior 34 (2), 89–104. Hawkins, S.A., Cockburn, M.G., Hamilton, A.S., Mack, T.M., 2004. An estimate of physical activity prevalence in a large populationbased cohort. Medicine & Science in Sports & Exercise 36 (2), 253–260. Health Bureau, 2002. Survey of Exercise Behavior. Health Bureau, Department of Health, Taiwan, Taipei. Health Education Authority UK, 1999. Physical Activity in Later Life. Health Education Authority, London, p. 96. Jagodzinski, W., Kuhnel, S.M., 1987. Estimation of reliability and stability in single-indicator multiple-wave models. Sociological Methods and Research 15 (3), 219–258. Judge, T.A., Erez, A., Bono, J.E., Thoresen, C.J., 2002. Are measures of self-esteem, neuroticism, locus of control, and generalized self-efficacy indicators of a common core construct? Journal of Personality & Social Psychology 83 (3), 693–710. Lee, L.L., Arthur, A., Avis, M., 2007a. Evaluating a communitybased walking intervention for hypertensive older people in Taiwan: a randomized controlled trial. Preventive Medicine 44 (2), 160–166. Lee, L.L., Avis, M., Arthur, A., 2007b. The role of self-efficacy in older people’s decisions to initiate and maintain regular walking as exercise—findings from a qualitative study. Preventive Medicine 45 (1), 62–65. Loehlin, J.C., 2004. Latent Variable Models: An Introduction to Factor, Path, and Structural Equation Analysis. Lawrence Erlbaum Associates, Mahwah, NJ. Long, J.D., Stevens, K.R., 2004. Using technology to promote selfefficacy for healthy eating in adolescents. Journal of Nursing Scholarship 36 (2), 134–139.
237
Mason-Hawkes, J., Holm, K., 1993. Gender differences in exercise determinants. Nursing Research 42 (3), 166–172. McAuley, E., 1992. The role of efficacy cognitions in the prediction of exercise behavior in middle-aged adults. Journal of Behavioral Medicine 15 (1), 65–88. McAuley, E., Courneya, K.S., 1992. Self-efficacy relationships with affective and exertion responses to exercise. Journal of Applied Social Psychology 22 (4), 312–326. Miller, C.A., 1999. Nursing Care of Older Adults: Theory and Practice. Lippincott Williams & Wilkins, Philadelphia. Muris, P., 2001. A brief questionnaire for measuring self-efficacy in youths. Journal of Psychopathology and Behavioral Assessment 23 (3), 145–149. Nunnally, J., Bernstein, I., 1994. Psychometric Theory. McGrawHill, New York. Resnick, B., 1998a. Efficacy beliefs in geriatric rehabilitation. Journal of Gerontological Nursing 24 (7), 34–44. Resnick, B., 2002. Geriatric rehabilitation: the influence of efficacy beliefs and motivation. Rehabilitation Nursing 27 (4), 152–159. Resnick, B., 1999. Reliability and validity testing of the self-efficacy for functional activities scale. Journal of Nursing Measurement 7 (1), 5–20. Resnick, B., Jenkins, L.S., 2000. Testing the reliability and validity of the self-efficacy for exercise scale. Nursing Research 49 (3), 154–159. Resnick, B., Palmer, M.H., Jenkins, L.S., Spellbring, A.M., 2000. Path analysis of efficacy expectations and exercise behavior in older adults. Journal of Advanced Nursing 31 (6), 1309–1315. Roach, J.B., Yadrick, M.K., Johnson, J.T., Boudreaux, L.J., Forsythe 3rd, W.A., Billon, W., 2003. Using self-efficacy to predict weight loss among young adults. Journal of the American Dietetic Association 103 (10), 1357–1359. Rodin, J., 1989. Sense of control: potentials for intervention. Annals of the American Academy of Political & Social Science 503 (1), 29–42. Ronen, G.M., Streiner, D.L., Rosenbaum, P., Canadian Pediatric Epilepsy Network, 2003. Health-related quality of life in children with epilepsy: development and validation of self-report and parent proxy measures. Epilepsia 44 (4), 598–612. Sallis, J.F., Criqui, M.H., Kashani, I.A., Rupp, J.W., Calfas, K.J., Langer, R.D., Nader, P.R., Ross Jr., J., 1990. A program for health behavior change in a preventive cardiology center. American Journal of Preventive Medicine 6 (Suppl. 2), 43–50. Scholz, U., Gutierrez-Dona, B., Sud, S., Schwarzer, R., 2002. Is perceived self-efficacy a universal construct? Psychometric findings from 25 countries. European Journal of Psychological Assessment 18 (3), 242–251. Schunk, D.H., Carbonari, J.P., 1984. Self-efficacy models. In: Matarazzo, J.D., weiss, S.M., Herd, J.A., Miller, N.E., Weiss, S.M. (Eds.), Behavioural Health: A Handbook of Health Enhancement and Disease Prevention. Wiley, New York, pp. 230–247. Shellman, J., Shellman, J., 2006. Development and psychometric evaluation of the eldercare cultural self-efficacy scale. International Journal of Nursing Education Scholarship 3 (Article 9). Sims, J., Smith, F., Duffy, A., Hilton, S., 1999. The vagaries of selfreports of physical activity: a problem revisited and addressed in a study of exercise promotion in the over 65s in general practice. Family Practice 16 (2), 152–157.
238
L.-L. Lee et al. / International Journal of Nursing Studies 46 (2009) 230–238
Stutts, W.C., 2002. Physical activity determinants in adults. Perceived benefits, barriers, and self efficacy. American Association of Occupational Health Nurses (AAOHN) Journal 50 (11), 499–507. The Department of Statistics Ministry of the Interior ROC, 2007. Percentage of Older People Over 65 in Total Population Among
Major Countries. The Department of Statistics Ministry of the Interior ROC, Taipei. Tipton, R.M., 1984. The measurement of generalized self-efficacy: a study of construct validity. Journal of Personality Assessment, Lawrence Erlbaum Associates 545–548.
International Journal of Nursing Studies 46 (2009) 230–238 www.elsevier.com/ijns
A preliminary reliability and validity study of the Chinese version of the self-efficacy for exercise scale for older adults Ling-Ling Lee a,*, Shoa-Jen Perng a, Chin-Chih Ho b, Hsiang-Ming Hsu c, Shu-Chuan Lau d, Antony Arthur e a
Department of Nursing, Tzu Chi College of Technology, and School of Nursing, Tzu Chi University, Hualien, Taiwan b Department of Health Administration, Tzu-Chi College of Technology, Hualien, Taiwan c Department of Public Health, Tzu Chi University, Hualien, Taiwan d Medical School, Fu Jen Catholic University, Taipei, Taiwan e School of Nursing, University of Nottingham, Nottingham, UK Received 28 February 2008; received in revised form 16 August 2008; accepted 4 September 2008
Abstract Background: Population ageing affects most undeveloped, developing and developed countries. Less than a quarter of older people undertake a level of physical activity worldwide that is sufficient to lead to health benefits. Understanding older people’s confidence regarding engaging in exercise helps to structure physical activity interventions that motivate them to initiate and adhere to regular exercise. Estimates of the reliability and validity of the English version of the self-efficacy for exercise (SEE) scale has been widely tested and shown to be valid for use in various settings and among older people. The reliability and validity of a Chinese version of the SEE for older adults has not been tested. Aim: To undertake a preliminary assessment of the reliability and validity of the SEE scale when applied to the older Chinese adults. Methods: The Chinese version of the self-efficacy for exercise (SEE-C) scale was tested on a sample of 192 older people from Taiwan with a mean age of 71.2 years recruited between October and December 2003. Results: There was acceptable internal consistency of the SEE-C scale. The fit of the measurement model to the data for the SEE-C scale was acceptable. There was evidence of validity of the measure based on hypothesis testing: health status predicted exercise self-efficacy, and exercise self-efficacy predicted physical activity. Those who exercised regularly (n = 102) had a higher mean SEE-C score (5.3) than those who did not (2.9, n = 90). Conclusion: This preliminary validation study provided evidence for the reliability and validity of the Chinese version of the SEE scale. Future testing of the SEE-C scale needs to be carried out to see whether these results are generalisable to older Chinese people living in urban areas and with different characteristics. A test of the scale among younger adults for a wider use of the instrument is also warranted. # 2008 Elsevier Ltd. All rights reserved. Keywords: Chinese self-efficacy for exercise scale (SEE-C); Older people; Confidence; Validity; Reliability
What is already known about the topic? * Corresponding author at: Tzu Chi College of Technology, Department of Nursing, Room 429, No. 880, Chien-Kuo Road, Section 2, Hualien 970, Taiwan. E-mail address: [email protected] (L.-L. Lee).
Self-efficacy plays a role in the maintenance of physical activity beyond programme involvement, as well as in the early stages of exercise adoption.
0020-7489/$ – see front matter # 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijnurstu.2008.09.003
L.-L. Lee et al. / International Journal of Nursing Studies 46 (2009) 230–238
Understanding older people’s confidence regarding engaging in exercise will help to structure physical activity interventions that motivate them to initiate and adhere to regular exercise programmes. English version of the self-efficacy for exercise (SEE) scale has been shown to be valid for use in various settings and particularly among older people. What this paper adds The Chinese version of the SEE scale is likely to be internally consistent across its nine items although individual items explain relatively little of the overall variance. This study provided evidence for good criterion-related validity of the SEE-C but not for its construct validity. 1. Background Population ageing affects most undeveloped, developing and developed countries (The Department of Statistics Ministry of the Interior ROC, 2007). Despite the well-established health benefits of physical activity, surveys suggest that over 60% of older people do not participate in regular exercise, and only about 25% of this age group report regular participation in physical activity (Hawkins et al., 2004; Health Education Authority UK, 1999). In Taiwan, up to 27% of men and 39% of women aged over 65 years do not participate in exercise activity (Health Bureau, 2002). Understanding older people’s confidence regarding engaging in exercise will help to structure physical activity interventions that motivate them to initiate and adhere to regular exercise programmes. Self-efficacy theory is one of the most widely applied theories in predicting health behaviour. Self-efficacy, a central concept of Bandura’s social learning theory (Bandura, 1977), attempts to predict and explain human behaviour by referring to an individual’s assessment of their competence or effectiveness to perform a specific behaviour successfully. The theory of self-efficacy suggests that the stronger the individual’s self-efficacy and outcome expectations, the more likely it is that he or she will initiate and persist with a given activity. The theory of self-efficacy incorporates self-efficacy expectations, which are an individual’s beliefs in their capabilities to perform a course of action to attain a desired outcome; and outcome expectations, which are the beliefs that a certain consequence will be produced by personal action (Bandura, 1986a; Gecas, 1989; Schunk and Carbonari, 1984). There is an evidence from empirical studies of a relationship between self-efficacy and a variety of health-related behaviour, such as healthy eating (Long and Stevens, 2004) and weight loss (Roach et al., 2003). Regarding exercise behaviour, self-efficacy plays a role in the maintenance of physical activity beyond programme involvement, as well as in the early stages of exercise adoption (Aittasalo et al.,
231
2007; Garcia and King, 1991; McAuley, 1992; Sallis et al., 1990). The relationship between self-efficacy expectations and level of exercise is shown in various studies, in spite of the diverse samples, measurement instruments, and measurement timing (Clark, 1996; Gortner and Jenkins, 1990; Stutts, 2002). People with higher self-efficacy expectations were found to maintain a sense of energy during exercise, perceive less effort being expended during exercise, and report both more positive affect during and after exercise and feeling more revitalised (Bozoian et al., 1994; McAuley and Courneya, 1992). To improve exercise activity in older adults, it is suggested that self-efficacy expectations related to exercise are considered, because these beliefs influence motivation to exercise and actual exercise activity (Resnick and Jenkins, 2000). Several generic self-efficacy measurements have been developed (Judge et al., 2002; Muris, 2001; Tipton, 1984) as self-efficacy does enjoy certain generic features concerning the judgments about one’s self-efficacy (Bandura, 1997). However, the usefulness of the generic self-efficacy measurements has been questioned (Bandura, 1986b; Eysenck, 1978). An individual’s confidence in carrying out different behaviours may vary across behaviours and contexts. Bandura (1997) argues that there is no universal measure of perceived self-efficacy and that any scale should be tailored to the particular domain of interest. Without reference to the specific behaviour and situation, it is inappropriate to label an individual as having ‘high’ or ‘low’ self-efficacy. Therefore, the task of the researcher is to measure self-efficacy within a specific focus, such as self-efficacy in drug abstention, quitting smoking, or selfefficacy in doing exercise. An English version of the SEE scale was designed to test people’s confidence to continue exercising in the face of barriers to exercise (Resnick and Jenkins, 2000). Estimates of the reliability and validity of the nine item SEE scale have been widely tested and shown to be valid for use in various settings, particularly among older people with good internal consistency and validity. For example, there is evidence of internal consistency (alpha = 0.93) and validity with efficacy expectations significantly related to exercise activity, and factor loadings all greater than 0.50 (Resnick, 1999,2002; Resnick and Jenkins, 2000; Resnick et al., 2000). The theoretical basis, clinical feasibility, and some acceptable psychometric test features of the SEE scale have made it a valuable tool for the evaluation of SEE among North American populations. However, whether or not these estimates hold true for older people from different social and ethnic backgrounds needs to be examined. Little data exist on the use of the SEE scale among older Chinese people living in the community and how cultural differences might affect the administration of the SEE scale has not been investigated. Before the SEE scale can be applied more broadly, a cross-cultural evaluation is warranted. The purpose of the present study is to evaluate if there is evidence for the reliability and validity of
232
L.-L. Lee et al. / International Journal of Nursing Studies 46 (2009) 230–238
the SEE scale when used with older Chinese adults in Taiwan.
2. Methods The participants were obtained from the baseline measures of a randomised controlled trial carried out in Taiwan (n = 202) (Lee et al., 2007a,b) and recruited between October and December 2003. The trial was designed to examine a 6-month community-based walking intervention using selfefficacy theory to enhance confidence in regular walking as exercise and to examine the effect on blood pressure control. Data were collected concerning the participants’ age (over 60 years), systolic blood pressure (between 140 mm and 179 mmHg), and able to carry out regular walking as exercise. The original English version of the SEE scale was translated into Chinese. As with the original English version, the translated instrument contained nine items measuring the participants’ perceived ability to exercise under various circumstances (such as bad weather and tiredness). Two professionals in the fields of gerontology and public health in Taiwan were consulted for the wording and phrasing of each item of the initial Chinese version of the scale. Two items of the translated SEE scale were raised as being of concern with regard to the Chinese words employed to describe ‘‘enjoy’’ (item 5 ‘‘You did not enjoy it’’) and ‘‘depressed’’ (item 9 ‘‘You felt depressed’’). A modified Chinese version was developed after a series of meetings with members of the study team. This procedure was intended to ensure the Chinese version of the SEE (SEE-C) scale to be linguistically appropriate (Bracken and Barona, 1991; Bradley, 1994; Chang et al., 1999). Twenty-two older people living in another community were recruited to pilot test the translated measure for readability and feasibility of completion. The data of the SEE-C scale were collected by a set of questionnaires and administered by a nurse researcher to a group of older people living in the community via face-toface interviews. The participants were instructed to listen to the statement and then choose an option from 0 (not confident) to 10 (very confident) that represented their perception of confidence regarding engaging in regular exercise for 20 min three times per week. In prior research, older adults were able to respond to the 0–10 scale questionnaire (Resnick, 1998a; Resnick and Jenkins, 2000). Therefore, the original format of the 10-point scale was retained in the Chinese version. In addition, an answer card was designed with larger words stating all possible choice of responses and used to facilitate participants to rate their perceptions of confidence about doing exercise. The scale was scored by summing the numerical ratings for each response and dividing the total by the number of non-missing responses. The mean scores for the self-efficacy of exercise ranged from 0 to 10, with the higher scores representing greater exercise selfefficacy.
Several other measures were used to test the validity of the SEE-C scale. These measures included demographic information, physical activity status, and perceived health. The demographic information included age, gender, and years of education. Physical activity was measured by asking how often the participant exercised (e.g. playing table tennis, cycling, hiking, and swimming) and walked for at least 20 min in one go. The participants rated their physical activity as one of: never; less than once a week; once a week; two to three times a week; or more than three times a week. Regular exercisers were defined as adopting any form of physical activity three times a week and at least 20 min in one go and maintaining the physical activity for 3 months. Perceived health was measured by three items used in others’ work (Mason-Hawkes and Holm, 1993) that are summed to create a score of between 3 and 10 with higher scores indicating better perceived health. The participants rated their health as ‘very good’, ‘good’, ‘poor’, or ‘very poor’. They then compared their health to others similar to in age to themselves and rated it as ‘better than average’, ‘average’, or ‘worse than average’ (Mason-Hawkes and Holm, 1993). Finally, participants rated the extent to which their health interfered with their desired activities as ‘not at all’, ‘somewhat’, or ‘a great deal’. The criterion validity of this measure of perceived health has been supported by findings that suggest positive correlations with physician visits (r = 0.25) (Mason-Hawkes and Holm, 1993), number of reported symptoms (r = 0.38) (Mason-Hawkes and Holm, 1993), number of chronic illnesses (r = 0.46) (Conn, 1998) and number of prescriptive medications (r = 0.40) (Conn, 1998). The internal consistency of the perceived health scale, as measured by Cronbach’s alpha was 0.67 from the pilot testing (n = 22), consistent with an alpha of 0.76 reported in a previous study (alpha of 0.76) (Conn, 1998). 2.1. Statistical analysis Descriptive analyses of the data were performed with all study variables to describe the study participants. Of the 202 participants recruited, 10 were not used because of missing responses to at least three questions in the SEE-C scale resulting in a sample of 192. 2.1.1. Reliability testing Estimates of internal consistency were carried out using Cronbach’s alpha. In the present study, an alpha coefficient of 0.7 or greater was considered as evidence of the internal consistency of the SEE-C scale (Nunnally and Bernstein, 1994). This indicates that 70% of the variance of the observed score is systematic, and 30% is due to random errors. However, internal consistency cannot be used in a single item measure. An alternative estimate of reliability using a structural equation modelling, a squared multiple correlation coefficient (R2) was used (Bollen, 1989), as R2 estimates the systematic variance in the observed score that can be explained by each item in the measurement model
L.-L. Lee et al. / International Journal of Nursing Studies 46 (2009) 230–238
(Bollen, 1989; Jagodzinski and Kuhnel, 1987). For example, if item 1 R2 = 0.68, 68% of the variance in item 1 is accounted for by the variance in exercise self-efficacy. The remaining 32% of the variance cannot be accounted for and is attributed to error. Ideally, the R2 for each item, or the amount of variance accounted for, should be at least 0.5 (Bagozzi and Yi, 1988). 2.1.2. Validity testing Validity testing for the SEE-C scale was based on a confirmatory factor analysis using structural equation modelling. The chi-square statistic, the Normed Fit Index (NFI), and Steigers’ root mean square error of approximation (RMSEA) were used to estimate model fit. The larger the probability associated with the chi-square, the better the fit of the model to the data (Loehlin, 2004). Because the chisquare statistic is influenced by sample size, the chi-square divided by degrees of freedom was used to evaluate the model fit. Ideally, this ratio should be below 3 (Arbuckle, 1997). The NFI tests the hypothesised model against a reasonable baseline model and should yield a value of 1.0. The RMSEA, unlike the chi-square statistic, gives an estimate of model fit that is not influenced by sample size. An RMSEA of <0.10 is considered good, and <0.05 is very good (Loehlin, 2004). Validity testing for the SEE-C scale was also evaluated by examining its construct validity and concurrent validity. Construct validity of the SEE-C scale was tested using an evidence supported hypothesis that higher self-efficacy expectations are more likely to be observed among individuals with better health status (Grembowski et al., 1993; Rodin, 1989) and operationalised by observing whether there was a positive correlation between perceived health and SEE-C score. Concurrent validity was used to examine criterion-related validity. As criterion validity refers to the ability of an instrument to correlate well with a criterion that is related to the construct measured by the instrument (Cronbach, 1990; Nunnally and Bernstein, 1994), it is assumed that self-efficacy expectations had a directly significant effect on exercise behaviour (Ronen et al., 2003) in the present study. This had been operationalised by observing if there would be a positive
233
correlation between SEE-C score and level of physical activity frequency. The correlations were assessed using Pearson’s correlation coefficient (r). Multiple regression was used for the hypothesis testing. Four variables (age, gender, education year, and perceived health) were included as predictors of self-efficacy expectations for exercise. Multiple regression was also used for testing criterion-related validity with age, gender and education year adjusted for. Additionally, the score for self-efficacy expectations were included as a predictor of physical activity. The original trial was reviewed by the University of Nottingham Medical School Ethics Committee and ethical approval was granted in July 2003 (Study reference number: D/6/2003). Signed informed consent was obtained prior to the data collection.
3. Results There were more male subjects than females in the study (57.8% vs. 42.2%), with a mean age of 71.2 and mean education year of 4.2 (n = 192). The mean scores of the study participants on the SEE-C scale, their perceived health, and levels of physical activity frequencies are presented in Table 1. The mean SEE-C score was 4.2 on a scale of 0– 10 and the mean perceived health score was 5.7 on a scale of 3–10. Slightly more than half (102/192, 53.1%) of the participants took part in regular physical activity. 3.1. Reliability and validity The Cronbach’s alpha coefficient was 0.75 (n = 192), which was lower than the original English version, which was 0.92 (n = 187) (Resnick and Jenkins, 2000) but acceptable for the internal consistency of the SEE-C measure. The squared multiple correlation coefficients (R2) were all less than 0.50 and ranged from 0.15 (item 6) to 0.36 (item 4). There was support for the confirmatory factor analysis of the Chinese version of the SEE scale. Both the chi-square (x2 = 45, d.f. = 27, p = 0.016) and x2 divided by degrees of
Table 1 Demographic data and mean scores for the study measures (n = 192).
Age (years) Educational (years) Resting systolic blood pressure (mmHg) Resting diastolic blood pressure (mmHg) Chinese version of self-efficacy for exercise (SEE-C) scale Perceived health Men/women Participants doing regular physical activity S.D. = standard deviation.
Mean (S.D.)
Range
71.2 4.2 151.9 82.3 4.2 5.7
60–91 0–15 140–180 60–112 0–10 3–10
(6.0) (4.1) (10.6) (9.8) (2.6) (1.7)
n (%)
111(57.8)/81(42.2) 102 (53.1)
234
L.-L. Lee et al. / International Journal of Nursing Studies 46 (2009) 230–238
Table 2 Bivariate Correlations of the major study variables (n = 192). SEE-C SEE-C Physical activity Perceived health Outcome expectations for exercise
1.00 0.46** 0.17* 0.36**
Physical activity 1.00 0.19* 0.23**
Perceived health
Outcome expectations for exercise
1.00 0.13
1.00
Note: SEE-C = Chinese version of self-efficacy for exercise scale. *p < 0.05, **p < 0.01 (Pearson’s correlation coefficient).
freedom (x2/d.f. ratio = 1.67) suggested a fair model fit (Arbuckle, 1997). Moreover, the NFI was 0.90, and RMSEA was less than 0.10 (0.059). In total, the nine items explained 45% of the variance in exercise self-efficacy. The evidence for the criterion-related validity of the SEE-C scale was supported by the positive correlation between SEE-C score and level of physical activity (r = 0.46, p < 0.0001) (Table 2). After adjusting for age and gender, the SEE-C scores predicted exercise activity (F = 16.83, p < 0.0001), accounting for 44.6% of the variance in exercise activity. Evidence of criterion-related validity of the SEE-C scale was further supported by a statistically significant difference between SEE-C scores of those exercising regularly (at least 20 min three times per week) and those who did not (t = 6.88, p < 0.0001). Those who exercised regularly had a mean SEE-C score of 5.3 (S.D. = 2.4), whereas those who did not had a mean score of 2.9 (S.D. = 2.3). The evidence for the construct validity of SEE-C scale is weak. Although the Pearson’s correlation revealed a statistically significant correlation between perceived health and SEE-C score (r = 0.17, p = 0.019) (Table 2), the association between perceived heath and exercise self-efficacy was weak. When controlled for age, gender and education year, the scores of perceived health subscale significantly predicted the SEE-C scores (F = 3.43, p = 0.04), but accounted for only 15.1% of the variance in exercise self-efficacy. However, there was evidence for the construct validity of the SEE-C scale based on a statistically significant correlation between exercise self-efficacy and exercise outcome expectations (r = 0.36, p < 0.0001) (Table 2).
4. Discussion The present study provides preliminary evidence for the reliability and validity of the SEE-C scale. The participants in this study were community-based, from a rural area of Taiwan, with at least one chronic health problem (mild to moderate hypertension). The diversity of the participants contributes to the generalisability of the use of the scale and the findings can be used to strengthen the measure of exercise self-efficacy among older Chinese people for future use. There was acceptable evidence of internal consistency, tested by Cronbach’s alpha. However, there was limited
evidence of the reliability, based on squared multiple correlation coefficients (R2). Only 36% of the variance in item 4 was explained by the model. There was evidence for the construct validity of the measure based on statistically significant relationship between perceived health and exercise self-efficacy and between exercise outcome expectations and exercise selfefficacy. The evidence of the criterion-related validity of the measure was supported by a statistically significant correlation between exercise self-efficacy and physical activity behaviour. Although these relationships were statistically significant, it is recognised that these correlations were weak. This may be related to the fact that many factors influence exercise and other forms of physical activity behaviours, particularly among older people. Confirmatory factor analysis indicated that the fit of the full model to the data was acceptable but relatively lower than that of the original English version. Further testing of this scale with older Chinese people living in urban areas is needed to evaluate the reliability of the SEE-C scale and explore the reasons for the low R2 values. The differences in the participants’ characteristics between the present study and the previous investigations of the SEE scale (Resnick and Jenkins, 2000) include race (Chinese vs. American), education (fairly educated vs. well educated), gender (more male vs. more female), the mean age of the participants (71.2 6.0 years vs. 85 6.2 years), and marital status (mainly married vs. unmarried). There were more participants involved in regular physical activity in the present study than in the previous one (53% vs. 38%) (Resnick and Jenkins, 2000), whilst the exercise self-efficacy scores were higher in Resnick et al.’s study (2000) (4.2 2.6 vs. 5.5 3). These differences may be explained by the sources of the population – communitydwelling older residents vs. a continuing care retirement community who had an ongoing walking programme, exercise room, and a safe indoor environment. As a result, those individuals were exposed to many sources of efficacy information to reinforce a generally high level of efficacy expectations. The above differences may cause the variation in the results of reliability and validity testing and need further study. The cultural relevance of the SEE scale is another area warranting further examination. Although acceptable levels of reliability and validity of the SEE-C scale for older
L.-L. Lee et al. / International Journal of Nursing Studies 46 (2009) 230–238
Chinese people in Taiwan were attained, cultural factors appeared to play a role in the applicability of some of the items on the scale. For example, the reason why item 2 (you were bored by the programme or activity) and item 6 (you were too busy with other activities) had the lowest squared multiple correlation coefficients (R2) may relate to the intention to adopt and maintain exercise activity and the time management of daily living among older people in Taiwan. As older people’s intention to exercise is mainly for the sake of their health to prevent becoming a burden (Lee et al., 2007a,b) and participants would usually exercise very early in the morning, those two items may not be the factors that influence their confidence to ‘‘exercise three times per week for 20 min’’. Perhaps then items reflecting the typical cultural behaviour of older Chinese adults (such as the responsibility for looking after their grandchildren) should be added to obtain a more reliable and valid measure. Several studies have also illustrated that self-efficacy might be influenced by some culturally specific elements, such as language (Scholz et al., 2002) and cultural beliefs (Shellman and Shellman, 2006). The present preliminary validation study of the SEE-C scale is part of a randomised controlled trial carried out among a group of mild to moderate hypertensive older people. These older participants with high blood pressure may limit the generalisability of the findings although previous study has found that there were about 80% of older adults having at least one chronic condition (Miller, 1999). Future studies are recommended to examine the reliability and validity of the SEE-C scale among different Chinese populations with different characteristics. There are a number of possible reasons for the issue of a slight male majority in our sample. Participants were recruited from each village’s community activity centre. The village’s community activity centres were usually used by older male people as they were spaces that encouraged ‘male’ activities, such as playing billiards, cards, etc. Recruitment to the study took place during the day when many older women were looking after their grandchildren. Another limitation of the present study is that the exercise behaviour was based on self-report. Prior research has found that the reliance on self-reported walking and physical activity may limit the validity of the data, as participants over 65 may tend to overestimate the amount of their physical activity (Sims et al., 1999). More objective measures of exercise, such as including
235
an accelerometer, are needed to better establish the relationship between exercise self-efficacy and exercise behaviour.
5. Conclusion To our knowledge, the SEE-C scale is the first Chinese scale for measuring SEE among older people, and this study is preliminarily systematic examination of its reliability and validity. The results obtained in the present study provided acceptable levels of reliability and validity for the scale when used with older people in Taiwan. This indicates that the Chinese version of the SEE scale can be employed to measure exercise self-efficacy among older people with at least one chronic health condition. Among the older Chinese adults, the findings can be used to facilitate the measure for assessing confidence regarding carrying out exercise on different occasions. Data obtained from the scale may also provide information for the design of community-based physical activity interventions. Although the instrument showed moderate reliability and validity, further improvements are needed. The strengths and weaknesses of the scale will be identified clearly through its future use in clinical practice, including in the community, and in research studies.
Conflict of interest There are no potential conflicts of interest known to any of the contributing authors.
Acknowledgments The authors would like to express their sincere gratitude to all of the participants in this study, for the kind assistance of the Public Health Nurses at JiAnn Health Station in Hualien, Taiwan, and to the leaders and volunteers of the nine villages’ community activity centres for providing study accommodation; and advisory personnel Prof. Mark Avis and Dr. Aisha Holloway; the financial support from The Tzu-Chi College of Technology for Ling-Ling Lee’s PhD study. Deep appreciation is expressed to Dr. Barbara Resnick who kindly offered her guidance throughout the course of this research project.
236
L.-L. Lee et al. / International Journal of Nursing Studies 46 (2009) 230–238
Appendix A. Chinese self-efficacy for exercise scale (SEE-C)
References Aittasalo, M., Miilunpalo, S., Stahl, T., Kukkonen-Harjula, K., 2007. From innovation to practice: initiation, implementation and evaluation of a physician-based physical activity promotion programme in Finland. Health Promotion International 22 (1), 19–27. Arbuckle, J., 1997. Amos Users’ Guide Version 3.6. Small Waters Corporation, Chicago. Bagozzi, R.P., Yi, Y., 1988. On the evaluation of structural equation models. Journal of Academy of Marketing Science 16 (1), 74–94. Bandura, A., 1986a. Cognitive regulators. In: Bandura, A. (Ed.), Social Foundations of Thought and Action: A Social Cognitive Theory. Prentice-Hall, Englewood Cliffs, NJ, pp. 467–480.
Bandura, A., 1986b. Self-efficacy. In: Bandura, A. (Ed.), Social Foundations of Thought and Action: A Social Cognitive Theory. Prentice-Hall, Englewood Cliffs, NJ, pp. 399–409. Bandura, A., 1997. Self-efficacy: The Exercise of Control. W.H. Freeman, New York. Bandura, A., 1977. Self-efficacy: toward a unifying theory of behavioral change. Psychological Review 84 (2), 191–215 (comment). Bollen, K.A., 1989. Structural Equations With Latent Variables. Wiley, New York. Bozoian, S., Rejeski, W.J., McAuley, E., 1994. Self-efficacy influences feeling states associated with acute exercise. Journal of Sport and Exercise Psychology 16 (3), 326–333. Bracken, B.A., Barona, A., 1991. State of the art procedures for translating, validating and using psychoeducational tests in
L.-L. Lee et al. / International Journal of Nursing Studies 46 (2009) 230–238 cross-cultural assessment. School Psychology International 12 (1–2), 119–132. Bradley, C., 1994. Translation of questionnaires for use in different languages and cultures. In: Bradley, C. (Ed.), Handbook of Psychology: A Guide to Psychological Measurement in Diabetes Research and Management. Harwood Academic Publishers, Singapore, pp. 43–56. Chang, A.M., Chau, J.P., Holroyd, E., 1999. Translation of questionnaires and issues of equivalence. Journal of Advanced Nursing 29 (2), 316–322. Clark, D.O., 1996. Age, socioeconomic status, and exercise selfefficacy. Gerontologist 36 (2), 157–164. Conn, V.S., 1998. Older adults and exercise: path analysis of selfefficacy related constructs. Nursing Research 47 (3), 180–189. Cronbach, L.J., 1990. Essentials of Psychological Testing. Harper & Row, New York. Eysenck, H.J., 1978. Expectations as causal elements in behavioural change. Advances in Behaviour Research and Therapy 1, 171– 175. Garcia, A.W., King, A.C., 1991. Predicting long-term adherence to aerobic exercise: a comparison of two models. Journal of Sport and Exercise Psychology 13 (4), 394–410. Gecas, V., 1989. The social psychology of self-efficacy. Annual Review of Sociology 15, 291–316. Gortner, S.R., Jenkins, L.S., 1990. Self-efficacy and activity level following cardiac surgery. Journal of Advanced Nursing 15 (10), 1132–1138. Grembowski, D., Patrick, D., Diehr, P., Durham, M., Beresford, S., Kay, E., Hecht, J., 1993. Self-efficacy and health behavior among older adults. Journal of Health & Social Behavior 34 (2), 89–104. Hawkins, S.A., Cockburn, M.G., Hamilton, A.S., Mack, T.M., 2004. An estimate of physical activity prevalence in a large populationbased cohort. Medicine & Science in Sports & Exercise 36 (2), 253–260. Health Bureau, 2002. Survey of Exercise Behavior. Health Bureau, Department of Health, Taiwan, Taipei. Health Education Authority UK, 1999. Physical Activity in Later Life. Health Education Authority, London, p. 96. Jagodzinski, W., Kuhnel, S.M., 1987. Estimation of reliability and stability in single-indicator multiple-wave models. Sociological Methods and Research 15 (3), 219–258. Judge, T.A., Erez, A., Bono, J.E., Thoresen, C.J., 2002. Are measures of self-esteem, neuroticism, locus of control, and generalized self-efficacy indicators of a common core construct? Journal of Personality & Social Psychology 83 (3), 693–710. Lee, L.L., Arthur, A., Avis, M., 2007a. Evaluating a communitybased walking intervention for hypertensive older people in Taiwan: a randomized controlled trial. Preventive Medicine 44 (2), 160–166. Lee, L.L., Avis, M., Arthur, A., 2007b. The role of self-efficacy in older people’s decisions to initiate and maintain regular walking as exercise—findings from a qualitative study. Preventive Medicine 45 (1), 62–65. Loehlin, J.C., 2004. Latent Variable Models: An Introduction to Factor, Path, and Structural Equation Analysis. Lawrence Erlbaum Associates, Mahwah, NJ. Long, J.D., Stevens, K.R., 2004. Using technology to promote selfefficacy for healthy eating in adolescents. Journal of Nursing Scholarship 36 (2), 134–139.
237
Mason-Hawkes, J., Holm, K., 1993. Gender differences in exercise determinants. Nursing Research 42 (3), 166–172. McAuley, E., 1992. The role of efficacy cognitions in the prediction of exercise behavior in middle-aged adults. Journal of Behavioral Medicine 15 (1), 65–88. McAuley, E., Courneya, K.S., 1992. Self-efficacy relationships with affective and exertion responses to exercise. Journal of Applied Social Psychology 22 (4), 312–326. Miller, C.A., 1999. Nursing Care of Older Adults: Theory and Practice. Lippincott Williams & Wilkins, Philadelphia. Muris, P., 2001. A brief questionnaire for measuring self-efficacy in youths. Journal of Psychopathology and Behavioral Assessment 23 (3), 145–149. Nunnally, J., Bernstein, I., 1994. Psychometric Theory. McGrawHill, New York. Resnick, B., 1998a. Efficacy beliefs in geriatric rehabilitation. Journal of Gerontological Nursing 24 (7), 34–44. Resnick, B., 2002. Geriatric rehabilitation: the influence of efficacy beliefs and motivation. Rehabilitation Nursing 27 (4), 152–159. Resnick, B., 1999. Reliability and validity testing of the self-efficacy for functional activities scale. Journal of Nursing Measurement 7 (1), 5–20. Resnick, B., Jenkins, L.S., 2000. Testing the reliability and validity of the self-efficacy for exercise scale. Nursing Research 49 (3), 154–159. Resnick, B., Palmer, M.H., Jenkins, L.S., Spellbring, A.M., 2000. Path analysis of efficacy expectations and exercise behavior in older adults. Journal of Advanced Nursing 31 (6), 1309–1315. Roach, J.B., Yadrick, M.K., Johnson, J.T., Boudreaux, L.J., Forsythe 3rd, W.A., Billon, W., 2003. Using self-efficacy to predict weight loss among young adults. Journal of the American Dietetic Association 103 (10), 1357–1359. Rodin, J., 1989. Sense of control: potentials for intervention. Annals of the American Academy of Political & Social Science 503 (1), 29–42. Ronen, G.M., Streiner, D.L., Rosenbaum, P., Canadian Pediatric Epilepsy Network, 2003. Health-related quality of life in children with epilepsy: development and validation of self-report and parent proxy measures. Epilepsia 44 (4), 598–612. Sallis, J.F., Criqui, M.H., Kashani, I.A., Rupp, J.W., Calfas, K.J., Langer, R.D., Nader, P.R., Ross Jr., J., 1990. A program for health behavior change in a preventive cardiology center. American Journal of Preventive Medicine 6 (Suppl. 2), 43–50. Scholz, U., Gutierrez-Dona, B., Sud, S., Schwarzer, R., 2002. Is perceived self-efficacy a universal construct? Psychometric findings from 25 countries. European Journal of Psychological Assessment 18 (3), 242–251. Schunk, D.H., Carbonari, J.P., 1984. Self-efficacy models. In: Matarazzo, J.D., weiss, S.M., Herd, J.A., Miller, N.E., Weiss, S.M. (Eds.), Behavioural Health: A Handbook of Health Enhancement and Disease Prevention. Wiley, New York, pp. 230–247. Shellman, J., Shellman, J., 2006. Development and psychometric evaluation of the eldercare cultural self-efficacy scale. International Journal of Nursing Education Scholarship 3 (Article 9). Sims, J., Smith, F., Duffy, A., Hilton, S., 1999. The vagaries of selfreports of physical activity: a problem revisited and addressed in a study of exercise promotion in the over 65s in general practice. Family Practice 16 (2), 152–157.
238
L.-L. Lee et al. / International Journal of Nursing Studies 46 (2009) 230–238
Stutts, W.C., 2002. Physical activity determinants in adults. Perceived benefits, barriers, and self efficacy. American Association of Occupational Health Nurses (AAOHN) Journal 50 (11), 499–507. The Department of Statistics Ministry of the Interior ROC, 2007. Percentage of Older People Over 65 in Total Population Among
Major Countries. The Department of Statistics Ministry of the Interior ROC, Taipei. Tipton, R.M., 1984. The measurement of generalized self-efficacy: a study of construct validity. Journal of Personality Assessment, Lawrence Erlbaum Associates 545–548.