Limits...
Development and validation of the short-form Adolescent Health Promotion Scale.

Chen MY, Lai LJ, Chen HC, Gaete J - BMC Public Health (2014)

Bottom Line: The first subsample (calibration sample) was used to modify and shorten the factorial model while the second subsample (validation sample) was utilized to validate the result obtained from the first one.The results of the CFA supported a six-factor model and 21 items were retained in the AHP-SF with acceptable model fit.Primary health care providers and clinicians can use the AHP-SF to assess these behaviors and evaluate the outcome of health promotion programs in the adolescent population.

View Article: PubMed Central - PubMed

Affiliation: College of Nursing, Chang Gung University of Science and Technology, No, 2, Chiapu Rd, West Sec,, Putz City, Chiayi County 61363, Taiwan. meiyen@gw.cgust.edu.tw.

ABSTRACT

Background: Health-promoting lifestyle choices of adolescents are closely related to current and subsequent health status. However, parsimonious yet reliable and valid screening tools are scarce. The original 40-item adolescent health promotion (AHP) scale was developed by our research team and has been applied to measure adolescent health-promoting behaviors worldwide. The aim of our study was to examine the psychometric properties of a newly developed short-form version of the AHP (AHP-SF) including tests of its reliability and validity.

Methods: The study was conducted in nine middle and high schools in southern Taiwan. Participants were 814 adolescents randomly divided into two subgroups with equal size and homogeneity of baseline characteristics. The first subsample (calibration sample) was used to modify and shorten the factorial model while the second subsample (validation sample) was utilized to validate the result obtained from the first one. The psychometric testing of the AHP-SF included internal reliability of McDonald's omega and Cronbach's alpha, convergent validity, discriminant validity, and construct validity with confirmatory factor analysis (CFA).

Results: The results of the CFA supported a six-factor model and 21 items were retained in the AHP-SF with acceptable model fit. For the discriminant validity test, results indicated that adolescents with lower AHP-SF scores were more likely to be overweight or obese, skip breakfast, and spend more time watching TV and playing computer games. The AHP-SF also showed excellent internal consistency with a McDonald's omega of 0.904 (Cronbach's alpha 0.905) in the calibration group.

Conclusion: The current findings suggest that the AHP-SF is a valid and reliable instrument for the evaluation of adolescent health-promoting behaviors. Primary health care providers and clinicians can use the AHP-SF to assess these behaviors and evaluate the outcome of health promotion programs in the adolescent population.

Show MeSH

Related in: MedlinePlus

The standardized estimates of CFA model for calibration and validation sample. (The values of validation sample were in parenthesis).
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4216378&req=5

Fig1: The standardized estimates of CFA model for calibration and validation sample. (The values of validation sample were in parenthesis).

Mentions: As shown in Table 1, the initial 40-item CFA of the calibration sample revealed that half of the model fit indices failed to meet the established criteria, including GFI (0.81), AGFI (0.78), NFI (0.75), NNFI (0.82), and CFI (0.84). Three indices met the criteria, including PNFI (0.70), PCFI (0.78), and χ2/df (2.43). The SRMR (0.064) and RMSEA (0.059) were both between 0.05 and 0.08, which indicated an acceptable, fit. Generally, the initial 40-item model did not fit very well and thus needed modification. As a first step, 16 items were sequentially deleted (one by one) due to their low factor loadings (<0.50). However, after this 16-item deletion, the 24-item AHP short form still showed an inadequate relative model fit in terms of the NFI (0.88 which is less than the cutoff value of 0.90). In step two, we examined the modification indices of the 24 AHP items and found that three pairs of items appeared to measure the same three concepts (items 26 and 35, items 9 and 37, and items 20 and 40). After review by a panel of experts, items 35, 37, and 20 were deleted so that the final AHP-SF contained 21 items. For this set of items, all model fit indices met the criteria, suggesting a good model fit. Figure 1 shows the correlation matrices among the latent variables and factor loadings for the calibration and validation samples. As Figure 1 illustrates, all standardized factor loadings exceeded the threshold of 0.50, indicating an acceptable convergent validity of the AHP-SF[24]. In addition, McDonald’s omega for the AHP-SF was 0.904 (Cronbach’s alpha 0.905) indicating excellent internal consistency (data are not shown in the table).Table 1


Development and validation of the short-form Adolescent Health Promotion Scale.

Chen MY, Lai LJ, Chen HC, Gaete J - BMC Public Health (2014)

The standardized estimates of CFA model for calibration and validation sample. (The values of validation sample were in parenthesis).
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4216378&req=5

Fig1: The standardized estimates of CFA model for calibration and validation sample. (The values of validation sample were in parenthesis).
Mentions: As shown in Table 1, the initial 40-item CFA of the calibration sample revealed that half of the model fit indices failed to meet the established criteria, including GFI (0.81), AGFI (0.78), NFI (0.75), NNFI (0.82), and CFI (0.84). Three indices met the criteria, including PNFI (0.70), PCFI (0.78), and χ2/df (2.43). The SRMR (0.064) and RMSEA (0.059) were both between 0.05 and 0.08, which indicated an acceptable, fit. Generally, the initial 40-item model did not fit very well and thus needed modification. As a first step, 16 items were sequentially deleted (one by one) due to their low factor loadings (<0.50). However, after this 16-item deletion, the 24-item AHP short form still showed an inadequate relative model fit in terms of the NFI (0.88 which is less than the cutoff value of 0.90). In step two, we examined the modification indices of the 24 AHP items and found that three pairs of items appeared to measure the same three concepts (items 26 and 35, items 9 and 37, and items 20 and 40). After review by a panel of experts, items 35, 37, and 20 were deleted so that the final AHP-SF contained 21 items. For this set of items, all model fit indices met the criteria, suggesting a good model fit. Figure 1 shows the correlation matrices among the latent variables and factor loadings for the calibration and validation samples. As Figure 1 illustrates, all standardized factor loadings exceeded the threshold of 0.50, indicating an acceptable convergent validity of the AHP-SF[24]. In addition, McDonald’s omega for the AHP-SF was 0.904 (Cronbach’s alpha 0.905) indicating excellent internal consistency (data are not shown in the table).Table 1

Bottom Line: The first subsample (calibration sample) was used to modify and shorten the factorial model while the second subsample (validation sample) was utilized to validate the result obtained from the first one.The results of the CFA supported a six-factor model and 21 items were retained in the AHP-SF with acceptable model fit.Primary health care providers and clinicians can use the AHP-SF to assess these behaviors and evaluate the outcome of health promotion programs in the adolescent population.

View Article: PubMed Central - PubMed

Affiliation: College of Nursing, Chang Gung University of Science and Technology, No, 2, Chiapu Rd, West Sec,, Putz City, Chiayi County 61363, Taiwan. meiyen@gw.cgust.edu.tw.

ABSTRACT

Background: Health-promoting lifestyle choices of adolescents are closely related to current and subsequent health status. However, parsimonious yet reliable and valid screening tools are scarce. The original 40-item adolescent health promotion (AHP) scale was developed by our research team and has been applied to measure adolescent health-promoting behaviors worldwide. The aim of our study was to examine the psychometric properties of a newly developed short-form version of the AHP (AHP-SF) including tests of its reliability and validity.

Methods: The study was conducted in nine middle and high schools in southern Taiwan. Participants were 814 adolescents randomly divided into two subgroups with equal size and homogeneity of baseline characteristics. The first subsample (calibration sample) was used to modify and shorten the factorial model while the second subsample (validation sample) was utilized to validate the result obtained from the first one. The psychometric testing of the AHP-SF included internal reliability of McDonald's omega and Cronbach's alpha, convergent validity, discriminant validity, and construct validity with confirmatory factor analysis (CFA).

Results: The results of the CFA supported a six-factor model and 21 items were retained in the AHP-SF with acceptable model fit. For the discriminant validity test, results indicated that adolescents with lower AHP-SF scores were more likely to be overweight or obese, skip breakfast, and spend more time watching TV and playing computer games. The AHP-SF also showed excellent internal consistency with a McDonald's omega of 0.904 (Cronbach's alpha 0.905) in the calibration group.

Conclusion: The current findings suggest that the AHP-SF is a valid and reliable instrument for the evaluation of adolescent health-promoting behaviors. Primary health care providers and clinicians can use the AHP-SF to assess these behaviors and evaluate the outcome of health promotion programs in the adolescent population.

Show MeSH
Related in: MedlinePlus