Development and Psychometric Testing of a New Instrument to Measure Factors Influencing Women's Breast Cancer Prevention Behaviors (Assists) Publisher Pubmed



Khazaeepool M¹ ; Majlessi F² ; Montazaeri A³ ; Pashaei T⁴ ; Gholami A^{5, 6} ; Ponnet K^{7, 8, 9} Show Affiliations
Source: BMC Women's Health Published:2016

Abstract

Background: Breast cancer preventive behaviors have an extreme effect on women's health. Despite the benefits of preventive behaviors regarding breast cancer, they have not been implemented as routine care for healthy women. To assess this health issue, a reliable and valid scale is needed. The aim of the present study is to develop and examine the psychometric properties of a new scale, called the ASSISTS, in order to identify factors that affect women's breast cancer prevention behaviors. Methods: A multi-phase instrument development method was performed to develop the questionnaire from February 2012 to September 2014. The item pool was generated based on secondary analyses of previous qualitative data. Then, content and face validity were applied to provide a pre-final version of the scale. The scale validation was conducted with a sample of women recruited from health centers affiliated with Tehran University of Medical Sciences. The construct validity (both exploratory and confirmatory), convergent validity, discriminate validity, internal consistency reliability and test-retest analysis of the questionnaire were tested. Results: Fifty-eight items were initially extracted from the secondary analysis of previous qualitative data. After content validity, this was reduced to 49 items. The exploratory factor analysis revealed seven factors (Attitude, supportive systems, self-efficacy, information seeking, stress management, stimulant and self-care) containing 33 items that jointly accounted for 60.62 % of the observed variance. The confirmatory factor analysis showed a model with appropriate fitness for the data. The Cronbach's alpha coefficient for the subscales ranged from 0.68 to 0.85, and the Intraclass Correlation Coefficient (ICC) ranged from 0.71 to 0.98; which is well above the acceptable thresholds. Conclusion: The findings showed that the designed questionnaire was a valid and reliable instrument for assessing factors affecting women's breast cancer prevention behaviors that can be used both in practice and in future studies. © 2016 Khazaee-Pool et al.