A Scale for Older adults' decisional balance regarding physical ACTIVity (SO-ACTIV): development and validation in a French sample.

The initial item pool was reduced from 55 items to 35 items through expert review, and confirmatory factor analyses supported a final 26-item instrument.

• An initial pool of 55 items was developed and submitted to expert review.
• Expert review retained 35 items divided into two overarching factors (facilitators and barriers) and four domains (intrapersonal, interpersonal, environmental, organizational).
• Confirmatory factor analyses further reduced the instrument to a final 26-item scale.
• The final structure is a second-order hierarchical model with two upper-level factors and four specific domains under each factor.
• The second-order hierarchical model yielded the best fit indices compared to alternative models tested.

The SO-ACTIV demonstrated acceptable to excellent internal consistency across domains and factors.

• Internal consistency ranged from acceptable to excellent across the four domains and two upper-level factors.
• Test-retest reliability over a 2-week interval was 'mainly demonstrated' across the scale.
• The online survey was completed by 452 French-speaking older adults.
• Both facilitator and barrier subscales showed adequate reliability properties.

Convergent validity of the SO-ACTIV was supported through correlations with motivation toward health-oriented physical activity and self-reported physical activity level.

• Facilitators correlated positively with self-determined forms of motivation and with physical activity level.
• Facilitators correlated negatively with amotivation.
• Barriers showed the opposite pattern: negative correlations with self-determined motivation and physical activity, and positive correlations with amotivation.
• Construct validity was assessed through correlations with motivation toward health-oriented physical activity and a self-reported physical activity level score.

Concurrent validity of the SO-ACTIV was supported by correlations between its subscales and an existing decisional balance measure for exercise in adults.

• Facilitators on the SO-ACTIV correlated positively with the 'pros' subscale of an existing decisional balance measure for exercise in adults.
• Barriers on the SO-ACTIV correlated positively with the 'cons' subscale of the existing decisional balance measure.
• This concurrent validity assessment used an established decisional-balance measure for exercise in adults as the comparison instrument.

Known-groups validity was supported by significant differences in SO-ACTIV scores across stages of change.

• Participants in the action and maintenance stages of change scored higher on facilitators compared with participants in earlier stages of change.
• Participants in the action and maintenance stages scored lower on barriers compared with participants in earlier stages of change.
• Known-groups validity was tested across all stages of change as defined by the transtheoretical model.

Existing decisional balance measures for physical activity were identified as having gaps that the SO-ACTIV was designed to address.

• Existing measures overlook age-related concerns specific to older adults.
• Existing measures lack theoretical grounding in models such as the transtheoretical model.
• Existing measures fail to reflect the multilevel socioecological structure (intrapersonal, interpersonal, environmental, organizational).
• Physical activity declines with age despite its established health benefits, motivating the need for an age-specific instrument.