Anxiety about aging scale: factorial invariance in mexican women aged 40–90

Published: 26 September 2026| Version 1 | DOI: 10.17632/jc6ssjv4br.1
Contributors:
,
,
,
,
,
,
,

Description

This study examined the factorial structure, internal consistency, and age invariance of Lasher and Faulkender's Anxiety about Aging Scale. The dataset comprised 23 variables, with participant anonymity maintained through a numerical identifier (Variable 1). Variables 2-21 (AAE ) captured responses to questionnaire items on a 0-10 Likert scale, where higher scores indicated greater anxiety about aging. Variable 22 recorded participant age, and Variable 23 (CAT_AGE) categorized participants into adult (40-59 years) and older adult (60-90 years) groups. Confirmatory factor analysis evaluated two measurement models. Model 1, representing the original scale structure, demonstrated acceptable fit. Model 2, refined by removing items with factor loadings below 0.70 or conceptual issues, exhibited a significantly improved fit. This model revealed a robust four-factor structure: Fear of older people, Psychological concerns, Physical appearance, and Fear of loss, representing key dimensions of anxiety about aging. These findings suggest that the reduced model is psychometrically sound and that anxiety about aging decreases with age, highlighting the potential need for age-specific interventions. Overall, the results support the factorial validity and internal consistency of the Anxiety about Aging Scale, confirming its theoretical relevance and psychometric utility for assessing anxiety about aging in Mexican women across different adult age groups.

Files

Steps to reproduce

The study used a convenience sample of 805 Mexican women aged 40-90 (454 adults, 351 older adults) who completed the Lasher and Faulkender Anxiety about Aging Questionnaire. The Spanish version adapted by Fernández-Jiménez et al. was used, with three modifications for the Mexican context and sample characteristics: (1) an expanded 11-point response scale (0-10) to increase response sensitivity, categorized as Strongly disagree (0), Disagree (1-3), Neither agree nor disagree (4-6), Agree (7-9), and Strongly agree (10); (2) cultural and linguistic adjustments to ensure relevance for the Mexican population and age range; and (3) digital administration for direct data entry, eliminating manual coding and improving data processing.

Institutions

Categories

Anxiety, Aging, Structural Equation Modeling, Factor Analysis, Psychometrics

Licence