INS NYC 2024 Program

Poster	Poster Session 11 Program Schedule 02/17/2024 10:45 am - 12:00 pm Room: Shubert Complex (Posters 1-60) Poster Session 11: Cultural Neuropsychology \| Education/Training \| Professional Practice Issues Final Abstract #17 Assessment of Measurement Invariance between Hispanic/Latino Spanish- and English- Test Takers in the NACC Carlos Araujo-Menendez, San Diego State University/University of California, San Diego Joint Doctoral Program in Clinical Psychology, San Diego, United States Rachel Membreno, San Diego State University, Department of Psychology, San Diego, United States Ruby Carpio, San Diego State University, Department of Psychology, San Diego, United States Shaun Goycoochea, San Diego State University, Department of Psychology, San Diego, United States Armando Lemus, San Diego State University, Department of Psychology, San Diego, United States Ariana Stickel, San Diego State University, Department of Psychology, San Diego, United States Category: Cross Cultural Neuropsychology/ Clinical Cultural Neuroscience Keyword 1: diversity Keyword 2: aging (normal) Keyword 3: neuropsychological assessment Objective: Non-English speakers are often misdiagnosed with cognitive impairments and disorders due to neuropsychological testing limitations. Misdiagnosis can have consequences on one’s ability to maintain a job, drive, and complete daily functions, in addition to having implications for health and treatment. Examining psychometric properties of tests within linguistically diverse populations may be the first step in addressing testing limitations. Previously researchers analyzed neuropsychological assessments in English and Spanish and found assessments in both languages measured the same constructs in a community-based sample of older adults in Washington Heights/Inwood Columbia Aging Project (WHICAP). The present study aimed to test measurement invariance in English- and Spanish-speaking Hispanic/Latino adults across the nation using data from the National Alzheimer’s Coordinating Center (NACC). Participants and Methods: This study included Hispanic/Latino adults (n=2,360), ages 18-104 (mean=69.41, SD=9.52) without dementia, from 40 sites. Language at testing (Spanish, English) and clinical status [cognitive normal (CN) or mild cognitive impairment (MCI)] was used to group participants (four groups total). Participants' first assessment was used for analyses. Tests included in the analyses were the Number Span Test forward and backwards (number of correct trials and longest span), Craft Story 21 immediate and delayed recall, Trial Making test part A and B, Animal and Vegetable verbal fluency, and the Multilingual Naming Test (MINT). First, among the CN English speakers, an exploratory factor analysis was run to identify the best-fitting factor model. Second, multiple group confirmatory factor analysis were performed using the best fitting model obtained from CN English speakers in the other samples to test for configural, metric, and scalar invariance across the groups. Results: The exploratory factor analysis in the CN English speaker group identified 5 factors which represented the following cognitive domains (attention, working memory, verbal episodic memory, processing speed, and language). Multiple group analyses yielded the following results: Configural invariance test showed acceptable fit (CFI=0.99, RMSEA=0.024, SRMR=0.041). Restricting the factor loadings to be equal across groups did not significantly change the fit of our model (CFI=0.991, RMSEA=0.025, SRMR=0.046), indicating good metric invariance. Restricting the factor loadings and intercepts resulted in a poorer overall fit (CFI=0.986, RMSEA=0.031, SRMR=0.051), indicating possible scalar invariance. Conclusions: We found mixed evidence that the neuropsychological assessments of Hispanic/Latino adults in the NACC dataset may not measure the same cognitive attributes in English and Spanish. This study addressed one aspect of possible psychometric limitations. However, future studies are needed to test other aspects of potential linguistic bias in neuropsychological assessments, including validity (e.g. predictive validity).

Poster

02/17/2024
10:45 am - 12:00 pm
Room: Shubert Complex (Posters 1-60)

Poster Session 11: Cultural Neuropsychology | Education/Training | Professional Practice Issues

Final Abstract #17

Assessment of Measurement Invariance between Hispanic/Latino Spanish- and English- Test Takers in the NACC

Carlos Araujo-Menendez, San Diego State University/University of California, San Diego Joint Doctoral Program in Clinical Psychology, San Diego, United States
Rachel Membreno, San Diego State University, Department of Psychology, San Diego, United States
Ruby Carpio, San Diego State University, Department of Psychology, San Diego, United States
Shaun Goycoochea, San Diego State University, Department of Psychology, San Diego, United States
Armando Lemus, San Diego State University, Department of Psychology, San Diego, United States
Ariana Stickel, San Diego State University, Department of Psychology, San Diego, United States

Category: Cross Cultural Neuropsychology/ Clinical Cultural Neuroscience

Keyword 1: diversity
Keyword 2: aging (normal)
Keyword 3: neuropsychological assessment

Objective:

Non-English speakers are often misdiagnosed with cognitive impairments and disorders due to neuropsychological testing limitations. Misdiagnosis can have consequences on one’s ability to maintain a job, drive, and complete daily functions, in addition to having implications for health and treatment. Examining psychometric properties of tests within linguistically diverse populations may be the first step in addressing testing limitations. Previously researchers analyzed neuropsychological assessments in English and Spanish and found assessments in both languages measured the same constructs in a community-based sample of older adults in Washington Heights/Inwood Columbia Aging Project (WHICAP). The present study aimed to test measurement invariance in English- and Spanish-speaking Hispanic/Latino adults across the nation using data from the National Alzheimer’s Coordinating Center (NACC).

Participants and Methods:

This study included Hispanic/Latino adults (n=2,360), ages 18-104 (mean=69.41, SD=9.52) without dementia, from 40 sites. Language at testing (Spanish, English) and clinical status [cognitive normal (CN) or mild cognitive impairment (MCI)] was used to group participants (four groups total). Participants' first assessment was used for analyses. Tests included in the analyses were the Number Span Test forward and backwards (number of correct trials and longest span), Craft Story 21 immediate and delayed recall, Trial Making test part A and B, Animal and Vegetable verbal fluency, and the Multilingual Naming Test (MINT). First, among the CN English speakers, an exploratory factor analysis was run to identify the best-fitting factor model. Second, multiple group confirmatory factor analysis were performed using the best fitting model obtained from CN English speakers in the other samples to test for configural, metric, and scalar invariance across the groups.

Results:

The exploratory factor analysis in the CN English speaker group identified 5 factors which represented the following cognitive domains (attention, working memory, verbal episodic memory, processing speed, and language). Multiple group analyses yielded the following results: Configural invariance test showed acceptable fit (CFI=0.99, RMSEA=0.024, SRMR=0.041). Restricting the factor loadings to be equal across groups did not significantly change the fit of our model (CFI=0.991, RMSEA=0.025, SRMR=0.046), indicating good metric invariance. Restricting the factor loadings and intercepts resulted in a poorer overall fit (CFI=0.986, RMSEA=0.031, SRMR=0.051), indicating possible scalar invariance.

Conclusions:

We found mixed evidence that the neuropsychological assessments of Hispanic/Latino adults in the NACC dataset may not measure the same cognitive attributes in English and Spanish. This study addressed one aspect of possible psychometric limitations. However, future studies are needed to test other aspects of potential linguistic bias in neuropsychological assessments, including validity (e.g. predictive validity).