Harmonizing Depression Measures Across Studies: a Tutorial for Data Harmonization

Zhao, Xin; Coxe, Stefany; Sibley, Margaret H.; Zulauf, Courtney; Pettit, Jeremy W.

doi:10.1007/s11121-022-01381-5

Cited by 6 publications

(4 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Data harmonisation processes between different surveys, or between historical data from the same survey, would inform public and global mental health decision-making. We suggest reviewing the tutorial by Zhao et al as an example of data harmonisation with complex sampling 35…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Mental health data available in representative surveys conducted in Latin America and the Caribbean countries: a scoping review

Ramírez-Bontá,

Vásquez-Vílchez,

Cabrera-Alva

et al. 2023

BMJ Open

View full text Add to dashboard Cite

BackgroundMental health data from Latin America and the Caribbean countries (LACC) national and international surveys are essential for public health surveillance. This review aimed to identify and describe available mental health survey data in LACC, providing access details for researchers.MethodsOur study was a scoping review. The search for available mental health survey data was conducted in PubMed and through grey literature searches, and the search dates were between 26 August 2021 and 15 October 2021. Included survey data were/had (1) nationally representative, (2) the latest version available from 2012 onward, (3) collected in at least one LACC and (4) at least one mental health variable or related factor. We accepted all written languages, including Spanish and English.ResultsA total of 56 national and 13 international surveys were included, with data available on 95 mental health variables classified into 10 categories. Most national surveys were performed in upper-middle-income countries. Variables categorised as ‘Substance use’ and ‘Violence’ were the most frequent. Mexico and Colombia had the highest production in both the national and international surveys. The main target population was the adult population. However, there are several mental health topics and LACC yet unsurveyed.ConclusionWe identified a total of 69 representative surveys from LACCs since 2012. We categorised the available data on mental health variables into 10 categories, and provided technical details to facilitate the future selection and use of these surveys.

show abstract

Section: Discussionmentioning

confidence: 99%

“…We suggest reviewing the tutorial by Zhao et al as an example of data harmonisation with complex sampling. 35 Open access…”

Section: Recommendationsmentioning

confidence: 99%

Mental health data available in representative surveys conducted in Latin America and the Caribbean countries: a scoping review

Ramírez-Bontá,

Vásquez-Vílchez,

Cabrera-Alva

et al. 2023

BMJ Open

View full text Add to dashboard Cite

show abstract

“…However, there seems to be a conception, among both applied researchers and methodologists, that one can improve upon sum scores by computing factor scores from a partial invariance model, as the factor scores will be free of systematic measurement bias and can be validly compared across groups. For example, Curran and Hussong (2009) recommended researchers assess measurement invariance across samples and then calculate person scores “by using one of several available factor score estimates in the factor model” (p. 97; see also Bauer & Curran, 2016; Curran et al, 2016, 2014; Davoudzadeh et al, 2020); subsequent applied research generally follows such a practice of using factor scores after invariance testing (e.g., Luningham et al, 2019; MacDonald & Park, 2022; Zhao et al, 2022). McNeish (2022), on discussing a potential benefit of using factor scores over sum scores, suggested that “it would not make sense to compare sum scores across populations” when invariance is violated, but “[F]actor scoring can address some of these issues” by “allowing partial measurement invariance” (p. 4281).…”

Section: Example 1: When Invariance Holdsmentioning

confidence: 99%

Are factor scores measurement invariant?

Lai,

Tse

2024

Psychological Methods

View full text Add to dashboard Cite

“…However, there seems to be a conception, among both applied researchers and methodologists, that if one uses factor scores from a partial invariance model, the factor scores will be free of systematic measurement bias and can be validly compared across groups. For example, when Curran and Hussong (2009) proposed the integrative data analysis framework for harmonizing measures from different samples, they recommend researchers to assess measurement invariance across samples, and calculate person scores "by using one of several available factor score estimates in the factor model" (p. 97; see also Curran et al, 2014Curran et al, , 2016Davoudzadeh et al, 2020); subsequent applied research generally follows such a practice of using factor scores after invariance testing (e.g., Luningham et al, 2019;MacDonald & Park, 2022;Zhao et al, 2022). McNeish (2022), on discussing a potential benefit of using factor scores over sum scores, suggested that "it would not make sense to compare sum scores across populations" when invariance is violated, but " [F]actor scoring can address some of these issues" by "allowing partial measurement invariance."…”

Section: Are Factor Scores Measurement Invariant?mentioning

confidence: 99%

Are Factor Scores Measurement Invariant?

Lai,

Tse

2023

Preprint

View full text Add to dashboard Cite

There has been increased interest in practical methods for integrative analysis of data from multiple studies or samples, and using factor scores to represent constructs has become a popular and practical alternative to latent variable models with all individual items. Although researchers are aware that scores representing the same construct should be on a similar metric across samples---namely they should be measurement invariant---for integrative data analysis, the methodological literature is unclear whether factor scores would satisfy such a requirement. In this note, we show that even when researchers successfully calibrate the latent factors to the same metric across samples, factor scores---which are estimates of the latent factors but not the factors themselves---may not be measurement invariant. Specifically, we prove that factor scores computed based on the popular regression method are generally not measurement invariant. Surprisingly, such scores can be noninvariant even when the items are invariant. We also demonstrate that our conclusions generalize to similar shrinkage scores in item response models for discrete items, namely the expected a posteriori scores and the maximum a posteriori scores. Researchers should be cautious in directly using factor scores for cross-sample analyses, even when such scores are obtained from measurement models that account for noninvariance.

show abstract

Harmonizing Depression Measures Across Studies: a Tutorial for Data Harmonization

Cited by 6 publications

References 42 publications

Mental health data available in representative surveys conducted in Latin America and the Caribbean countries: a scoping review

Mental health data available in representative surveys conducted in Latin America and the Caribbean countries: a scoping review

Are factor scores measurement invariant?

Are Factor Scores Measurement Invariant?

Contact Info

Product

Resources

About