What does this research mean for the field?

Extending the measurement error autoregressive model with a Diggle-Kenward selection process allows for accurate recovery of within-person reliability estimates in intensive longitudinal data, even when large amounts of data are missing not at random. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

To determine how missing observations distort within-person reliability estimates in intensive longitudinal data and evaluate an extended measurement error autoregressive model tailored for missing not at random data.

August 23, 2026Psychological Methods

Within-person reliability of composite scores or single-item responses with missing intensive longitudinal data.

View Full Paper

Ask AI

Bookmark

Authors

DMDaniel McNeish

Discussion

Member takes

Overview

Simulation study reveals that extending autoregressive models with selection processes recovers within-person reliability despite missing data, highlighting robust psychometric evaluation methods.

Key Points

To determine how missing observations distort within-person reliability estimates in intensive longitudinal data and evaluate an extended measurement error autoregressive model tailored for missing not at random data.
Extended the measurement error autoregressive model with a Diggle-Kenward selection process to account for missing not at random (MNAR) data mechanisms.
Conducted Monte Carlo simulations evaluating parameter recovery and within-person reliability estimation accuracy under typical missingness rates (30%-40%) across varying missing data mechanisms.
Standard psychometric approaches exhibited deteriorating reliability estimation accuracy as missing data rates increased.
The proposed selection-model extension accurately recovered true within-person reliability parameters even when substantial MNAR data were present under evaluated conditions.

Cite This Study

Daniel McNeish (2026) studied this question.

synapsesocial.com/papers/6a8aad667677a3411444591f https://doi.org/10.1037/met0000865

View Full Paper

Ask AI

Bookmark

August 23, 2026Psychological Methods

Within-person reliability of composite scores or single-item responses with missing intensive longitudinal data.

View Full Paper

Ask AI

Bookmark

Authors

DMDaniel McNeish

Discussion

Member takes

Overview

Simulation study reveals that extending autoregressive models with selection processes recovers within-person reliability despite missing data, highlighting robust psychometric evaluation methods.

Key Points

To determine how missing observations distort within-person reliability estimates in intensive longitudinal data and evaluate an extended measurement error autoregressive model tailored for missing not at random data.
Extended the measurement error autoregressive model with a Diggle-Kenward selection process to account for missing not at random (MNAR) data mechanisms.
Conducted Monte Carlo simulations evaluating parameter recovery and within-person reliability estimation accuracy under typical missingness rates (30%-40%) across varying missing data mechanisms.
Standard psychometric approaches exhibited deteriorating reliability estimation accuracy as missing data rates increased.
The proposed selection-model extension accurately recovered true within-person reliability parameters even when substantial MNAR data were present under evaluated conditions.

Cite This Study

Daniel McNeish (2026) studied this question.

synapsesocial.com/papers/6a8aad667677a3411444591f https://doi.org/10.1037/met0000865

View Full Paper

Ask AI

Bookmark