Integrating region- and network-level contributions to episodic recollection using multilevel structural equation modeling

Repository Home of the MemoLab’s manuscript:

Kurkela, K., Cooper, R., Ryu, E., & Ritchey, M. (submitted). Integrating region- and network-level contributions to episodic recollection using multilevel structural equation modeling.

This repository tries to follow the tidyverse’s style guide and BIDS formatting

Directories

intermediate/ = contains data written in between analysis steps
mplus/= mplus specific files

Analysis Steps

Step 1: Single Trial Estimates

• Estimate neural activation at each retrieval trial using Mumoford et al. 2012’s multi-model method.
• 01_SingleTrialEstimates/A_condition_retrieval_regressors_RSA.m – generates multiple conditions *.mat files of the retrieval trial. See SPM12 manual.
• 01_SingleTrialEstimates/A_nuisance_retrieval_regressors_RSA.m – generates multiple regressors *.mat files of nuisance variables. See SPM12 manual, manuscript for more information.
• 01_SingleTrialEstimates/B_first_level_RSA_singleTrial.m – specifies a SPM.mat file. This model is NOT estimated, just specified. See SPM12 manual.
• 01_SingleTrialEstimates/C_generate_RSA_singleTrial.m – Takes the model specification SPM.mat files from the previous step and runs a Mumford et al. 2012 mult-model single trial estimate analysis

Step 2: Extract ROI Data

• Extract single trial estimates (SPM_T values from a multi-model single trial estimate analysis, see Mumford et al. 2012; Maureen Ritchey’s Generate SPM Single Trial) from PM Network ROIs (see Cooper, Kurkela, Davis, & Ritchey 2021; Publically Available ROIs)
• Assumes ROIs are stored in a local directory: rois/
• Assumes single-trial estimates are stored in a local directory: st_estimates/
• Assumes SPM12 is located in spm12/
• See: 02_ExtractROIData/Extract_ROI_data.m – Extracts single trial estimates from select ROIs and writes them as a csv file
• See: 02_ExtractROIData/Reslice_ROIs.m – reslices ROIs to be in the same space as the single trial estimates.
• See: intermediate/02_Extracted_ROI_data_1s.csv

Step 3: Tidy Data

• Take Extracted Single Trial Estimates and appends behavioral data ‘tidying’ the data along the way
• Assumes behavioral data are stored in a local directory: orbit-data/
• See: 03_Tidy/tidy.R
• See: intermediate/03_tidy_roi_data_1s.dat

Step 4: MPLUS Modeling

• Run a series of SEM models in MPLUS. See README in mplus/
• See: 04_RunMPLUSModels/runMPLUESmodels.R

Step 5: Create Publication Tables

• After running the MPLUS models, these scripts automate extracting the results from the MPLUS output files (*.out) and writing the results to be formatted for publication.
• See: 05_CreatePublicationTables/CreateTable1.R
• See: 05_CreatePublicationTables/CreatePublicationTables.R

References

Mumford, J. A., Turner, B. O., Ashby, F. G., & Poldrack, R. A. (2012). Deconvolving BOLD activation in event-related designs for multivoxel pattern classification analyses. NeuroImage, 59(3), 2636–2643. https://doi.org/10.1016/j.neuroimage.2011.08.076.

Cooper, R. A., Kurkela, K. A., Davis, S. W., & Ritchey, M. (2021). Mapping the organization and dynamics of the posterior medial network during movie watching. NeuroImage, 236, 118075. https://doi.org/10.1016/j.neuroimage.2021.118075