See also `README.md` in the [gitlab repo][1]. Contact: [Verena R. Sommer][2] ## Summary ## This projects provides code for the main analyses and figures reported in ["Neural pattern similarity differentially relates to memory performance in younger and older adults"][3] by V.R. Sommer, Y. Fandakova, T. Grandy, Y.L. Shing, M. Werkle-Bergner, and M.C. Sander (2019) *Journal of Neuroscience* ([preprint](https://doi.org/10.1101/528620)). Also additional control analyses are provided. To be able to reproduce the main results, averaged behavioral and EEG similarity data are made available. Since the statistical tests include random permutations which can result in slight variations in the statistical outputs, files with the original statistical results are also provided. ## Details ## The main analysis includes the between-item pattern similarity of EEG time-frequency data at each electrode and for each subject (`GS1_GlobalSimilarity.m`). These analyses were run on a high-performance computer cluster. The resulting individual channel x time x time similarity data for each memory quality are provided. For example, `YA_GAGS_half_11.mat` contains the similarity matrices of the high memory quality items for all young adults (YA). `11` denotes high, `01` medium, and `00` low memory quality. The individual single-trial time-frequency EEG data are not public. The provided **scripts** include analysis of: 1. Global EEG pattern similarity: `GS...m` 2. Behavior, i.e. memory performance, strategy use, imagery rating: `Beh...m` and all custom and third-party dependencies including (modified) raincloud plots (by Micah Allen, Davide Poggiali, Kirstie Whitaker, Tom Rhys Marshall and Rogier Kievit; https://github.com/RainCloudPlots/RainCloudPlots) and Colorbrewer (by Cynthia Brewer; https://www.mathworks.com/matlabcentral/fileexchange/34087-cbrewer-colorbrewer-schemes-for-matlab), except the Fieldrip Toolbox which is available under http://www.fieldtriptoolbox.org/download/. The provided **data** include: 1. Individual global similarity matrices (subject x channel x time x time, whereby one time-dimension is labeled as frequency in order to use standard Fieldtrip functions): `YA_GAGS...mat` and `OA_GAGS...mat` 2. Averaged individual memory performance, i.e. number of items per memory quality: `BehDataMerlinAll.mat` 3. Output from cluster-based random permutation statistics: e.g., `YA_11_vs_00.mat` To reproduce the **plots** presented in the paper, run: 1. Figure 4: `Beh1_memory_performance.m` 2. Figure 5A and B: `GS4b_plot_global_similarity.m` 3. Figure 5C and D: ` GS5c_cluster_plot.m` 4. Figure 5E: `GS7_individual_regr_z_cluster.m` [1]: https://gitlab.com/verysummer/merlin.-age-differences-in-neural-similarity-and-memory [2]: https://www.mpib-berlin.mpg.de/en/staff/verena-r-sommer [3]: https://www.jneurosci.org/content/39/41/8089