## Tutorial Find a detailed description and step-by-step tutorial of how to run spectral pattern similarity analysis in our preprint: https://psyarxiv.com/87hpb/ ## Code Get all necessary scripts from the [gitlab repository][1] linked under Files. (Except FieldTrip; download here: https://www.fieldtriptoolbox.org/download/). **Run the complete pipeline with `config_and_run.m`** (see details below). ## Data Download sample dataset under Files (figshare) including time-frequency representations (TFRs) of 10 children and 10 adults as well as all interim results to run the analysis pipeline. Children's IDs: 2102, 2104, 2110, 2112, 2113, 2114, 2201, 2203, 2205, 2206 Adults' IDs: 3113, 3115, 3119, 3122, 3132, 3208, 3213, 3214, 3218, 3219 ### Included data: https://osf.io/jbrsa/files/ 1. Single-subject single-trial TFRs of 10 children and 10 adults in folder "Time-frequency EEG data": `<id>_Enc_reduced_tfr_2_125Hz.mat` (trial × electrode × frequency × time) 2. Single-subject single-trial similarity matrices (within-item and within-category similarity) in folder "Pattern similarity data": `<id>_within-item_sim_ST.mat` (trial × electrode × time × time) `<id>_within-cat_sim_ST.mat` (trial × electrode × time × time) 3. Grand average similarity matrices (within-item and within-category similarity) in folder "Grand average pattern similarity": `<group>_within-item_sim_GA.mat` (subject × electrode × time × time) `<group>_within-cat_sim_GA.mat` (subject × electrode × time × time) 4. Individual mean similarities (within-item and within-category) and item specificities in folder "Grand average pattern similarity data": `CH+YA_mean_within-item_sim_in_cluster.mat` (subject × 1) `CH+YA_mean_within-cat_sim_in_cluster.mat` (subject × 1) `CH+YA_mean_item_specificity_in_cluster.mat` (subject × 1) 5. Results from statistics (first level, second level, third level) in folder "Statistics": `<group>_within-item_vs_within-cat_1st_level_stat.mat` (subject × electrode × time × time) `<group>_within-item_vs_within-cat_2nd_level_cluster_stat.mat` (electrode × time × time) `CH_vs_YA_item_spec_in_cluster_3rd_level.mat` 6. Individual mean item memory performance in folder "Memory performance": `CH+YA_mean_item_memory.mat` (subject × 1) 7. Electrode layout in folder "Time-frequency EEG data": `elec.mat` ### Analysis steps: For detailed description of the `config` input struct, see code documentation in `config_and_run.m`. **Step 1**: Run RSA for each subject `step1_rsa_get_sim_matrices(config)` **Step 2**: Compute grand average `step2_rsa_group_ga(config)` **Step 3**: Plot averaged similarity matrix `step3_plot_sim_matrices(config)` **Step 4**: Test within-item versus between-item similarity 1. Test within-item similarity against between-item (within-category) similarity for each subject (first level) `stat = step4a_sim_comparison_1st_level(config);` 2. Test first-level t-values (stat; see above) against 0 at the group level (second level) `stat2 = step4b_sim_comparison_2nd_level(config);` 3. Extract similarity values from identified clusters (stat2; see above) and contrast between age groups `step4c_sim_comparison_3rd_level(config)` **Step 5**: Plot resulting clusters from step 4 (time-time and topographies) `step5_plot_clusters(config)` **Step 6**: Plot similarity comparision `step6_plot_sim_comparison(config)` **Step 7**: Correlate with memory performance + plot `step7_correlation_with_behavior(config)` [1]: https://gitlab.com/verysummer/dcn-tutorial/