NB. _.exp file for subject 085DAC (word generation) was added August 2023. This subject did the Word Generation task but not the Pyramids and Palm Trees task, and had been omitted from the main GAM analysis. Data are included here as needed for subsequent project https://osf.io/7v9jh_ **Raw Data** **Dataset 1.** Data from this project is taken from Bruckert (2016) chapter 4, using a word generation task during fTCD recording. The exp format has some header information about data acquisition, then 8 columns of data: 1. Time sample commenced 2. Sample number 3. Cerebral blood flow velocity (CBFV, in cm/s) for left fTCD probe 4. CBFV for right fTCD probe 5. Heart beat averaged CBFV for left probe (DO NOT USE) 6. Heart beat averaged CBFV for right probe (DO NOT USE) 7. Stimulus onset channel 8. Another stimulus onset channel (DO NOT USE) We normally extract columns 2, 3, 4 and 7 for processing. **Task** The word generation task is reported in Bruckert (2016). In brief, there are 23 trials, each of which last 25 seconds with 25 seconds rest between them. The trials proceed as follows: - Each trial starts with a 'CLEAR MIND' visual cue, which is the trigger for the onset of the trial. This is presented for 5 seconds. - Next, a letter appears on screen for 2.5 seconds, and the participant is required to start generating words beginning with that letter. - A blank screen follows for 12.5 seconds (i.e. 15 seconds of covert generation time in total) - Next, a 'SAY WORDS' cue is presented for 5 seconds, during which the participant overtly reports the words they thought of. - Finally, 'RELAX' is presented on screen for 25 seconds prior to the next trial. Some values from the original processed data are saved with demographics in WordGen_results.csv. A data dictionary is provided for this file. The analysis also uses 2 columns from Chapter5_fMRIdata.csv, which provides fMRI laterality indices on the Word Generation task for a subset of participants. The columns are: fMRI_new_wg_MCA: conventional LI fMRI_diff_wg_MCA: different in N voxels L and R (This file contains a lot of other extraneous material, not used here). **Dataset 2** Data from Woodhead et al (2021). Sentence Generation task. The original computed LIs are provided in A2_SG_LI. This just has the mean LI for sessions 1 and 2 with the subject ID. **Dataset 3** Data from Woodhead et al 2018. The WSLG_AnalysisData.csv is summary processed data from the original paper. There is a data dictionary provided. **Notes 2nd Nov 2022** Minor corrections to figures 3-5. Polarity was reversed for LI estimates in figs 4-5, and the CI in figures 3-5 had been computed for 90% CI rather than 95% CI. Script was modified to correct these points.