The write-up of the report is in the PDF file in the OSF storage repo. This document also refers to the data processing and analysis scripts that were used. The scripts and processed data are available through the github storage link. Below is a step-by-step guide to reproducing the analysis workflow. 1. Download the repository. 2. Copy the "production/audio" folder from the main project osf (https://osf.io/5agn9/?view_only=6fec040a9589499bb9998240bdab12a1) into "data" in the downloaded repository 3. If you wish to regenerate the alignments: - edit the paths in the praat script scripts/edit_textgrids_and_audio.praat - run this script - follow the instructions in alignments/alignment_notes - edit the paths in the praat script scripts/merge_mfa_and_full_textgrids.praat - run this script 4. If you wish to regenerate the duration measurements (this relies on steps 1, 2 and 3): - edit the paths in the praat script scripts/get_durations.praat - the duration measurements are saved in data/all_durations.tsv 5. If you wish to go through the data wrangling process: - work through many_speech_data_wrangling.Rmd - the part where the R script opens Praat to manually check the textgrids may require editing of the paths in the R script (but not in the Praat script); this may not run straightforwardly on Windows / Linux; also, this requires step 3 to have been run - the final file is saved in processed_data/NF_durations.csv 6. If you wish to work through the statistical analysis, run the script many_speech_analysis.Rmd