Repo contains scripts and data necessary to replicate analyses in:
Cychosz, M., Munson, B., & Edwards, J. (accepted). Practice and experience predict coarticulation in child speech. Language Learning and Development. (https://psyarxiv.com/vwhtk/)
get_mfcc_coefs_nw.py - example script to calculate Mel spectral vectors for each consonant and vowel, CV sequence duration, and word duration for nonwords; generates nwr_mfcc_complete.csv for further processing
1a-get_repetition_scores.R - combines dataframes from two sets of acoustic measures (rwr_mfcc_complete.csv and nwr_mfcc_complete.csv); generates repetition accuracy scores on the basis of consonant manner, prosody, etc.
1b-get_repetition_scores_partII.R - finishes 1a-get_repetition_scores.R and generates dataframe with acoustic measures and repetition scores; generates repetition_scores_mfcc.csv
2-match_words.R - select repetitions that were 100% correct from repetition_scores_mfcc.csv and generates final_mfcc_scores.csv
3-map_vocab_scores.R - reads in final_mfcc_scores.csv and vocab_scores.csv and merges the dataframes; generates final_mfccs_w_vocab.csv
4-model_coartic.R - reads in final_mfccs_w_vocab.csv to process spectral vectors data and calculate Euclidean distance; modeling and figures for paper
coartic_experience_accepted.pdf - compiled paper
coartic_experience_accepted.tex - tex source of coartic_experience_accepted.pdf
phono_prbability.csv - phonotactic transition probabilities between CV sequence and first C of second syllable
rwr_mfcc_complete.csv - acoustic measures (spectral vectors, word and phone duration) for each C and V in the real words
nwr_mfcc_complete.csv - acoustic measures (spectral vectors, word and phone duration) for each C and V in the nonwords
repetition_scores_mfcc.csv - dataframe with acoustic measures and repetition scores from word repetition task
final_mfcc_scores.csv - dataframe with 100% correct repetitions and acoustic measures
vocab_scores.csv - dataframe with vocabulary scores, LENA measures, and (limited) participant demographic information
final_mfccs_w_vocab.csv - dataframe of 100% correct repetitons with acoustic measures, vocabulary scores, LENA measures, and demographic information