Author: Trevor KM Day, day00096@umn.edu
This repository is for analysis of MacArthur-Bates Communicative Development Inventory (American English) forms. The abbreviation MCDI is typically used throughout, rather than MB-CDI, as a quirk of how the form was referred to in internal Elison Lab materials.
This code both analyzes data from:
- Stanford Wordbank (Frank et al., 2016)
- University of Minnesota (UMN) Baby Connectome Project (BCP; Howell et al., 2018)
- UMN Phenoscreening (PHE)
All three projects use both Words & Gestures (WG) and Words & Sentences (WS).
The analysis is built up in multiple stages:
- Analysis of Wordbank WG and WS data individually.
- Joint Wordbank WG/WS analysis. Through steps 1-2, various functions for processing of Wordbank-format data are developed (
code/Wordbank/wordbank-functions.R) that are shared across many scripts. - Conversion of BCP/PHE data to Wordbank format.
- Use of Wordbank analysis of BCP/PHE data (
code/BCP/). - Growth curve modeling of BCP data (
code/growth-curving/).
wordbank-csv2rds.R converts WG/WS data to RDS to speed read-in and also extract demographic information and saves that to another RDS for easy manipulation.
- Frank, M. C., Braginsky, M., Yurovsky, D., & Marchman, V. A. (2016). Wordbank: An open repository for developmental vocabulary data. Journal of Child Language. doi: 10.1017/S0305000916000209.
- Howell, B. R., ... Elison, J. T. (2018). The UNC/UMN Baby Connectome Project (BCP): An overview of the study design and protocol development. NeuroImage. 10.1016/j.neuroimage.2018.03.049