A meta-analysis of infant speed of processing and vocabulary development

shutterstock 543939718

Though infant language development has been widely studied, individual differences that affect language development such as speed of processing has recently received increased attention. Speed of processing refers to how quickly infants process signals such as actions and speech and how this affects their acquisition of knowledge.

In most literature on this topic, infants’ vocabulary size is measured using CDI questionnaires and responses from their caregivers. Infants will then take part in an eye-tracking task based on the looking-while-listening paradigm where they are presented with visual and speech stimuli in the form of two images and a sentence including a target word. The reaction time from the speech stimuli until the infant looks at the image corresponding to the target word is measured. It has often been found that processing speed predicts vocabulary size and acquisition.

While the amount of research on this topic is growing, we have yet to find a meta-analysis consolidating these findings. A meta-analysis is vital for furthering our understanding of how infants’ speed of processing influences their ability to process and acquire new vocabulary. Three research questions were proposed for this project:

  1. Are individual differences in speed of processing stable across development?

  2. What is the magnitude of the correlation between speed and concurrent vocabulary size and is this moderated by other variables?

  3. Is speed of processing related to prior and subsequent vocabulary size and at what distance / direction is this effect maximised?

I was provided with a list of target papers and was tasked to come up with the search terms to find more relevant papers in the databases we planned to use. This was a learning curve for me as I had never used Boolean searches1 before.  Learning the modified versions of it for different databases took a few tries, but I had plenty of help from Andrew and Sam. I devised a list of terms based on the target papers and with some alterations, we went ahead and used them for searches in PsychoInfo, Web of Science and Scopus. This resulted in a large number of papers being identified as potential hits.

During this internship, I have learned the importance of using the right software for dealing with the number of hits we received. We began using Zotero but the de-duplication process (removal of duplicate items pulled from multiple databases by deleting or merging articles) was slow as I had to go through each individually. Following discussion, we decided to use EndNote where the import and de-duplication was much quicker. Having the right tools for the project is so essential for saving time and resources, and for getting a more accurate analysis.

I was also tasked with filling out the pre-registration form prior to the abstract screening such as our inclusion and exclusion criteria, our search terms and methods, and our intentions for the project. I was not aware at the time of what analysis we would be conducting as my internship will have ended before that step in the process, so I completed what I could and sent it off for the team to complete the rest. I was also introduced to PRISMA regulations for systematic reviews and meta-analyses at this stage and began noting down how many papers we imported, how many were duplicates and how many were excluded.

Prior to my arrival, inclusion and exclusion criteria were drafted. This made the next step, the abstract screening process, significantly simpler for me as I could follow the exact criteria and could send any papers that I wasn’t sure of to the team for review. I enjoyed the abstract screening as I was able to not only identify papers that were relevant for the current project, but also find ones that I found personally interesting.

Unfortunately, the 6-week internship ended before the abstract screening was complete and I was unable to see the project to completion due to the time commitment required for a meta-analysis. I have greatly enjoyed my time working with LuCiD and the team on this meta-analysis and am very grateful for the opportunity that they have provided me with. I was able to attend a conference on language development and have gained vital first-hand experience in research. I will take everything that I have learned and apply it to my future endeavours in psychological research.

I am excited to see the results of the meta-analysis as this could be a major step in our understanding of infant vocabulary development and its relation to speed of processing. Faster linguistic processors may have larger vocabularies and may take to reading and writing sooner than slower processors. This could have positive implications for children's education, with children being educated at a rate that suits them.

1Boolean search is a type of search allowing users to combine keywords with operators (or modifiers) such as AND, NOT and OR to further produce more relevant result

Author

Leave a Comment

* Indicates fields are required