I am visiting researcher at the Tavistok Aphasia centre, Newcastle University, where I previously was Associate working on the project "Reading comprehension in adults after stroke".
The project, founded by the Stroke Association and leads by Dr Julie Morris, involves a large-scale data collection of people with aphasia, in collaboration with North East NHS Services.
The main aim of the project is to propose a psychometric test for assessing reading in adults after stroke from word level to sentence level. The project will end in Spring 2015.
I am Assistant Professor at the School of Life science in Heriot Watt University. You can find about Teaching and Project supervision at http://www.psych.hw.ac.uk/people/f-m/dr-maria-garraffa.html
My research on Language Pathologies dealt with two major areas:
- the study of sentence processing in acquired and developmental language disorders;
- the discrimination of language vulnerabilities in early second language learners (eL2) and in children with primary language impairment (LIc).
Syntactic Priming in Primary Language impaired children (PLIc)
I am currently working on implicit language learning mechanisms in preschool PLIc, investigating their syntactic development during Structural Priming trials. Together with Prof Holly Branigan and Dr Moreno Coco, we have completed a series of experiments using structural priming paradigm in preschool PLIc and obtaining different long term priming effects compared to typical developing children (Garraffa, Coco, Branigan, 2014).
We are now collecting data on a follow up study exploring the time course of priming to specify the effect rapidly dissipate in PLIc (relative to TDc) targeting implicit learning of passives (with different degree of morphosyntactic cues: verb categories, gender mismatch).
The main aim of this line of research is to investigate the language impairment from the point of view of the language learning mechanisms embedding the study of vulnerable sentences and the possible changes in their acquisition.
We also looked at the modeling of language learning in LIc using Bayesian unsupervised models (Coco, Garraffa, Branigan, 2013). The classification performance of the BA models demonstrates that we can infer the developmental group of a child very accurately with unsupervised model (same as a Binomial Logistic Classifier) introducing the necessity of modeling learning rates of selected structures.
Early second language learners and children with primary language disorders
I was recently involved in two successful external grant applications aiming at studying grammatical development in in two different social contexts: early acquisition in immigrant children and bilingual acquisition in heritage languages (Italian/Sardinian).
Together with Prof Antonella Sorace and Prof Maria Teresa Guasti we just completed the study on development of grammar and executive control in eL2 immigrant children. Overall the study demonstrated a gradual emergence of changes in receptive grammar in children who are becoming bilingual, already visible over a period of two years. A sentences types analysis indicates a possible route of the linguistic advantage in bilingual children, evident in a better comprehension of passives and datives.