Multilingual construction of Communicative Development Inventories in Southern Africa

Alcock, Katie and Dowling, Tessa and Pascoe, Michelle and Mahura, Olebeng and Brookes, Heather and Ndhambi, Mikateko and Southwood, Frenette and Oosthuizen, Helena (2019) Multilingual construction of Communicative Development Inventories in Southern Africa. In: Child Language Symposium 2019, 2019-07-10 - 2019-07-12, University of Sheffield.

Text (Multilingual CDIs poster)
Multilingual_CDIs_poster_2.pdf
Download (601kB)

Abstract

The Communicative Development Inventories (CDIs) are a set of validated, parent-completed questionnaires assessing children's vocabulary, gesture, and grammatical development. They have many benefits but in particular administration does not require a qualified psychologist or speech-language therapist, making them ideal for settings with poor access to professionals. In addition they are cheap to administer at scale, again benefitting low resource settings. Further, they are relatively easy to develop for languages that are not well described in terms of development, meaning they can be adapted easily to under-studied languages. The CDIs have previously been adapted for Asian (1), Indo-European (2), other European (3), Pacific Island (4) and some African languages (5,6). They have been used successfully in studies of health risks (nutrition, infection) in sub-Saharan Africa (7), again meaning they are useful for impoverished settings and child development issues prevalent in such settings. We are in process of developing CDIs for six languages spoken in Southern Africa (IsiXhosa, Setswana, Sesotho, Xitsonga, Afrikaans and South African English). We have developed a common method for each language starting with a long list of words taken from other CDIs, following on to parent/practitioner assessment of face validity of these words in their translations, and continuing to piloting, instrument reduction, and validation. Within validation we look at correlation with age, and relationship to other variables including family socio-economic status, and child language performance measured in the home or lab. We also apply these methods to the assessment of gesture using the CDI. Some of the languages to be studied are related to each other (of the Bantu group of African languages) and to two languages for which CDIs have been developed in East Africa (5); for these we will highlight common grammatical development to be examined that exists in all of the related languages. In the East African setting parents were found to be valid observers of their child's morpheme omissions so this is a particular focus for assessment of grammatical development. To date we have developed four pilot versions (IsiXhosa, Setswana, Sesotho, and Afrikaans). Vocabulary in isiXhosa is found to be significantly related to toddler age, mother education, and first-born status, among other known correlates of child language (8). In Sesotho, toddler vocabulary is also significantly related to age (9). In the other languages (Xitsonga and SA English) we have created pre-pilot word lists and carried out focus groups with parents enabling us to determine the acceptability of the words on our long-lists. We conclude that construction of CDIs in multiple (some unrelated) languages simultaneously is possible with rigorous application of the same methods to each language. Where languages are under-studied, we can benefit from previous research and descriptions of development in related languages. 1.Tardif, J. Child Lang. 36, (2009). 2.Caselli, J. Child Lang. 26, (1999); Bleses, J. Child Lang. 35, (2008). 3.Barrena, in A portrait of the young in the new multilingual Spain. (2008), vol. 9. 4.Reese, First Language 35, (2015). 5.Alcock, J. Child Lang. 42, (2015). 6.Prado, Journal of Health, Population and Nutrition 37, (2018). 7.Prado, J. Child Psychol.Psychi. 58, (2017); Alcock, BMC research notes 9, (2016). 8.Whitelaw, University of Cape Town (2018). 9.Horn, Cape Town (2018).