Cognition and reading in Low SES vs. low SES+ oral tradition children

Shany, M., & Geva, E. (2012). Cognitive, language, and literacy development in socio-culturally vulnerable school children–the case of Ethiopian Israeli children. In Current issues in bilingualism (pp. 77-117). Springer Netherlands. http://www.oise.utoronto.ca/gevalab/UserFiles/File/20142015_Publications/2012_Shany_Geva_in_Leikin_Schwartz_Tobin.pdf

This is an interesting chapter by Michal Shany and Esther Geva, discussing research that compared elementary school children whose parents immigrated to Israel from Ethiopia with children from low SES who are not of Ethiopian origin – in aspects of reading and cognitive abilities. 

The majority of Ethiopian immigrants to Israel come from a rural, oral tradition society, where over 90% of the community members are illiterate.  About 35% of  Ethiopian immigrant adults studied in an Ulpan (Hebrew courses funded by the government), and for most of them this has been their first exposure to a written language of any kind.  Fifty five percent of men and 70% of women of Ethiopian origin are illiterate in Hebrew.  Parents of Ethiopian children typically have 1 or 2 years of education, compared to an average of over 11.5 years for parents of Israeli Jewish children.  Over 52% of families coming from Ethiopia live in poverty, as compared to 16% of families in the general population.

Studies found that there are no significant differences between immigrant children and native speaking children in skills at the single word level like single word identification, single word decoding, non-word decoding and spelling.  This finding is true for many languages.  Thus we can expect that there will be no difference in these skills between a child of Ethiopian origin and his classmates who are not of Ethiopian origin, unless this child has learning disabilities. 

In a number of studies that focused on the word level skills of English speaking children who were learning Hebrew in bilingual English-Hebrew education , it was shown that in spite of limited language proficiency in Hebrew, these middle-class children can learn to decode and spell Hebrew words with accuracy, and decode Hebrew words with efficiency. Thanks to the simplicity of the phonological structure of Hebrew, the lexical outcome of assembling into words a series of matched grapheme-phonemes in vowelled Hebrew is unequivocal, and can be accomplished with accuracy even in the absence of linguistic proficiency.  When spelling-sound relations are straightforward (shallow orthography, there is only one way to read a word) – syntax and vocabulary don't affect single word decoding.  When the orthography is deep (the relation between letters and sounds are not unequivocal, the same word can be read in different ways giving different meanings) – syntax and vocabulary have a significant effect on single word decoding. 

Phonological processing, speed of lexical access and phonological memory predict word recognition skills in English as a second language.  Geva found that phonological awareness is more important in learning to read English than Hebrew.  She found a correlation of 0.62 between single word reading and phonological awareness in the native language (English), but a correlation of only 0.32 between word reading and phonological awareness in Hebrew (the second language). 

Naming speed may be a better predictor of reading fluency in languages that have shallow orthographies like Dutch, than in languages that have deep orthographies like English.   

Bilinguals or people learning a second language, who have problems in decoding and spelling in their native language, tend to have problems in decoding and spelling in the second language as well. 

Compared to single word decoding skills, immigrant children's reading comprehension is affected by many factors like oral language skills, vocabulary, the level of acculturation and familiarity with conventions of text structure and home literacy.  Homes that are literate in the native language (there are book in the native language at home, parents read stories in the native language) increase their children's chances to succeed in reading comprehension in the second language.  Reading comprehension skills transfer from the native language to the second language, when the child reaches an appropriate level of command of the second language.

Three hundred and twenty six children participated in this study.  One hundred and seventy five were of Ethiopian origin and 151 were of other origins.  The children studied in the first, second, fourth and sixth grades in six cities in northern and central Israel.  The children of both groups in each school and grade level came from the same classes.  The children of Ethiopian origin were born in Israel or came to Israel before they were 2 years old.  The other children were born in Israel. 

Although both groups of children came from low SES, the situation of the students of Ethiopian origin was worse:  their mothers spoke less Hebrew, their fathers were more unemployed, they had more people per room at home. 

Beginning in first grade, there was no difference between the two groups in phonemic discrimination and phonemic awareness.  There were also no differences between the two groups and the general population in these skills. 

In another study Shany found differences between kindergarten children of Ethiopian and non-Ethiopian origin in phonemic awareness and discrimination.  It seems that the gaps between the groups gradually close with the onset of formal education.

There was no difference between the two groups in single word reading accuracy in all grade levels.   However, compared with the national norms, both groups performed at the low average range. 

This finding is in line with other studies showing that with the onset of formal education, there are no differences between English as a second language speakers and native speakers in phonemic awareness and in single word reading skills.  These studies also show that regardless of the language spoken at home, the performance of children from low SES homes is lower than that of the general population.  Thus this delay can be attributed less to the language spoken at home than to insufficient exposure to reading, to a rich language and to literary contexts.  Children in both groups were hardly familiar with popular children's books. 

In each grade level, children of Ethiopian origin performed worse than children in the other group on a test of working memory.  Compared to the national norms, the performance of both groups was within average limits.  

In rapid naming of letters and numbers, single word reading speed and spelling, significant gaps between the groups closed by fourth grade.  In lower grades, both groups performed at the low to very low levels compared with the population average.  When the gaps between the groups closed, both of them also performed within average limits. 

Grammar skills:  in grades 1, 2 and 4, the average of the children from Ethiopian origin was in the 25^th percentile, and the average of the other group was in the 38^th percentile.  In 6^th grade, both groups' average was in the 25^th percentile.  This means that the children who were not of Ethiopian origin got worse. 

The low performance of both groups in oral language is probably related to their low SES, and to underexposure to rich language in written and oral contexts.  What's important is not only the fact that the parents don’t speak Hebrew or don't speak it well, but that the parents have low literacy in both languages, and that the quality of the language spoken at home, Hebrew or Amharic, doesn't prepare the children to the quality of language and concepts needed to succeed in school.