Sally Thomason, at Language Log, discusses a recent psychological study by a University of Michigan professor that makes the following claim: "the greater transparency of numeral words for 11-19 in East Asian languages accounts in part for young Chinese, Japanese, and Korean students' superior learning of math by comparison to American students." The key seems to be the descriptive power of names for the numbers. As Thomason puts it, "both the ordering of the numerals in the compound English 'teen words (3+10, 4+10, ...) and the semantic opacity of the words eleven and twelve make math learning harder for American first-graders."
Hindi and Urdu actually follow the English system, roughly, in having, non-transparent names for the 'teen' numbers (11-19), so if you follow this psychologist's argument, Hindi and English speaking children should have a tougher time understanding relationships between numbers in grades 1 and 2.
Thomason points out her many objections to the study, which all make sense to me. This seems like a very flawed study of an interesting issue.
My suggestion would be to experimentally teach one group of American first-graders a set of alternate names for 11-19, with new names that actually are transparent. Instead of 'eleven,' and 'twelve,' then, one could try 'teni-one' and 'teni-two'...
Then, compare their scores on a basic math test that only uses numerals (i.e., '11' and not 'eleven'). If there's anything in this theory, the 'teni-one' students should do somewhat better.
Something equivalent could be tried in Hindi: for one group of students, get rid of 'gyara,' 'bara,' etc. Replace with 'ek-das', 'bai-das', or something similar.