Developing Low-Cost Ai Proxies for Phonological Awareness in Pakistan's Mother Tongues (Saraiki, Punjabi, Balochi, Sindhi)

https://doi.org/10.5281/zenodo.17494858

Abstract

This paper is concerned with the challenge of critical literacy in the multilingual environment of Pakistan in which children whose mother tongues include Saraiki, Punjabi, Balochi and Sindhi are disadvantaged by an educational system that preserves the primacy of Urdu and English. The study reveals a pronounced phonological awareness gap, one of the cornerstones of reading in such first languages, which is aggravated by the fact that educational resources are lost. The research had two major aims, to determine the current phonological awareness of kindergarten students in these languages and to determine the efficiency of a purpose-specific, low-cost digital tool that seeks to fill this gap. Using a descriptive research design, a baseline data were gathered on 372 children in 25 schools with the help of a culturally based assessment tool. The outcome established that there was a serious deficit and that the mean pre-test result was 43.8% on average. Some level of comparison provided an immense influence of the digital tool, where the experimental condition got a mean of 28.6 percentage points improvement in six times of the control condition. This was observed to have been the same in all four languages and both the high and low priced privately run schools, and indicated a lot of usefulness and scalability of the tool. The research concludes that there is a need to have a set of mother-tongue phonological training to develop literacy. It develops strong support that an intervention of such low cost and technological intensive nature can successfully close this underlying learning discrepancy that offer a feasible avenue to more educational equity and higher literacy rates among millions of children in Pakistan and similar multilingual contexts.

Keywords: Phonological Awareness, Low-Cost AI, Mother Tongue Education, Regional Languages of Pakistan, Language Technology