The vast majority is very quick to criticize ICT4D projects and highlight statistics such as the World Bank statistic that states that about 70% of ICT projects fail, without even understanding the context of these numbers. In this case, how does the World Bank define failure? What constitutes a project as a failure? Some projects may be black and white with a clear boundary between success and failure; however, most projects lack this definitive boundary. For example, the Zambian Interactive Radio Instruction (IRI) program, known as Learning at the Taonga Market (LTM) was launched in 2000 to create low cost, high quality education for educationally and geographically marginalized areas in Zambia. The LTM integrates IRI, which acts as an active teaching tool, and the Lifeline radio, which is a dual-powered device that uses both wind-up and solar technology minimizing the dependency on other energy sources to teach lessons written and recorded by the Educational Broadcasting Services in conjunction with the Education Development Center. This program was designed to use existing technology, such as the radio, to provide high-quality education for over 800,000 children who cannot attend school. Since its implementation, over 160,000 children have received education through the LTM and these children have tested better than the children attending mainstream schools.
While this program appears to be working, some people argue it is another failed ICT project. Even though the demand for the LTM program and the enrolment of G1 participants have steadily increased, the retention rate is uncomfortably low as only 2,916 of the total 7,782 learners completed G5. Additionally, when testing the participants’ literacy and numeracy skills, it was apparent that the children had gained knowledge. However, the mean numeracy score dropped from 71.5% in 2001 to 63% in 2003 and the literacy skills dropped from 56.6% in 2001 to 48.8% in 2003. Even though observers noticed an improvement in literacy and numeracy skills, the tests proved otherwise. Does this mean that the project failed?
The lowered retention rate could be due to a lack of monitoring and evaluation; some people could be counted as “drop outs” even if they just switched IRI centers. Additionally, the discrepancies in the numeracy and literacy tests could be due to the different sample sizes tested in 2001 and 2003. Therefore, is it accurate to consider this project a failure on the basis of somewhat skewed data? And even if the data were accurate, should this project be classified as a failure based on two statistics, even when vast improvements and increases in demand have been noted? All these questions cannot be answered unless we define failure.