Strategies for assessing the impact on primary school pupils of a language development programme directed at teachers.
AbstractMillions of rands have been invested in teacher development programmes provided by South African non-governmental organisations (NGOs) since the mid-1970s. This investment increased dramatically through the 1980s and levelled off in the first half of the 1990s. Money will continue to be invested in in-service teacher education and training (INSET) as the need is still huge. However funders are asking the challenging question of what impact these programmes have on pupils. The Teacher’s English Language Im provem ent Program m e (TELIP), a University based educational NGO which provides English language im provem ent courses for primary school teachers m ust also address this question: w hat im pact does it's programme have on pupils? TELIP does not prescribe how and what teachers should be teaching their pupils but it directly develops teachers’ language skills and confidence. Therefore, assessing the impact on pupils is a challenge. This study aim ed to find an appropriate model of impact assessment for TELIP specifically and for other NGOs with similar orientations. The research design was based largely on a literature review , interviews with specialists in the field and workshops with TELIP staff. The four research methods studied were; experimental design, longitudinal model, action research, and ethnographic research. The research findings indicate that a longitudinal study would be appropriate for TELIP to assess the impact of its English language development programme on pupils. There are four arguments favouring this model. Firstly, it provides an opportunity to do an in-depth study tracing the progress o f the pupils in relation to the skills learnt and put into practice by the teacher during and after the TELIP course; secondly, baseline data could be collected by testing pupils; thirdly, pupils’ written and spoken English can be monitored according to the criteria set by the project; and fourthly, the need to find a matching control group is eliminated. The findings also emphasise the need to work out specific assessment criteria and to consider contextual factors which impact on learning, such as, the authoritarian, bureaucratic nature o f the education system, poor facilities and equipment, ethos o f learning and teaching, other INSET courses that teachers do, and pupil maturation.