This final report describes a theory-driven evaluation of the Trinidad and Tobago Continuous Assessment Programme (CAP) implemented in the primary school system. Evidence was collected using a multiphase mixed methods research design, with information gathered on the fidelity and strength (intensity) of the programme in a sample of 60 schools in the seamless project and 40 schools in the original CAP pilot project.
Phase I was an exploratory qualitative study of eleven sites, Phase II was a quantitative modelling study based on a multi-instrument survey of 378 teachers in 35 schools, and Phase III was an explanatory qualitative study of two schools with different levels of implementation and two principals at relatively high implementation sites.
The evidence collected suggests that programme strength was variable across sites. In schools reporting high programme strength, most of the original CAP activities were still being done. These schools usually had strong leadership by principals who were originally trained under the CAP Pilot programme. However, in schools with low programme strength, some teachers were not even aware of the CAP.
Overall, the concerns based profiles suggest that the system was dominated by nonusers of CAP, although there might be lower resistance in schools that reported high implementation.
The most frequently implemented activity was the CAP project, but this activity was not conducted in a way that provided the full benefits of formative assessment. Neither was there efficient use of the data collected even at high implementation sites. Thus, programme fidelity was judged to be universally low, with inadequate formative assessment, a lack of feedback to students, and poor or inappropriate data use. The quantitative modelling study confirmed that fidelity outcomes, such as providing feedback and using multimodal assessment, were very different to ?doing? CAP and, as such, the independent variables in the model were much less predictive of fidelity measures.
The integrated meta-inferences confirmed that several contextual, organizational, and teacher variables were important, with user variables more important in fidelity.
Most students in schools had very positive views of assessment, but tended to see classroom assessment as a mechanical recording of grades or used as an accounting procedure, rather than as a tool for improving learning in the classroom.
The key recommendations focused on the need to develop a coherent and balanced comprehensive assessment policy and a simplified, targeted and restructured CAP, with an emphasis on formative assessment.
Caution should be exercised when developing new policy that seeks to integrate classroom assessment with the placement function currently held by the SEA. The wide variation in practice, universal absence of training, and general lack of support at building sites suggested that teacher measurements and judgements would lack validity and integrity within a high stakes context.