How will the sampling methodology be developed?

The committee 'sampling methodology, standards and procedures' made up from experts from each jurisdiction will oversee samples.

• • The study population for iPIPS will be children who are starting school. Given the range of different national and regional policies on early schooling, ‘starting school’ is defined as the point at which children are starting to learn to read formally. This could be called “school” in some countries and Kindergarten in others. This will vary considerably from one system to another, so discussions will take place with each jurisdiction to determine the most appropriate time for the initial baseline assessment.

• • The sample of children must allow us to collect enough data in relation to the different variables to estimate accurately the parameters of interest in the most cost effective way. So we will begin by drawing up a representative sample of schools, then sample the classes, and then sample the children, using a probability proportional to size sampling technique. Our sampling experts together with country experts who contribute to the sampling committee will determine the precise arrangements. Underrepresented groups will be over-sampled as necessary, to provide valid data for these groups.