How WLZ identifies a cohort of children

By Bridget Suthersan, Senior Data Analyst

We identify a cohort as being eligible for support through West London Zone by analysing what we call their risk factors. By this we mean early indications that a child may experience poor outcomes in later life, such as unemployment, substance misuse, damaged relationships, depression, crime.

In identifying students, we consider a number of risks, including poor Maths and English attainment levels, poor attendance at school, and demographic risks, such as whether or not a student is Pupil Premium (PP) or has an Education, Health and Care Plan (EHCP). We also make use of a WLZ survey, and particularly data from the Strengths and Difficulties Questionnaire (SDQ), a validated behavioural screening tool for young people.

Ultimately, in identifying students we aim to construct as detailed and comprehensive a picture of a whole school population as possible, so that we can ensure that we identify the students who can most benefit from our support as well as understand their daily context and provide the school with valuable insights into their community.

But it is not as simple as it sounds. Schools vary, with different data types, systems and collection methods. One school may record academic data in terms of levels or categories (e.g. ‘Above’, ‘At’, ‘Well Below’), whilst another may make use of comprehensive testing with standardised scores.

That’s why West London Zone operates under certain methodological principles, in order to ensure robustness and reproducibility of cohort identification, even across different data sets.

First, risks need to be evidence-based. That is, we only analyse risks that are known to have a deleterious impact on young people’s wellbeing. Further, in classifying a student as being at risk or not, we will always refer to validated measures, where available.

Second, risk factors should be maximised. That means, where the data is available on risks that are proven to be associated with poor outcomes in later life, we should incorporate this data into the analysis. This means that certain risks may carry different weightings, and it does allow us to build a more comprehensive and granular picture of the whole school population as well as the identified cohort.

Thirdly, data quality is paramount. That means using the highest quality data available for classification of risks. As an example, we prioritise standardised, nationally representative academic data over teacher-assessed levels or scores because we can compare it across schools with reliability and rigour.

Finally, in undertaking our risk analysis, we pay particular attention to what we call ‘malleable’ risks – that is, the risks that we aim to influence through our work, namely attendance, attainment and SDQ scores.

This is the first in a series of blogs on our cohort identification methodology by our Senior Data Analyst Bridget Suthersan. The next topic will be an explanation of the WLZ data collection survey, designed in partnership with Dartington Social Research Unit, which includes questions on a range of topics such as emotional well-being peer relationships, social isolation, and family background.