This article has been updated. See bottom for details.

Readers of the Guardian's datablog on Monday would have seen a detailed analysis conducted by the newspaper of the latest Department for Education (DfE) figures.

Their headline, however, provoked some controversy:

Church schools shun poorest pupils

This conclusion was based on a number of figures that appear to show faith schools tend to contain a smaller proportion of pupils on free school meals than the standard rate for their local area.

According to the data, around 73 per cent of Catholic primaries and 72 per cent of Catholic secondaries have a lower proportion of pupils who are eligible for free school meals than the average for their local authority, with a similar picture for Church of England schools.

Meanwhile, only around half of non-faith schools had a smaller proportion than average for their local authority.

However, this appeared to contrast starkly with the picture painted by Maeve McCormack, representing the Catholic Education Service (CES). The CES comment that:

"18.6% of pupils at Catholic primary schools live in the 10% most deprived areas, compared to 14.3% of pupils at primary schools across England. At secondary level, 17% of pupils at Catholic secondary schools live in the 10% most deprived areas compared to 12% of pupils nationally."

So how can we reconcile these opposing comments?

The Catholic Education Service draw their statistics from the Department for Education school census for 2010. They measure the level of deprivation based on where a child lives, and chart the proportion of children residing in the most deprived areas, measured by decile.

The graphs below, presented in the CES's recent report, outline their findings:

A closer look at the methodologies behind the Guardian and CES statistics reveals the main cause of the difference - the measure of deprivation.

The Guardian use eligibility for 'free school meals' as their proxy for deprivation. In fact, this is a measure used by the DfE for allocating the Pupil Premium. They state:

"Free school meals are a key accessibility indicator, as only the least well-off children are eligible for them"

The Catholic Education Service, meanwhile, are less keen on using the measure. They comment:

"It is an imperfect indicator for several reasons: for example, parents cannot be obliged to claim FSM or to declare their financial situation, and anecdotal evidence suggests that many do not claim because of shame at revealing their poverty. In consequence, the figures usually collected are ‘for those known to be eligible for and claiming FSM’"

Instead, the CES data uses the 'Income Deprivation Affecting Children Index' (IDACI). This index combines a number of measures of deprivation in an area including households in receipt of benefits, houses without central heating, air quality and even prevalence of severe road traffic accidents.

The Department for Communities and Local Government outline the details behind this measure in their Indices of Deprivation report from 2007.

The CES also commented that the catchment area for Catholic schools were geographically wider than the local authority and postcode data used in the Guardian's analysis. 

Full Fact was unable to contact Ms McCormack to clarify the caveat, however the issue of catchment areas does seem crucial given that the newspaper's analysis compares deprivation based on a different sample than what the schools actually draw their pupils from.

So the Guardian's headline, while it reflects their own research, should be considered in the context of their own methodology. They arrive at the conclusions in their piece using local authority and postcode areas of comparison and free school meals as an index of deprivation.

Meanwhile, the CES use a broader definition of deprivation and claim that more precise catchment areas are needed to make the analysis more meaningful.

Update (08/03/2012)

The Catholic Education Service have got back to us and clarified their critique of the free school meals measure of deprivation.

Eligibility for free school meals is not automatic, and requires parents to register. This can mean a potentially significant number of pupils may be missed by free school meals eligibility statistics. The CES emphasise that the IDACI used in their report is better in so far as it is based on a survey rather than requiring the parents themselves to register.

Update (28/03/2013)

Since the time of writing we've received feedback from a number of readers pointing out the drawbacks of using the Catholic Education Service's figures as a proxy for measuring whether faith schools are indeed 'shunning' the poorest pupils. These are valid and worth spelling out in more detail.

Using the Income Deprivation Affecting Children Index can tell us about which postcode areas in the country are more or less deprived than others, and we can compare this to the locations of Catholic and non-Catholic schools. This confirms, as the CES state, that Catholic schools have a larger proportion of pupils from the most deprived areas than the national average.

However we can't then assume this tells us anything about selection, since Catholic schools could simply tend to be located in more deprived areas. IDACI as an index can't distinguish between different families' income and so wouldn't be an useful measure for what the Guardian are setting out to achieve. The Guardian's analysis gives us a useful insight into these differences, but we should remain aware that since schools' catchment areas are different, this analysis does not entirely compare like-with-like.