What is amiss in school bullying statistics? The data or teachers?

Over the last twenty years, I have worked with many schools wanting to reduce pupil bullying. I have seen many schools draw up anti-bullying action plans, but often those plans have struggled to ground themselves with clear data and demonstrable outcomes.

Unless a strategy is grounded in reliable data, there is always a risk that the strategy is misfocused. Without data, it is difficult to monitor the impact of the strategy. When the data is problematic an anti-bullying strategy can end up being ineffective.

In this piece I shall clarify the problem, hoping that the clearer the confusion becomes, the more likely it is that solutions will begin to emerge.

Bullying continues to be a serious problem in schools

We know that school bullying has very serious consequences. In the UK it drives children out of school, accounting for 3% of pupil absence (according to parents) or 9% of pupil absence (according to pupils).

We also know that it has very serious health consequences with almost a third of those who have been bullied in the UK going on to self-harm, and 44% going on to develop mental health issues, including depression (Ditch the Label). 

Victims of bullying are up to 9 times more likely to commit suicide and it has been estimated that almost half of the 176 suicides of UK children (aged 10-14) which took place between 2000 and 2008 can be linked to bullying (Beat Bullying).

No one can be in any doubt that bullying is potentially one of the most serious threats to the health and wellbeing of many children. For all schools there is an urgent and pressing need to robustly monitor and reduce incidents of bullying.

The limits of data

One of the most fundamental issues confronting a school improvement plan is getting an accurate view of what bullying actually exists within a specific school. This is far from straightforward.

There is an impression that children will just tell teachers if bullying is taking place. OFSTED (the UK inspectorate of schools) reported that 83% of pupils believe that bullying will stop if it is reported (No Place for Bullying). This might suggest that all a school needs to do is ask the pupils about what bullying is taking place.

However, the 2016 UK Annual Bullying Survey gives a far more troubling impression. It reports that 40% of the children being bullied had not actually told anyone when they were bullied. This is a ‘high’ figure of children who seem to be keeping silent. Perhaps there is a temptation to question the reliability of the figure? However, there is a degree of independent confirmation provided by the separate UNESCO study School Violence and Bullying, which looked at bullying across the world and found that on average around one-third of pupils keep silent and do not tell anyone when they are bullied. 

There is a range of different reasons for why children are reluctant to speak up about bullying, ranging from fear of the consequences to a lack of confidence in a school’s ability to deal with bullying. Whatever the cause of pupil silence, what it means for schools trying to reduce bullying is that simply asking pupils if they are being bullied will not necessarily generate an accurate set of data about the scale or extent of the problem of bullying within a school.

When schools generate data about internal bullying they must ‘test’ their data and ask whether there is a risk that it is under-recording the scale of bullying. Unless they at least entertain that possibility and test for it, there is a very serious risk that a school will have a falsely optimistic view about bullying and so misfocus and under-resourced strategies to deal with it.

One of the basic ways for a school to ‘test’ its data is to compare it to ‘benchmark’ data on what we know to be average levels of school bullying in ‘other’ schools. Unfortunately, using benchmark data for bullying is much easier said, than done.

Establishing a general baseline of bullying

Working out a baseline of school bullying is very complicated as there are different surveys and studies, which often use different definitions of bullying and ask different questions about bullying. Some of the main studies used by educationalists in the UK include:

• Annual Bullying Survey (2016) with 9000 respondents aged 12-20
• PISA (2015) with 540,000 respondents aged 15, from 65 countries
• DFE Evidence from Longitudinal study (2015) 11,000 respondents aged 15
• UNESCO School Violence and Bullying (2017) with data from 100,000 children in 18 countries
• OFSTED No Place for Bullying (2012) with 1060 respondents aged 7-18.

There are methodological difficulties in trying to draw conclusions across the surveys, but there are also similarities which enable us to place elements of the surveys alongside each other. The table below summarises some outcomes from these surveys. The Annual Bullying Survey (ABS) includes percentages in square brackets which are reports of bullying by perpetrators, in addition to percentages which represent the reports of bullying by victims.

Daily bullying

Interestingly the UNESCO survey arrived at the same 5% figure as the ABS survey. UNESCO report that 66% of pupils around the world said that they had been bullied at some point, and of that 8 % reported being bullied daily. 8% of 66% is 5%. The ABS and UNESCO figure for daily bullying is also broadly in line with the OFSTED figure of 6%, which was the proportion of pupils (of all ages) who confirmed that they were currently being bullied. The broad convergence of ABS, UNESCO and OFSTED suggest a degree of credibility can be given to the figure of 5% as the proportion of pupils who are being bullied on a daily basis.

Monthly bullying

The PISA study tells us that a quarter of UK pupils report that they are bullied more than once a month. (This compares to a PISA international average of 19% across the OECD countries). The ABS survey doesn’t give an equivalent figure, but if we add up the tally of ‘daily,’ ‘several times a week’, ‘once a week’ and ‘once a fortnight’ this gives us a figure for the proportion of pupils reporting that they are bullied more than once a month. That figure comes to 27% (55% of 50%). This is close to the PISA figure of 25%, suggesting that there are grounds for a conclusion that around a quarter of UK pupils are experiencing bullying more than once a month.

Annual bullying

As well as a daily and monthly figure for bullying, school statistics typically need a baseline figure for the proportion of pupils who are bullied over the course of a school year. This figure is much harder to arrive at. The ABS survey records that 50% of pupils have ‘ever’ been bullied, but it distributes that cohort into bullying which has occurred at different frequencies over the previous year, suggesting that 50% is the figure for the proportion of pupils bullied over a year.

There is a considerable difference between the OFSTED figure of ‘ever’ bullied (50% primary, 38% secondary) and the UNESCO figure (66%). We know from the PISA study that bullying in UK schools seems to be above international averages, but we also know that there are some extreme examples of bullying in some other international contexts, which massively skew data sets. It is difficult to close the gap between the UNESCO and OFSTED figures without doing further work on the data. However, neither figure is actually a figure for ‘bullying over the course of a single school year’ so even if convergence was achieved on those data sets it does not necessarily provide clarity about what proportion of pupils typically report that they have been bullied in a school year.

 Perhaps the most relevant data set is provided by the UK Department for Education (DfE), as one of the questions which it is looking at is whether bullying is reducing over time. According to the DfE study annual bullying has gone down in the UK from 41% to 36% during the first decade of this millennium (DFE Estimates of Y10 Bullying). However, at the same time that the DfE reports that bullying is reducing, Child Helpline International reports that the numbers of pupils asking for help with bullying has significantly increased. Of course, it is not impossible for bullying to be reduced, whilst the numbers of pupils asking for help is increasing (especially if there is a greater awareness of bullying driving the increased requests for help). However, the data is complex and so a caveat is needed.

With due regard to necessary caveats, what the DfE study is essentially stating is that bullying amongst 15-year-old children was running at a level of 41% and that is it ‘now’ running at a level of 36%. In so far as that data is reliable, this gives a figure for the proportion of pupils experiencing bullying in a school as at least one third.

The data from OFSTED, above, seems to suggest that reports of bullying amongst older pupils are generally less than that reported by younger pupils. So, as the DfE study relates to 15-year-olds, this means that the proportion of younger pupils experiencing bullying during the course of a school year could well be considerably greater than one-third.

Drawing conclusions about a baseline for school bullying

We have seen that there are a number of serious problems with trying to produce a figure for a general baseline of bullying. There are differences of definition and differences of understanding of what constitutes ‘bullying’. There is also the fact that some pupils do not report the bullying which they are suffering from and different age groups seem to report different levels of bullying. The contexts of different schools give rise to different causes and different types of bullying, and different schools can apply very different levels of resources to resolving bullying. So, all in all, drawing generalised conclusions about bullying is difficult.

However, problematic though school data on bullying undoubtedly is, we have also seen that there are surprisingly similar degrees of convergence between the outcomes of independent studies about bullying. Those studies seem to be telling us that in UK schools:

• around 5% of pupils experience daily bullying
• around a quarter of pupils experience bullying more than once a month
• at least a third of pupils experience bullying in the course of a school year.

When these kinds of figures are put in front of teachers, school leaders and inspectors, there is often incredulity about their accuracy. Teachers will often state that they have never worked in a school with bullying running at anything like these kinds of ‘high’ levels.

This isn’t just the incredulity of complacent teachers in bad schools trying to ignore the problem of bullying. It is an incredulity which I have seen amongst outstanding teachers, excellent school leaders and professionals with vast amounts of successful experience in reducing school bullying.

So, there is a problem. Either something is amiss in the data and independent studies are reaching a puzzling convergence which massively over-estimates the amount of bullying which takes place in schools. Or something is amiss in the experiences of teachers as there is a puzzling lack of evidence in school self-reporting about the scale of bullying which the data seems to suggest is taking place amongst young people.

Rory Fox CEO Edlumino. June 2016.