OnPoint by Keith Ng


Wrong. Wronger. Wrongerest.

The day before the stabbing at Avondale College, the Herald fortuitously began its own campaign on school violence. In that first story, it said that there were “40 police callouts to NZ schools each week”.

After the story was printed, the Herald issued a retraction and posted the “corrected” version of the story: “Police made an average of 32 apprehensions a week at schools and universities last year”.

Then, in a second story two days later: “Police are arresting an average of 31 people each week of the year at places designated for learning.”

So, from "callouts" to "apprehensions" to "arrests". Which one was it, really? We had a crack at this on Media7 last week and the answer was: d) None of the above.

It's kinda a cliché that those on the Paul Henry side of the media establishment get a little nuts about ... well, a lot of things. And it's kinda a cliché that those of us on the Russell Brown side of the blogosphere get a little nuts on the MSM for pandering to whatever the fear-of-the-day is.

But this time, it's quite a lot more than hoodie-bashing and moral high-horses. This time, the Herald printed – on the frontpage, in big fat frontpage headline fonts – something that was thrice wrong; the statistics that they were quoting, even if they had quoted them correctly, was not an indicator of school violence; and everyone else re-reported those figures verbatim – even though some of the errors should have been evident from the first reading of the stories (as Editing the Herald noticed right from the start).


These callout/apprehension/arrest figures all came from the same table, a petit wee thing that the police sent through in response to the Herald's OIA request on 21 January – almost six weeks before the stabbing took place.

It was a strange OIA, since half the stuff they asked for was on the StatsNZ website, and the other half was a slightly more specific cut of those same statistics, which they could've just asked for and got in two days.

Anyway, the story that came out of this OIA request started:

Police are being called to schools about 40 times every week of the academic year to deal with behaviour teachers say they cannot handle."

Really? The numbers came out of a table entitled: "Table 1: Recorded Offences in Schools and Other Educational Institutions for the Last 10 Fiscal Years".

Recorded offence means that the police were made aware of an incident. (e.g. "Hi. Some cars have been smashed.") It says a crime has occurred. It says nothing about police presence being required, it says nothing about the situation getting out of control, it doesn't even say that there is a situation that needs to be dealt with. Callouts and recorded offences are different. Pretty. Goddamn. Different.

And how'd they get 40 a week out of 1658 a year? By dividing it by the number of weeks in the academic year, rather than, you know, *a year*. As you can see two paragraphs above, the title pretty clearly says "Fiscal Year".

The other part that they didn't read was that the statistics are for "Schools and Other Educational Institutions". And if there was any doubt over the title, the email which came with the figures pushed the point:

The data provided includes only offences recorded with the location description showing as "education school or educational institution". Police do not produce more specific data as to the location of an offence; therefore it is not possible to break down the data into the various types of school (i.e. Primary, Intermediate or Secondary). Additionally, the data provided could include offences committed in other educational institutions such as Universities.”

The fact that it's not just about schools makes using the school year problematic, but it also intimates a much, much bigger problem: These are not schools statistics!

They include universities, polytechnics, wanangas, industry training, language schools, etc. One third of students in NZ study in tertiary education institutions.

So, how much of this crime is happening at those places, and how much is at schools? We don't know. All we know is, of the 1064 violence offences, ???? happened on school grounds and ???? happened on the grounds of some other kind of education institution. And if they happened at school grounds, they may have happened during school hours. Or not. And they may have involved students. Or not.

It doesn't tell us shit about schools.

And that was just the first fuck up. After the story came out, the police comms people called the Herald and told them they got it wrong. The "senior editorial staff" they spoke to were very obliging, says the police, and promptly issued a correction.

Unfortunately, they amended the story the wrong way.

Instead of calling "recorded offences" "callouts", they decided to call it "apprehensions" instead. Putting aside the fact that "recorded offences" is not "apprehensions", it was also quite (additionally) misleading because the stats definition of "apprehension" doesn't mean what we think it means (like, "arrest").

According to StatsNZ:

An ‘apprehension’ means that a person has been dealt with by the Police in some manner (eg: a warning, prosecution, referral to youth justice family group conference etc) to resolve an offence. In some circumstances ‘dealt with by the Police’ may mean that the offender has been found to have a mental condition or is in custody, so no further action is taken other than to document the offence."

That's why calling apprehension stats "apprehensions", without clarification, might be a bit misleading, because people will think that it's arrests.

Two days later, the same reporter comes back and calls them arrests, making it sound like each instance was a case of the police coming down and dragging someone from the school. The fact that she rounded it down to 31 didn't really make it better.

The same set of numbers changed definitions three times in three days - and was progressively further from the mark each time. But even if they got the definitions right, it would still have been useless as an indicator of school violence, because it includes a whole lot of crimes from other places, and there's no way to separate them out.

(And just for an additional level of wrongness, they failed at copying and pasting. They wrote down the 1998/99 figure as 869, when it was actually 839. Sigh.)


So what do we actually know about school violence?

Normally, for something like this we'd look to the Crime and Safety Survey for answers. But the survey excludes under-15 year olds, so it's no good here. The best data we have is from the schools themselves - for every stand-down and suspension, we can find out the reason for them. It's better than the police data because we can be reasonably certain that every incident that's serious enough to warrant police attention is serious enough to warrant attention from the school, but it's exclusively schools.

It's possible that schools dole out stand-downs and suspensions differently now than they did a decade ago, so a change in stand-down/suspension numbers might reflect a change in school attitudes rather than violence.


For physical assault against other students, there were 5.58 stand-downs for every 1,000 students in 2000. In 2007, there were 7.32.

For physical assault against teachers, there were 0.52 stand-downs for 1,000 students in 2000. In 2007, there were 0.81.


For physical assault against other students, there were 1.15 suspensions for every 1,000 students in 2000. In 2007, there were 1.22.

For physical assault against teachers, there were 0.18 suspensions for 1,000 students in 2000. In 2007, there were 0.31.

After we adjust for increased student numbers, schools were standing down and suspending more students for violence in 2007 than they were in 2000.

What does it mean? Schools are taking more serious steps more often when it comes to violence. Why? This doesn't tell us, but at least it gives us a question.

69 responses to this post

First ←Older Page 1 2 3 Newer→ Last

First ←Older Page 1 2 3 Newer→ Last

Post your response…

This topic is closed.