My Storify: Robots, Cows, Farmers

Superstitious? Don't be.

Its Friday 13th. You have something stressful to do. Do you put it off until another day because it'll be bad luck to do it on Friday 13th? If that stressful thing was a TB test would you think it might be better to arrange for the following week? And if you did, would you be right to do so?

Studies of medicine have a long history of looking to see if there is any temporal pattern to the outcomes of things like operations. You may have seen this on the news recently, a study which said that "patients who are admitted to hospital at the weekend are more likely to die than those admitted during the week". There are lots of other studies that do similar things but with days like Friday 13th, full moons, and superstitious periods found in different cultures and religions.

For example, according to farming folklore, dogs are more aggressive at full moon (Chapman and Morrell, 2000). However, analysis of dog bites requiring admission to hospital during the lunar cycle fails to find any relationship. Other studies also fail to find a relationship between lunar cycles and psychiatric disorders, anxiety, suicide, or emergency hospital admissions (see Raison et al, 1999; Owen et al, 1998; Wilkinson et al,1997; Mathew et al, 1991; Cohen-Mansfield et al, 1989; Thompson and Adams, 1996) but some weak associations with spontaneous full term delivery, aggression and crime (Ghiandoni et al, 1998; Lieber, 1978; Thakur and Sharma, 1984).

By contrast, Yang et al (2008) suggest that in southern China people avoid taking risks during so-called “ghost months” when superstition has it there are wandering ghosts released from hell. The study finds that the number of accidental drownings was significantly lower during ghost month periods. Similarly, another study (O’Reilly and Stevenson, 2000) shows that a superstition over being discharged from maternity on a Saturday leads to higher than expected discharges on Fridays and lower discharges on Saturdays.

The evidence for the links between Friday 13th and human health is mixed. One study (Scanlon et al, 1993) of admissions to accident and emergency on Friday 13th compared to Friday 6th shows that whilst some types of accident (such as poisoning) were not higher on Friday 13th, others specifically transport accidents were higher on five out of 6 friday the 13ths despite there also being fewer vehicles on the road. In fact, the study reveals a 52% increase in the risk of a traffic accident on Friday 13th. Other studies find a link between car accidents and Friday 13th for women but not men (Näyhä, 2002, but later contradicted by Radun and Summala, 2004) and a full moon (Laverty and Kelly, 1998). Friday the 13th is also unrelated to use of emergency services (Lo et al, 2011) blood loss, emergencies or intestinal perforations during operations (Schuld et al, 2011) and post-tonsillectomy haemorrhage (Kumar et al, 2004).

This may all sound like a bit of fun and jokey research: is there much you could do about any of this if it turned out to be true anyway - what is the causal link? Overall, if you lumped all these studies togeher and ran a meta-analysis, you'd probably find no effect.

But anyway, what about studies of animal health? Is there any evidence that Friday 13th is not a good day to be conducting a TB test, for example? Well, Ive looked at the evidence from thousands of TB tests going back to 1992. I compared results of tests read on Friday 13th with those read the day before, on thursday and friday the week before, and on thursday and friday the week after. Realistically, these are the days you would test if you hadnt wanted to read the test on Friday 13th.

For those of you who are superstitious, the results are disappointing. In the analysis, adverse tests (thats finding either a Reactor or IR, or both) on Friday 13th were never more likely to occur than on any of the other possible reading days. If there was any effect, then it was the other way around. The analysis seemed to show that bTB tests the week after Friday 13th turned up more adverse tests than on Friday 13th.

These are only some preliminary results and Ill be writing them up as soon as time allows. But, if you have any other superstitions when it comes to animal health please let me know: email them to me, or tweet them to @garethenticott. I'll also post the full references eventually, but you could probably find them in Google Scholar.

So, if you are testing today, good luck. You may well have been better off not waiting until next week.

A resurgence in wildlife crime? Or just bad stats?

This week in the Guardian, Patrick Barkham claimed that badger baiting was "making a comeback". Now, the idea of badger baiting is truly abhorrent, but does that mean that we can't look at this claim critically? The article presents some data saying that reports of badger baiting have increased. Here's exactly what Patrick wrote:

"but according to Operation Meles, a police and charity partnership to combat badger persecution launched this autumn, there were 243 reports of badger fighting in 2009 and 2010. The RSPCA recorded 355 incidents of badger persecution, including illegal snaring as well as digging and baiting, across Wales and England in 2010, compared with 255 reports in 2009".

There are a number of problems with this. Indeed, if the kind of stats used in this article were about cancer or homeopathy, the internet's stats geeks would be all over it. The fact that they are not, or that the comments section is similarly absent from dissent, says a lot. Does it mean that when it comes to things like this, things which arent fun or amusing, we can be a bit looser with data? I'm not sure about that. 

One by one, here's whats wrong:

There's no doubt that doing research on wildlife crime is complicated. The data often isnt there - many wildlife "crimes" are simply not "crimes". Some research - like this one on Dog fighting [free pdf] - uses interviews offenders to get round this. If you're interested in understanding the reasons and rationalisations for dog fighting, then you could watch this video too:


The Causes of Urban Dogfighting: Annual Lecture 2010 from League Against Cruel Sports on Vimeo.

These methodological challenges make doing criminological research on wildlife crime challenging to the say the least. The case is a bit different for badger baiting: it is a statutory offence, so data should be available. But to claim that something is making a "comeback" - a bold claim - we need to know what level it has been in the past. This is particularly important when the overall level is relatively low: small changes may result in apparently big differences (a comeback? see below). The questions we need to ask are what were levels of badger baiting like in the 1970s, 80s or 90s? In which year was the highest level of badger baiting incidents recorded? We're simply not told: hardly the basis to make a claim of a comeback. Indeed, basing this comeback on two years of data is hardly a trend either. Why isnt that data being presented? If it doesnt exist, how can a comeback be claimed? Is it down to bad data management? Or it could be down to all sorts of other reasons...

Secondly, the article mixes up reports with incidents - they are completely different. You may have multiple reports of just one incident: it may be that more people are reporting the same incidents. Ben Goldacre wrote a similar explanation on reports of a rise in the number of people taking anti-depressants based on the number of prescriptions. However, in this case as the data seems to come from the same source its likely this is just careless use of language. Nevertheless, the data mixes up all sorts of different types of wildlife crime under the heading of badger baiting: snaring is not badger baiting, at least as it is traditionally understood (but could conceivably contribute to it). Moverover, the data from the police does not even record a rise in reports of badger fighting (which the article seems to be about). For this data, it would be interesting to know how many reports of the same incident there are and whether this has changed. This would give us a clue to the explanation...

Thirdly, if incidents have risen, what is the explanation: there may be other things going on with the data that explain the changes rather than reaching for some societal breakdown explanation. Simply, the context of reporting may have changed. People may have become more aware of badger baiting through awareness campaigns, or through badger culling being in the news. Given the timescales this is possible.To be fair to Patrick, he does kind of undermine his resurgence argument in the same sentence by pointing this out:

"Most of those involved in wildlife crimes believe the current increase is due not to a contagion of cruelty but to the growing courage of witnesses...to report such vicious acts"

The other possibility is that more resources have been put into policing crimes against badgers. When you look more often for something, you tend to find it more often. So any rise in incidents may not be a real increase at all. Again, this is quite possible. During this time there has been increased resources put into the creation of the National Wildlife Crime Unit (launched in 2006). Badger baiting has always been on its list of priorites. The article itself makes the point that Operation Meles was set up to combat these crimes (although this started after the reporting period). Again, Patrick undermines the "resurgence" argument when he also points this out:

"North Yorkshire police have become more robust with it all. We've started to go after them. One of the reasons the police are so interested in poaching, apart from trespassing, is that quite often people are suspected of going into outbuildings on farms and stealing equipment, tools, diesel and metals"

Of course, its conceivable that a reduction in resources/changes to police priorities explains the rise: as the risk of prosecution declines, so might offenders be more likely to engage in these activities. Given that resources are pretty low for wildlife crime though, I think this is unlikely. Whatever, the point is that changes to the policing context mean that "rises" are not necessarily real.

The increased resources issue is probably a good explanation for this. In Fyfe and Reeves chapter in this book, they describe how police forces have been taking wildlife crime more seriously, thanks to extra resources. The establishment of the NCWU and an increase in the number of dedicated wildlife crime police officers, but it hasnt been plain sailing either: even those dedicated wildlife crime police officers still make the distinction between wildlife crime and "real" crime. Its this real crime that is the priority whatever anyone may say about wildlife crime. Again, it may be. Their intention is to plough more money into catching offenders by working with the police whilst also effectively funding their own wildlife crime enforcement agency. But the absence of an equivalent baseline means that what we won't know from this is whether there has been a rise in badger baiting. Perhaps the most interesting thing about this is the "Big Society" approach to the enforcement of wildlife crime. Despite a recent increase in policing resources, you could surmise that LACS dont have much faith in the police's ability to really deal with the problem of wildlife crime (as they see it). That opens up a whole can of worms over the purpose of regulation, whose role is it to enforce statutory regulations, and the point of statutory regulations without statutory enforcement. Returning to the badger cull, if the Government have privatised wildlife control by licensing farmers to cull badgers, then LACS actions show the extent to which the protection of wildlife is also being privatised.

Finally, I was interested to see the article make a link between bovine TB and the "rise" in badger persecution.

"According to Ian Hutchison of Operation Meles, some of the increase in the persecution of badgers in particular is connected to their unpopularity with the farming community because of bovine TB in cattle. Until there is a cull, some people are taking the control of badgers into their own hands. But the setting of dogs on them is a different kind of killing. "It's a macho thing, to be honest," says Hutchison".

Depending on how you want to define persecution, I think the point being made here is mostly wrong. You can read some of my research on farmers' attitudes to badgers and culling them here. Of all the farmers that have participated in our research, most of them regard badger baiting as from another world. Some of them have talked about the people involved in it as being from another world; they mention dodgy people in pubs; and talk of not wanting to get mixed up in that business. Sure there are those who will turn a blind eye if people turn up on their land. Neither does it mean that farmers dont kill badgers. But what our research shows is that farmers views on badgers are intensely complicated: protective, hostile, ambiguous, contradictory and changeable.

Its also interesting to see a police officer recognise a distinction within the treatment of badgers: he seems to suggesting that farmers killing badgers is not "real" wildlife crime. Its seems likely that its those "real" criminals that he's going to go after, and this distinction may well legitimise the continued actions of those others. This is what wildlife criminologists agree on: that there is a broad spectrum of what constitutes wildlife crime. And what regulation researchers agree on is whatever the agreed definition, only part of it will be enforced - it will vary according to a range of social factors. That's probably not good news if you're a badger, but it doesn't mean there's been a resurgence in wildlife crime.

The origins of the badger cull

Sorting out my clutter, I came across this on my hard drive. In 2006 Jim Paice gave a presentation to a TB Forum at the Hatherleigh show. There were some other speakers as well, but here he is outlining how important a badger cull would be (along with other measures - you can hear his desire for PCR too). All of this was a year before the ISG released their final report.


Jim Paice MP at the Hatherleigh show 2006 by GarethEnticott