Did you hear? Apparently the average four-year-old gets hit 936 times a year. Blimey, that’s a huge number. I must be a super parent – I don’t smack my kid nearly that much! Mind you, most parents these days are just drug-addicted benefit-scrounging criminals who shut their kids up with chocolate so they don’t make noise during the Jeremy Kyle show, right? So I guess you can’t be surprised.
Did that number sound suspiciously high to you? It did to me. So luckily for you, I’ve tracked down the original study. And guess what – that number doesn’t stand up to scrutiny AT ALL.
This study wasn’t even TRYING to find out how often kids were spanked. It was a tiny study of just 33 mothers from one Southwestern US town who were asked to wear audio recording devices on their arms for four or 6 evenings in a row. These kids weren’t even four years old – they were aged from 2 to 5 and a half. Needless to say, this would NOT give you a good estimate for how often four-year-olds in general are spanked over a whole year.
Far from demonstrating an epidemic of child abuse sweeping our country, only 15 of the 33 families in the study were heard to spank their kids – so, more than half the kids weren’t spanked AT ALL. And out of those 15, ten were only spanked once or twice during the 5 evenings they were recorded.
So how does this translate to the average four-year-old being spanked 936 times a year? Well, the authors say that their data adds up to kids being spanked on average 18 times per week. BUT – they calculate this average only from kids who were spanked at all – which we know was less than half of the kids in this study!
The average rate (the median, for stats geeks) used by the authors to get to this stat is 0.22 ‘spanks per hour’. To make this add up to 18 spanks a week they must have presumed that the mothers were in contact with their children for 82 hours a week, or 11.7 hours a day. This just wasn’t the case – more than 60% of the mothers worked full-time and only 14% didn’t work – not to mention that some of the children were school-aged.
So let’s try and get a better answer. The researchers saw 41 incidents of spanking, giving an average of 1.24 per family. The families were recorded for an average of 12.95 hours each, giving a ‘spank rate’ of 0.096 ‘spanks per hour’ across the whole group. Let’s assume that the full-time-working mothers were away from their kids for 44 waking hours a week, and the part-timers and students for 22. Plugging all those numbers in gives an average of 150 spanks a year, or three a week. Not ideal perhaps, but it hardly suggests the US is a nation of child abusers.
Let’s also think about when the recordings were made – between 5pm and bedtime. I don’t know about you, but our house is fairly calm until you try and get a toddler to do anything on an adult schedule. And when we’ve both come in tired from a long day, we’re not exactly in the mood to be patient while our toddler throws his dinner on the floor, wants to watch just a bit more TV or refuses to get in the bath (or out of it). We don’t tend to spank, but the rate of sub-par parenting (snapping, shouting, impatience) is definitely higher in the evenings, and I’d HATE if someone judged us on what goes down between 5pm and bedtime.
That said, there were troubling discoveries in this dataset. Most troubling was that the kids didn’t seem to be spanked just for serious misbehaviour. I’m sure we’re all sympathetic to parents who decide that a quick smack is needed to keep their children away from a hot stove or to stop them kicking the baby. Unfortunately, that sort of misbehaviour accounted for only 15% of offences which led to a spanking – the most common infractions were breaking simple social rules such as ‘sucking fingers, eating improperly, getting out of a chair or going outside without permission’. Spanking also didn’t seem to be the last resort – parents tried on average only one other disciplinary technique before spanking, and on average it took less than 30s for a misbehaviour to escalate to corporal punishment.
What about the study’s original aims? Well, it seems that people are reasonably honest when you ask them about spanking. Only two mothers who said they didn’t spank actually did, and parent’s reports of how often they spanked correlated well with how often they were heard hitting their kids in the recordings.
It’s traditional to blame the media for overblowing this sort of study, but frankly the authors are as much to blame here – they come up with the extremely unsound figure of 18 spanks per week and they write it in the paper itself – despite acknowledging some of the difficulties I’ve listed above. And it only takes a cursory glance through the paper to see that they come down heavily on the anti-spanking side of the debate – they claim spanking ‘doesn’t work’, because most kids are misbehaving again within ten minutes, but don’t say how long a telling-off or a time-out is effective for.
This whole thing makes me extremely angry. I’m angry at the study authors for putting forward such a bullshit statistic, and I’m angry at the peer reviewers for letting it through. I’m angry at the people who’ve been posting it on my Facebook without questioning how high that number seems (but gratified at the number who have). I’m angry because it does no-one any good if everyone thinks that most other parents are horribly abusive. It means we’re even less likely to be sympathetic to the parent who loses it in the supermarket, or lashes out during the school run. We won’t assume that they’re having a momentary loss of control, we’ll assume their part of the half of parents who hit their kids more than 900 times a year. And I’m sure we all know that when you’re at the end of your tether and your child is having a public meltdown you’re struggling to cope with, the MOST HELPFUL THING is judgement and alienation. Not.
I’m glad you took this on. I was incredulous when I saw that statistic.
Not to mention that a lot of behavioral economics research shows that when you imply that “everyone else” is doing something (even something unsavory) it makes people feel like it’s ok for them to do the same. So, promoting a fake stat saying “OMG YOU GUYZ ALL OTHER PARENTS ARE SPANKING THEIR KIDS A BUNCH OF TIMES A DAY!” is likely to 1) Make parents who are already spanking their children feel more ok with their decision since everyone else is doing it too and 2) Make parents who are not spanking their children feel like they are doing something wrong by bucking the parenting norm. In other words, it is more likely to cause people to be more likely to spank than to stop spanking. These anti-spanking researchers are likely undermining their own cause by releasing faulty information.
I can’t help myself. Average and median are so not the same thing! Otherwise, wow what a terrible misrepresentation of data! I hate it when science is distorted to make click-bait headlines, but this goes way beyond that.
I think this might be an American/British thing? Over here ‘average’ is mean, median or mode, whereas I think across the pond it’s sometimes used to refer to just the mean?
I was going to add a paragraph discussing how it’s disingenuous to use the median like a mean because you can’t multiply it around in the same way, but that seemed a little nerdy. It’s also virtually impossible to work out how much difference it would have made without having access to their raw data.
Yes, here average is only the mean. Of course the data you’re describing might be quartiles for all we can tell! (Why yes, I am a stats geek, why do you ask?)
I feel compelled to point out that while 1000 spankings a year is a big problem, so is 150/year. This is a spanking every 2-3 days. For me, that amount of spanking is a problem.
Even the figure of 150 a year is likely to be an overestimate, as they’re recording at a particularly fraught time of day. I agree it’s not great, but to me <150 a year suggests that most parents try not to spank but sometimes lose it, whereas 936 a year would make me think that most parents use it as a routine and habitual part of their discipline strategy. Either way, the authors could have been conservative with their statistics and still highlighted a problem.