Could someone please explain to me exactly why the ONS stats on domestic abuse are not accurate?

Another disingenuous contribution from Mr cadno.

I've never understood why sex of perpetrator isn't recorded. Is it recorded only for convictions,not at a first report, maybe?

That's being a victim of personal or intimate violence crime by sexuality (including from family - very sad to see such a higher prevalence of assaults from family for gay people)

Having a look and they've actually changed the way statistics are broken down since I last looked and published a report which confirmed that the stats were unreliable. Can't imagine it's an easy task though trying to get accurate statistics from 43 different police forces.

Yes Felas, I find it frustrating because they record sex of victim and have a breakdown of crime by sexuality (although not also sex) but there's no data on the sex of the perpetrators. I guess a way of finding out would be to look at convictions but we know that conviction rates for rape and domestic violence are low.

That's interesting oneflew. Wonder why that is.
I think records of gender of perpetrator would be useful as male victims may also be in a gay relationship.

I saw a Scottish study conducted by WA before, it said that when male victims were reinterviewed 12 months later 1/3 admitted they had made the offence up. I cannot find the research now, it's disappeared from WA website.

Only a third ? Well, I'm shocked - see.

I've done a bit of digging on the ONS stats in the past, and one thing that really skews the stats is the statistics around people murdered by a current or former partner.

For women, it only includes current and former partners, but for men it also includes being murdered by their partner's former partners as well. So in heterosexual relationships it's a man killing his ex's new boyfriend which to me is actually still a way to try to control their female ex.

Unfortunately the statistics don't show the sex of the perpetrators, but I'd be willing to bet that a lot of the violence against men is committed by men.

I think that's summarised in my second link buffy

Whoops forgot the link.
www.refuge.org.uk/about-domestic-violence/domestic-violence-and-gender/

There's this summary link from Refuge which seems to have some more detailed work linked within it.

I'd be interested in knowing what proportion of male victims are abused by male perpetrators, but haven't been able to find that statistic.

Here's another more recent analysis
www.statslife.org.uk/social-sciences/2308-official-stats-mask-extent-of-domestic-violence-in-the-uk

www.telegraph.co.uk/women/womens-life/11666990/Domestic-abuse-and-violence-is-not-gender-neutral.-Wake-up-Britain.html

This is a link criticising the way the ONS collect stats

www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/compendium/focusonviolentcrimeandsexualoffences/yearendingmarch2015/chapter1overviewofviolentcrimeandsexualoffences

Also note that sexual offences are counted separately from violence.

Here's a blog from the amazing Karen Ingala-Smith
kareningalasmith.com/2013/04/29/this-thing-about-male-victims/

I saw a Scottish study conducted by WA before, it said that when male victims were reinterviewed 12 months later 1/3 admitted they had made the offence up. I cannot find the research now, it's disappeared from WA website.

I think it's because they don't count repeat offences against the same victim. So they are counting how many victims there are. If you continue how many offences were committed against men v women it looks totally different. Basically men are far less likely to have repeated acts of DV against them.

I would want to know how many male victims are abused by perpetrators who themselves were born male, as opposed to born female.

That might be a start.

Then I'd want to break that down by category.

We all know that the category of dv most likely to end in death and very serious injury involves female victims of male aggressors.