Domestic violence - a statistical 'shock and awe' campaign?

An old statistical joke goes like this:

Apparently, one in five people in the world are Chinese. There are five people in my family. Mum isn’t Chinese and I’m sure dad isn’t. It could be my older brother Colin. Or it could be my younger brother Chang Fu Yat. Frankly, I think it must be Colin.

Pick up almost any newspaper on any given day and you will most likely find a by-line claiming: "Statistics show …"; "new survey finds …"; or, "new study proves …". Often accompanied by embellishments such as "shocking”, “appalling”, and so on.

Nowhere is this more prevalent than on the subject of gender relations and in particular the emotionally charged subject of domestic violence. (The new euphemism is "family violence" - but only lip service is paid to even the concept of male victims.)

The widely cited statistic on the subject is the Women’s Safety Survey (WSS), published in 1996 that claims, "One in four Australian women (25 per cent) experience domestic violence within their lifetime". This survey finding underpins government policy and legislation in every Australian state jurisdiction (with the exception of Victoria, which now evidently claims that "one in five women are victims of domestic violence", apparently suggesting that women would be much safer if they domicile in Victoria).

No "study" is of much value until it has been subjected to peer review. This has never occurred in relation to the Women’s Safety Survey. For a number of reasons, there is an urgent need to now do so.

The WSS study was released under the imprimatur of the Australian Bureau of Statistics (ABS) but was in fact a creature of the then bureaucratically powerful Office of Status of Women (OSW), which commissioned the survey. There was significant consternation reported at that time in relation to complaints by ABS officers that they were being "bullied" into undertaking unprofessional, and methodologically flawed "advocacy research"- research which is designed to prove the existence of something, whether it exists or not.

For example, the notion that one in four women are suffering from domestic violence is an alarming figure and conjures images, at the very least, of black eyes and bruises occurring on an appalling scale. How many Australian’s would know that the survey findings included such largely irrelevant questions as “Have you ever received an obscene phone call?” Or that the survey report obfuscated the fact that some 27 per cent of respondents were actually reporting violence caused by other women?

There were many other seriously disturbing aspects to this survey. Again, for example, it also involved only voluntary participation, which is a key source of survey bias as it attracts participants who may have an untoward fascination with the subject matter, a factor that can dramatically skew the results.

Similarly, the survey was a "life incidence" survey, thus inviting the recitation of some event far off in both time and in memory. The failings of human memory and perception with the passage of time are well recognised by our legal system, which, with very few exceptions, refuses to admit evidence that has been expired by time. Twenty years and a bitter divorce can change a memory from someone merely "pushing away" into "he threw me down the stairs".

The law recognises the frailty of old memories but our ever -increasing victim culture does not. Society would not entertain the concept that someone is currently a "road accident victim" based on a minor and non-permanent injury they had incurred from a vehicle accident 20 years previously. Nor would we necessarily put much faith in a 20-year-old version of how the accident occurred. Yet this is precisely what such surveys on domestic violence increasingly attempt to encourage society to accept as reality.

When citing the "one in four" statistic, some domestic violence literature oddly leaves out the phrase "within their lifetime", giving a false impression of immediacy (that one in four women are victims - right now, on this very day). Moreover, the Women’s Safety Survey did not actually purport to say that one in four women were victims of "physical" domestic violence, but included a range of other non-physical (and potentially uneventful) behaviours that were then (re)-classified as "domestic violence".