Taking more than one ineffective predictor doesn't magically make them better.
The article states there is no demonstrated link. That doesn't mean that it's not strong enough to detect on it's own, that means there is no link demonstrated.
Adding up a bunch of nothing doesn't make something. It just makes it easier to hand wave away the fact it doesn't actually mean anything.
Drug use has clear links with violence (although we don't know if it causes that violence, and sometimes we're talking about victims more likely to use drugs after violence). The article does nothing to explain why drug addiction is so strongly linked to violence in medical literature.
> Adding up a bunch of nothing doesn't make something. It just makes it easier to hand wave away the fact it doesn't actually mean anything.
There's a whole bunch of research showing this. Feel free to pick any one study and explain why it's wrong. That research is pretty clear: MI alone is a weak predictor; drug or alcohol addiction is a less weak predictor; previous violence is a stronger predictor; but if you have combinations of any two that's a stronger predictor than any one alone and if you have all three that's the strongest predictor.
The submitted article even quotes this increased risk from drug offenders:
> A 1997 survey of prisoners also indicated only 12% of federal drug offenders were ever convicted of a violent crime.
That's an increased risk of being a perpetrator of violence! 12% of the general population haven't been convicted of a violent crime.
According to Wikipedia the total number of adults under correctional supervision:
> In total, 6,899,000 adults were under correctional supervision (probation, parole, jail, or prison) in 2013 – about 2.8% of adults (1 in 35) in the U.S. resident population.
That's the total, which includes all non-violent crimes as well as violent crime.
• In Los Angeles, USA, 35% of methamphetamine users aged 18-25 years old were found to have committed violence while under the influence of the drug (7).
• In Memphis, USA, victims and family members believed that 92% of perpetrators of intimate partner violence had used drugs or alcohol during the day of the assault and 67% had used a combination of cocaine and alcohol (8). A study on intimate partner violence in China found that partners who used illicit drugsa were significantly more likely to abuse their spouses physically, sexually, or both (9).
• Results from the British Crime Survey 2007/08 showed that victims of violent crime believed the offender to be under the influence of drugs in 19% of incidents (10).
• In Australia, perpetrators of violence against nurses in emergency departments were perceived to be under the influence of drugs in 25% of cases (11).
• In Atlanta, USA, ecstacy users with higher levels of lifetime use exhibited higher rates of aggressive and violent behaviour (12).
• In Rhode Island, USA, a quarter of women arrested for intimate partner violence and referred by courts to intimate partner violence prevention programmes reported symptoms consistent with a drug-related diagnosis (13).
• In Canada, boys reporting sexual harassment perpetration were seven times more likely to use drugs and girls four times more likely to use drugs (14).
• In a study of violence in youth holiday resorts among young German, Spanish and British holidaymakers, the use of cocaine during the holiday was associated with triple the odds of involvement in fighting and use of cannabis with double the odds (15).
• In England and Wales, 12% of arrestees held for assault tested positive for cocaine use and 24% for opiate use (excluding methadone) (16).
---end quote---
---begin quote---
Psychiatric factors:
There are elevated levels of psychiatric conditions, particularly Post Traumatic Stress Disorder (PTSD), in drug users experiencing and perpetrating violence. For example, high rates of intimate partner violence have been found among women with both drug use and PTSD (80), while the presence of both cocaine dependence and PTSD is associated with increased perpetration of partner violence (81). Furthermore, psychological distress and PTSD associated with experiencing rape and physical assault are related to greater severity of drug use
The article states there is no demonstrated link. That doesn't mean that it's not strong enough to detect on it's own, that means there is no link demonstrated.
Adding up a bunch of nothing doesn't make something. It just makes it easier to hand wave away the fact it doesn't actually mean anything.