Copyshop suicide

Photo by Flickr user just.Luc. Click for sourceBad Science has a great article on the ‘copycat suicide’ effect, where media reporting of suicide can increase the chances of suicide in other people.

Copycat suicide is sometimes called the ‘Werther Effect’, after Goethe published his 1774 novel ‘The Sorrows of Young Werther’ which depicted Werther’s suicide and was reportedly followed by people imitating the same method to end their lives.

It’s an interesting effect because it shows the influence on the media on what people usually think of the most extreme of decisions.

An excellent 2003 review article on the subject found that the effect holds for all media reports of suicides (including fictional ones) but celebrity suicide is most associated with subsequent deaths. Interestingly, it notes that the largest known increase followed the death of Marilyn Monroe.

The review also found found that the greater the coverage of the suicide, and the more details in the reporting, the larger the increase in subsequent deaths.

Because of this, there are now media guidelines for reporting suicide, and the Bad Science article reports on a particularly bad example where the journalist reported exactly the sort of thing most associated with increased risk in a single story – virtually nothing except details of the suicide method.

One of the most interesting bits of the Bad Science piece doesn’t appear in the print version. However, it discusses research that found the majority of people who attempt suicide and survive are pleased they did some years later:

There is a literature which I think is extremely powerful, and yet unanimously ignored by mainstream media, and that is the follow-up data on what happens later in life to people who have felt so suicidal that they have made serious attempts on their own lives.

In extremis Pajonk et al followed up a large number of people who they picked up in intensive care after very serious suicide attempts. Amongst those who survived, and did not have serious psychotic illnesses, six years later, the majority were happy and well, living productive family lives, and were – we might reasonably interpolate Рglad to be alive.

Link to Bad Science article on media reporting of suicide.
Link to review article on media and suicide (with open-access link).

3 Comments

  1. Posted March 28, 2009 at 4:22 pm | Permalink

    That is important and I wish it would be reported more often. Is there a similar copycat effect with other kinds of reporting, for eg acts of kindness?

  2. StunnedMullet
    Posted March 30, 2009 at 2:54 am | Permalink

    Yup. That’s Bad Science alright.
    Unusual for you Vaughan, usually you pick up on bad science fairly fast. I would probably be worth contemplate what it is about suicide that makes it so hard for people to reason about it objectively.
    So in what way is it Bad Science?
    The title of the article is…”Long-term follow-up after severe suicide attempt by multiple blunt trauma ”
    Well, obviously that not just any sort of suicide, those are jumpers. ie. Preselected for..
    * Impulsive (requires no preplanning)
    * Urban (requires tall buildings)
    * Not (too) afraid of heights. (They actually jumped)
    Well, lets move past that on to what else is special about this group… ie. They were so smashed up they were in intensive care, but not so smashed up they died. ie. They have selected for the _maximum_ possible pain, disfigurement and disability short of death. And then you wonder why these guys didn’t try _that_ again! Oo! They suicides, they must be incapable of learning! :-)
    Their “Good Outcomes” can be caricatured as “ceased irritating society”. If they thought for a moment they’d realise a “Good Outcome” would be “they liked life and no longer wish they were dead”.
    “Predictive variables for an adverse outcome (10 patients, 28%) were found to be a diagnosis of schizophrenia, continued psychiatric treatment and being without employment.”
    Predictive? Did they really mean that? Wait a minute, what are they measuring and predicting here. A Good outcome was “good psychosocial ability”. ie. Employable.
    And how about that classic line… “continued psychiatric treatment predictive of adverse outcome.” Well we had better stop that stuff then!
    And as for that “we might reasonably interpolate – glad to be alive”, that’s way too large a leap to conclude from the data.

  3. StunnedMullet
    Posted March 30, 2009 at 3:59 am | Permalink

    Oh, actually the whole damn article is Bad Science now I look closer at it.
    Are the copy cats suiciding _because_ of the reporting or are the planning on it anyway but are unsure of how (most methods are painful and risky)…. and then someone proves a method so they copy that.
    ie. You getting a peak in the stats because they have been shown how, but would have eventually worked it out anyway. Or perhaps tried something that didn’t quite work and totally screwed up their life, but won’t appear as a stat.
    Bad Bad Science! Spank!


Post a Comment

Required fields are marked *
*
*

Follow

Get every new post delivered to your Inbox.

Join 2,590 other followers