Itac-24
..a substantial body of longitudinal research consistently points to a very small group of males who display high rates of antisocial behaviour across time and in diverse situations. The professional nomenclature may change but their faces remain the same as they drift through successive systems aimed at curbing their deviance: schools, juvenile justice programs, psychiatric treatment centres, and prisons.
Terrie Moffitt 1993 1
[We followed a group of] ..75 young offenders. All had conduct disorder and 68% had ADHD during the..school years. Between the ages of 6 and 30 the 75 offenders were sentenced for a total of 12 000 crimes, which allowing for [unreported and unsolved crimes] can be estimated to 1000 crimes per individual.
Darteg and Levander 1998 2
There is increasing support for the idea that children who exhibit both early hyperactivity-impulsivity-attention problems and conduct disorder may be the children who become lifelong persistent criminals. Early identification of these high-risk children would permit the implementation of interventions designed to prevent an ensuing lifelong pattern of antisocial behavior
Skilling, Quinsey and Craig 2001 3
The Risk Screen for Youth Offenders is a procedure for identifying teenage boys and young men who are at risk of poor adult outcomes, including criminal offending. The young men that we identify with this instrument are at risk of progressing to chronic antisocial behaviour, which can include victimising members of their family and the wider community4 . While this instrument can be used to identify teenagers headed towards a range of poor adult outcomes, its main use is as a risk assessment procedure for identifying young men at risk of progressing to adult crime. It can be used and interpreted in one to two hours.
The three quotations at the head of this short guide neatly sum up ten year’s of developments in our understanding of the life history of serious and chronic adult offenders. They also sum up the theory on which this instrument is firmly founded. In 1993, Terrie Moffitt was pointing out that in New Zealand, most chronic adult offenders are male, and they leave a characteristic trail of disordered and antisocial behaviour through childhood and adolescence. By 1998, Darter and Levander from Scandinavia had reported that their group of chronic adult offenders were once youth exhibiting recognisable behavioural disorders – attention deficit hyperactivity disorder (ADHD) and conduct disorder (CD). They also note their very high rate of offending - an average of one thousand offences each by the age of 30 years. By 2001 Skilling, Quinsey and Craig are pointing out that there is global data suggesting that the combination of ADHD and CD in child and adolescent males often progresses to chronic adult offending.
They also suggest that, if we know that, we ought to be able to recognise them and rehabilitate them before they become confirmed adult offenders. These new understandings open up the possibility of an exciting approach to crime prevention based on draining the pool of chronic adult offenders in society by shutting down the inflow of new recruits – by recognising and rehabilitating them before they become established adult offenders.
This is where this instrument comes in. It ‘recognises them’. This instrument is based on the studies quoted above, and many others, that have identified the life events that give rise to a sequence of behaviours that lead to antisocial adulthood. This instrument will have done its job if it focuses our attention and our resources on those young offenders who are headed towards an adult lifetime of victimising others, and other poor adult outcomes.
This instrument is intended to be used as part of a ‘diagnose and treat’ intervention. If this instrument suggests that intervention is appropriate, its use should lead to proven rehabilitative services that address each young person’s unique combination of current problems, or ‘needs’ as they are usually called. If there are no rehabilitation services available, or if they address the wrong problems, then this instrument is not a suitable aid to case management decisions. There is no point in diagnosing conditions that you cannot treat.
The ‘diagnose and treat the needs’ approach to managing teenagers experiencing problems is now firmly established as international best practice. For example, a recent review of Canadian youth services reported that:5
Our review firmly concludes that there is a great deal known about what constitutes effective interventions. Specifically, interventions reflecting appropriately targeted treatment with awareness of criminogenic risk and need (factors directly related to youth antisocial behaviour), and programs reporting high levels of program integrity, represent the most sound basis of intervention.
Interventions that work, therefore, are those that use risk and needs assessments to target research proven programmes (which must be delivered exactly as deigned) at the underlying causes of each individual’s offending.
Risks and needs
Since this instrument is all about ‘risks that contribute to re-offending’, and since the term ‘risk’ is used in different ways, we should make clear what we mean here. ‘Risk factors’ are life events that contribute to progress towards a disordered adulthood, like behavioural disorder in childhood, and drug dependence in adolescence. In general, the greater the number of risk factors showed by a young person, the higher the chance of a disordered adulthood6. All of the empirical information we have suggests that the same set of risks and needs are at work in all cultural groups7, including Maori8 .
Some risk factors are called ‘dynamic’, which means that they can be changed, now. Childhood behavioural disorder is not a dynamic risk factor, because when dealing with an antisocial adolescent, there is nothing that can be done about childhood behaviour. It is too late. Adolescent drug dependence, on the other hand, is a dynamic risk factor because, having identified it, it is possible to treat it successfully.
‘Criminogenic needs’ or just ‘needs’ are a particular set of dynamic risk factors. They are the dynamic risk factors (meaning that they are risks that can be changed) that are:
The core set of ‘needs’ that may appear in adolescence are:
The Risk Screen for Youth Offenders explores the behavioural history of the young person and determines whether the trajectory being followed is likely to lead to adult offending. Then, as a crosscheck, it estimates the overall risk of continued offending.
No use of protective factors
Readers familiar with the literature on risk assessment may have noticed that this instrument makes no use of protective factors. Protective factors are a group of life situations that reduce progress toward adult disorder, such as high IQ, and a strong relationship between the youth and some pro-social adult – a grandmother, perhaps. Interventions sometimes take the form of efforts to ‘pump up the protective factors’, and some approaches to risk assessment involve subtracting the protective factors score from the risk factors score. In addition to the protective factors, this instrument also does not use a wide range of powerful risk factors, such as mother’s age at the birth of the young person.9
Both protective factors and the wider set of risk factors are not used for the same reason – there is a more reliable and direct way to identify youth headed for chronic adult offending. Attempts to balance out the total set of risk and protective factors, and predict future behaviour from that balance can be subverted by the temperament of the youngster and/or mother (or other primary caregiver). Some children are ‘copers’ who survive the highest risk childhoods and the absence of any significant protective factors to become fully functional adults. Others show disturbed behaviour in response to a relatively small set of risks. Recent studies suggest that the varying degrees of resistance or susceptibility to risk factors is genetically determined. A determined caregiver who does not give up on a difficult child can also make a great difference to his or her adult outcomes. Overall, a young person’s future is not exclusively determined by the balance of risk and protective factors in his or her life.
The approach used here involves setting aside the complex balance of general risk factors, protective factors, and their interactions with the temperaments of child and caregiver, and simply collecting and interpreting information about the young person’s behaviour and functioning. If behaviour and functioning have been poor over a prolonged period, then the youth is at risk of poor adult outcomes, regardless of other risk and protective factors. A teenager’s future behaviour pattern is more likely to be a continuation of his or her past and present behaviour, and its accumulated consequences.
The Risk Screen for Youth Offenders uses 24 carefully chosen questions to explore the behavioural history of the young man or young woman that you are assessing, and to estimate his or her likely behavioural future. The 24 questions are put to two sources – some to the young person in question, and some to a ‘significant other’ - mother, a grandparent, or someone who knew the young person well as a child. The young offender is interviewed face-to-face, and the significant other by telephone. Significant others are sometimes hard to find but they are an essential part of the assessment procedure. Be prepared to try several sources if necessary for a reliable significant other.
The answers to the 24 questions are then combined in different ways and plotted on the scales on the last page of the instrument, which may be detached and filed. The pattern of results across the scales provides the information needed to assist your key case management decisions. Altogether, a trained and experienced user can collect the information, interpret the scores, make the key case management decisions, and file the one page summary in one or two hours. However, the instrument covers a lot of ground in those 24 questions so it is important that every question is asked properly and every answer interpreted properly.
The four scales are introduced in the Table below and discussed in the following sections.
Table one: The four scales
|
Scale number |
Purpose |
Number of questions |
Maximum score |
|
1 |
Checks the reliability of the answers you have been given |
6 |
6 |
|
2 |
Measures the extent of disruptive behaviours in early and middle childhood |
6 |
24 |
|
3 |
Measures the extent of delinquency in late childhood and early adolescence |
6 |
24 |
|
4 |
Measures the extent of adolescent offending |
6 |
24 |
Using scale one
Collecting information about young offenders is sometimes a risky business. Many offenders and their significant others who respond to your questions will provide factually correct information. Others may tell you what they think you want to hear, or provide you with their own version of history. A few will set out deliberately to mislead and to obscure past events. There is clear evidence from the trial conducted with this question set that, in families with a history of criminality, the significant other will attempt to minimise or deny the offender’s past behaviour problems.
Therefore, scale one deals with the first problem you have – deciding whether the information you have been given by your youth offender and significant other is accurate enough to base case management decisions on. There is no point in making key case management decisions on the basis of incorrect information. Scale one alerts you to the possibility that some of your information may be wrong, and gives you an estimate of how wrong.
Among the 24 questions there are six that are put to both offender and significant other. Each is a straightforward question, such as whether the young person was ever excluded from ever returning to a school, about which there ought to be no misunderstanding. On scale one you record the number of questions out of six that the two informants agreed on. If they agree on six or five, the rest of the information you collected is probably also correct or mostly correct. If they agree on four or less you are advised to collect more information before proceeding.
Our trial data suggests that in a third or less of cases, scale one will suggest to you that you do not have enough accurate information to proceed. Often, but not always, the reason for discrepancies may be a ‘colluding mother’ who is doing her best to paint a rosy picture of her offspring. Not unreasonable behaviour for a mother, but inconvenient when you are looking for accurate information about the progression of the young person’s antisocial behavior through time. In general, offenders’ responses are more reliable than their mothers’, but again, this is not always the case. Before you begin the telephone interview, it is important to impress three points on the significant other:
On the last page of the instrument summary page, scale one looks like Figure one below.
Figure one: Scale one on the final page of the instrument
If you are not satisfied with the quality of the data, you have a number of options. Interviewing a second significant other is a good option, and comparing the answers with those from the first significant other. This procedure may help you decide whether offender or significant other is the principal source of disagreement.
If disagreements persist, you may decide that your time and the rehabilitation resources you have could probably be invested more usefully in another case. This is not an invitation to drop difficult cases, because continuing disagreements between sources probably means that you are dealing with a high-risk young offender. It is a reminder that if you cannot get a confident diagnosis, you may not be able to line up an effective treatment.
Scale two: Childhood disruptive behaviour
Scale two measures early disruptive behaviour - behaviour characteristic of combinations of attention deficit hyperactivity disorder (ADHD), conduct disorder (CD) and oppositional defiant disorder (ODD). To score highly on scale two a young person must have shown behaviour characteristic of broadly based disorder, together with an early age of onset. Scale two is out of 24 points, and a score of above 12 is a positive identification of a childhood behavioural disorder that was sufficiently pervasive to affect normal development. The seven aspects of childhood behaviour that are examined in scale one include:
The question set in the Risk Screen for Youth Offenders was trialed using 150 teenage offenders drawn from both the youth justice and adult corrections systems. The distribution of trial scores for scale two is shown in Figure two, which also shows the cut-off between the scores of 12 and 13 points. Note that in Figure two the distribution of scores is ‘the tail end of a normal curve’ – indicating that scale two is sampling from a single population.
Some of the questions in the trial version have been replaced by improved questions here, so that the distribution of scores from this version may differ slightly from that shown in Figure two. The same applies to the distributions of scores for scales three and four, shown in Figures three and four below. These three figures will be replaced by new ones based on new data as soon as we have collected enough.
Scale three: Delinquency
Scale three measures progress to delinquency - whether by late childhood or early adolescence the young person was showing behaviour that was on the edge of criminal offending - truanting, running away from home, experimenting with drugs, shoplifting. Once again, the scale involves six questions and 24 points, one third of which are allocated for early onset. A score of more than 12 out of 24 is a positive identification. The six aspects of late childhood – early teenage behaviour that are assessed on the delinquency scale are:
Figure two: Scale two trial data: Distribution of childhood behaviour disorder scores
Figure three: Scale three trial data: Distribution of delinquent behaviour scores
Scale four: Teenage crime
Scale four measures the extent to which early delinquency progressed to adolescent criminal offending and associations with other offenders, including the age at which non-trivial offending began. To get a high score on this scale the young person would have committed a number of offences, had several contacts with the justice system, had friends who offend, and a pattern of regular drug or alcohol use. There are five questions in Scale four:
Note that ‘seriousness of current offence’ is not in any way a reliable indicator of risk or trajectory, and is not used in this instrument.
Figure four: Scale four trial data: Distribution of teenage offending scores
4: Interpreting the data
Behavioural trajectories
The youth offender’s scores on question sets two, three, and four are plotted on the corresponding three scales on the last page of the instrument. Draw a line from the point you have plotted on scale two to the point plotted on scale three, and from there to the point you plotted on scale four. These lines represent the youth offender’s antisocial behaviour trajectory from childhood into adolescence. Three general kinds of trajectories are shown in Figure five below.
Figure five: Trajectories across scales two, three and four
As Figure five above illustrates, children and teenagers who offend, both boys and girls, come in at least three kinds, each of whom requires a different response10. First, there are casual offenders – ‘kids who did something silly’, usually in association with several others. Once apprehended and sanctioned, they are not likely to offend again11 . Regardless of the severity of the offence, it is not good practice to spend time and other resources on casual offenders. Most of the children and young people who appear in the youth or adult justice systems are casual offenders. The trial of this instrument revealed than many teenagers currently in intensive, expensive rehabilitation programmes are casual offenders, often because their placement had been decided without the use of risk assessment.
The second group of youth offenders are the ‘adolescence limited offenders’, who mostly begin offending after the age of 13 years and mostly cease before early adulthood. In effect, they mature out of it12. If they are detected early, appropriate services can reduce the number of victims they will cause during their teenage years. However, if they are detected late – 17 years or older - they may be close to ‘ageing out’ of adolescence limited offending behaviour. Once again, scarce rehabilitative services could be used to better effect on the highest risk cases.
Finally there are those for whom this instrument is designed – boys and young men headed towards an adult lifetime of offending and other adult problem behaviours, which often include substance dependence, serial partnering, and spouse and child abuse. They identify themselves in several ways - by their trail of disruptive behaviours through childhood (high scores on scale two), the accumulating consequences of those disruptive behaviours (like use of cigarettes and alcohol, suspensions from school), their early-onset offending, and their associations with other youth offenders. They require intensive rehabilitation services directed squarely at their own individual set of dynamic risk factors, as described above. Heavy sanctions definitely do not work for this group and may well increase subsequent offending. It is very unlikely that their family, working unaided, will be able to bring about any significant change in their life course.
Scales two, three, and four make it possible to distinguish between these three general types of young offenders. Casual offenders, adolescence limited offenders and chronic offenders will show characteristic and different profiles across the three scales. Casual offenders will have low scores across all three scales. An adolescent limited offender will show rising scores across the three scales, and those headed towards chronic adult offending will show consistent, high scores on all scales. Figure six below illustrates a number of trajectories identified in the trial.
Sometimes, you will find that there is no recognisable pattern in the results on scales two, three and four. There are two possible reasons for this: It may be because of measurement errors - questions that were not answered honestly, or answers that were not interpreted correctly. Alternatively, it may be due to changing circumstances in the young person’s life resulting in changing behaviour patterns through time.
Note that scales two, three and four make use of ‘age of onset’ information. Recent research has shown that the earlier a particular form of problem behaviour appears, the more likely it is to persist and ultimately affect adult functioning13 . For example, conduct disorder that emerges before the age of 10 years, and ADHD that emerges before seven years indicate higher risk of progress to adult offending than later emergence14. Similarly, ‘appearance in an adult court for the first time at age 15 years’ indicates a higher risk of progress to adult crime than appearance for the first time at age 18 years
Figure six below provides further guidance about the implications of ‘age at first offence’. The data, which illustrates the percent of serious and chronic adolescent offenders who had committed their first non-trivial offence by ages from six to 16 years, is from an American study15. The graph shows the cumulative percentages by age at first criminal offence for youth who became chronic and serious adult offenders.
Figure six: Some representative trajectories from the trial
Click here to view the large version of this image.
Figure seven: Age at first self-reported offence for youth offenders headed toward adult crime
Age
As the graph shows, more than 50 percent of youth who ultimately progressed to adult crime had committed their first non-trivial offence before turning 12 years and almost 90 percent before turning 14 years. It follows that if you have a young offender with an offence record before the age of 14 years you are probably dealing with a young person who is likely to become a chronic adult offender. While it is unwise to base case management decisions on a single indicator, the data in Figure seven provides another source of advice that may assist case management decision-making. The data in Figure seven is American, because no New Zealand data is available, but it is likely that local data would show a similar relationship.
An example of the use of the Risk Screen for Youth Offenders
The summary page shown overleaf contains the results of an assessment conducted on a 15 year old boy in the trial. It provides us with the following case management advice:
Checking back on this young person’s scores on the instrument, which is not shown here, reveals that his ‘adolescent offending’ score resulted mainly from high scores on two questions – the drug and alcohol use profile, and the deviant friends question. While this young offender is not a priority for intensive rehabilitation services, he is probably going to engage in further adolescent offending. His assessment suggests that he may well justify an intermediate level of response, and in particular, services directed at supporting his caregivers in taking control of his use of drugs and alcohol and his associations with other offenders, both of which contribute strongly to adolescent offending.
Click here to view the large version of this image.
1. Moffitt T 1993 Adolescence-limited and life-cycle persistent antisocial behaviour: A developmental taxonomy Psychological Review n100 p 674-701
2. Dalteg A and S Levander 1998 Twelve thousand crimes by 75 boys: A 20 year follow-up study of childhood hyperactivity Journal of Forensic Psychiatry v9n1 pp39-57
3. Skilling T, V Quinsey and W Craig 2001 Evidence of a taxon underlying serious antisocial behaviour in boys Criminal Justice and Behavior v28n4 p450-470
4. See Pfiffner L Parent’s psychopathology and ADHD children Journal of Consulting and Clinical Psychology n67 p881-893 and Perry B et al 1995 Childhood trauma, the neurobiology of adaptation, and use-dependent development of the brain: How ‘states become traits.’ Infant Mental Health Journal v16n4
5. Leschield a et al 1993 Youth at risk: A review of Ontario young offenders programs and literature that supports effective intervention. www.lfcc.on.ca/risk1.htm
6. Lynam D 1996 Early identification of chronic offenders: Who is the fledgling psychopath? Psychological Bulletin v120n2 p209-34
7. See for example Kunitz S 1999 Alcohol dependence and conduct disorder among Navajo Indians Journal of Studies on Alcohol v60n2 pp159-67 and Kunitz S 1999 Risk factors for conduct disorder among Navajo Indian men and women Social Psychiatry and Psychiatric Epidemiology v34n4 pp180-89 and Loeber R and D Farrington 1997 Strategies and yields of longitudinal studies on antisocial behaviour in Stoff D et al Handbook of antisocial behaviour New York: John Wiley
8. Fergusson D et al 1993 Ethnicity, social background and young offending: A 14-year longitudinal study Australian and New Zealand Journal of Criminology n26 p155-170
9. As a group, socially isolated teenage mothers produce high-risk children with as much as ten times the average chance of progressing to a disordered adulthood. There are, of course, many teenage mothers who are exceptionally good caregivers.
10. Nagin et al 1995 Life-course trajectories of different types of offenders Criminology V33 p111-139 and Patterson G et al 1998 Variables that initiate and maintain an early-onset trajectory for juvenile offending Development and Psychopathology v10 p 531-547 and Fergusson D et al 2000 Offending trajectories in a New Zealand birth cohort Criminology v38n2 p525-551
11. At least one half of all young offenders appear in the justice system only once. So a good response is to sanction but not to rehabilitate, and get them out of the system quickly and cheaply, so that more resurces are available for managing high-risk cases.
12. Around one third of young offenders fall into this category.
13. See for example, Mason A and M Windle 2001 Delinquency risk as a function of number of early onset problem behaviours International Journal of Offender Therapy and Comparative Criminology v45n4 pp436-448.
14. See DSM-IV
15. Loeber and Stouthamer-Loeber 2002
Copyright © Department of Corrections | Feedback and queries email: webmaster@corrections.govt.nz