Global Risk Assessment Device

1
Global Risk Assessment Device

• Licking County Juvenile Court
• Follow-Up Training

2
Follow-Up Training Agenda

• Part 1 The Data
• Part 2 Current Usage and Upgrades

3
Follow-Up Training Agenda

• Part 1 The Data
• Overall Sample Characteristics
• Overall GRAD Scores and Trends
• Breakdown by Demographics
• Age, Gender, and Household Composition
• Youth and Parent reports compared and contrasted
• Overall Referral Patterns

4
Follow-Up Training Agenda

• Part 2 Current Usage and Upgrades
• Special Recognitions
• New Cut-off Scores
• New Interpretation and Recommendation Pages
• New Parent Pages
• Multiple Referrals
• New custom features

5
The Overall Sample

• Who is in your data deck?
• 100 cases as of 8/5/03!
• Excellent job gathering
• this data so quickly!

6
Slight Reduction in the Numbers

• Our stated goal was to have all paired data
• A hallmark that makes Licking County truly
  unique!
• Out of the first 100 cases, however
• 3 cases were not paired
• For todays report
• 97 youth
• 97 adults

7
Data from the Youth

• Who are the 97 youth in your data deck?
• N 91 White, non-minority youth
• N 6 Minority youth
• Overall sample
• 94 White
• 4 African American
• 2 Bi-racial

8
Data from the Youth

• Average age 14.9 years
• Approximately one-half of the sample is 14, 15,
  or 16 years of age (Range 10-18 years)
• Females are 45 of the sample
• Average age 14.6 years
• Males are 55 of the sample
• Average age 15.1 years
• This age difference between males and females is
  not statistically significant however

9
Data on the Families

• What do the families of these youth look like?
• Household composition
• 34 stepfamily arrangement
• 29 single-parent mother-headed
• 27 married biological parents
• 5 single-parent father-headed
• 2 grandparent-headed
• 2 foster family
• 1 other

10
Data on the Families

• Where does these families live?
• Almost 80 come from 3 zip codes
• 43055
• 43056
• 43062

11
Mental Health Issues

• 39 of youth had prior experience with some form
  of mental health counseling
• 19 of youth had family members with prior mental
  health counseling experience

12
Overall GRAD scores(Youth Report)

• Overall Licking County sample scores (using the
  youth reports) are lower than the Franklin County
  scores used to set your original cut-offs (and
  lower than Cuyahoga County scores as well)
• In a short while, we will be talking about how
  the parent reports are remarkable similar to what
  the youth reported

13
The Cutoff Scores

• Therefore, keep in mind that Licking Countys
  moderate and high risk youth may have been
  labeled low and moderate at times due to the
  higher cut points

14
GRAD scores(Youth Report)

• Variation by demographic characteristics
• Effects of age
• Gender differences
• Effects of household composition

15
Age and GRAD scores(Youth Report)

• Youth divided into 3 groups
• 13 years and under
• 14 and 15 years of age
• 16 and older
• 2 significant differences
• Education/vocation risk
• Mental health

16
(No Transcript)
17
(No Transcript)
18
Take Note (Part 1)!

• Your youngest youth seem to be presenting with
  the greatest educational risk levels
• Youth with the greatest mental health risks may
  be those in the early years of high school

19
Gender and Grad Scores(Youth Report)

• One significant gender difference
• Females displaying significantly higher risk in
  the peer relationships domain
• General trends regarding gender and GRAD scores
  are similar for all other domains except
  substance abuse
• this trend suggests that males are using
  substances at higher levels than females

20
Take Note (Part 2)!

• Females on average may be presenting with greater
  risk levels than males
• This finding is consistent across counties
• In fact, the Cuyahoga and Franklin data decks
  display many of these differences as significant
  ones

21
Differences by Gender and Household Composition

• Because of the gender differences noted above,
  Male and Female youth data were analyzed
  separately regarding the potential effects of
  household composition

22
Household Data

• Household composition divided into 3 groups as
  follows
• married
• stepfamily
• single-parent and all other forms

23
Gender and Houshold

• For the females, there were no significant
  differences among the three household composition
  categories
• 2 domains approached significance
• Education
• Traumatic events

24
(No Transcript)
25
(No Transcript)
26
Gender and Household

• For the males, significant differences by
  household composition were found in the following
  domains
• Education
• Accountability
• Leisure
• Sociability
• Traumatic events
• In all cases, males most at risk resided in
  stepfamily arrangements

27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
So What Does This Mean?

• There is a pronounced
• Home by Gender
• Interaction in your sample!

33
Differences by Household Composition

• Further analysis conducted with the entire sample
  on all GRAD domains with Gender and Household
  Composition controlled
• Significant interactions found on
• Mental health
• Family
• Traumatic events
• Education
• Accountability

34
(No Transcript)
35
(No Transcript)
36
(No Transcript)
37
(No Transcript)
38
(No Transcript)
39
Take note (Part 3)!

• Overall, youth at greatest risk in Licking County
  tend to be females from single-parent families
• Males residing in stepfamilies often are another
  group tending to display higher risk levels as
  well

40
Use of the Parent Reports

• What makes the Licking County
• Data truly unique!

41
Parent and Youth Data

• The relationships between parent and youth
  reports are all highly correlated
• Pearson r ranged between .45 and .85
• Hence, parents and youth tend to give very
  similar reports on average

42
Parent and Youth Data

• At the same time, there are differences between
  parent and youth reports that are noteworthy
• Youth on average report higher risk levels than
  parents in all domains except
• Accountability
• Sociability

43
Take note (Part 4)!

• On the one hand, there seems to be some solid
  evidence to indicate that what the youth and
  parent report is very similar
• As we shall see in a bit, however, the difference
  between what youth and parents report is
  predictive of higher risk levels as well

44
Take note (Part 4)!

• On the other hand, youth tend to report higher
  levels of risk in most all domains (probably
  because they know more!)
• Parents report tend to report higher levels of
  risk in 2 domains associated with responsibility
  and social competence
• The evidence seems to point directly to the need
  to gather multiple family member viewpoints
  whenever possible!

45
Last but not least.. where are these youth
going?

• Analysis of data regarding where the youth and
  family were referred for services

46
Referred versus Non-referred

• 47 youth were referred to some type of service
• 50 youth received no referrals

47
Who are the Non-referred Youth?

• Both youth and parents report lower risk on the
  prior offense domain for the non-referred youth
• At the same time, however..
• Non-referred Youth report Higher risk on five
  domains
• Family/parenting Peers Health
• Sociability Mental health
• Non-referred Parent report Higher risk on 1
  domain
• Mental health

48
What is happening with these youth?

• Looking at low prior involvement in illegal
  behavior alone, a case can be made for the least
  amount of intervention.
• Left unaddressed is what happens to these youth
  and families when other risk domains are not
  taken into account.

49
Top 5 Most Utilized ServicesWhen Referral Made

• Community service (53)
• Individual Counseling (40)
• Urinalysis (38)
• Victim awareness (28)
• Victim apology (15)

50
Who Is Being Referred toCommunity Service?

• Significantly higher GRAD scores on the substance
  abuse domain as reported by both youth and
  parents
• No other significant differences

51
Who Is Being Referred toIndividual Counseling?

• Significantly higher scores on
• Education/vocation
• While not significant, the majority of other GRAD
  domains also trended toward higher risk levels,
  including the mental health domain
• Another important note the discrepancy between
  youth/parent reports on educational risk was
  significantly greater for those referred to
  individual counseling

52
More on mental health

• Having reported previous mental health counseling
  experience was not related to whether or not the
  youth was referred to individual counseling

53
Still more on mental health

• However, those youth with previous counseling
  experience also reported significantly higher
  risk in ALL domains except substance use and
  prior offenses
• And their parents also reported significantly
  higher risk levels in the same domains

54
Even more on mental health

• And finallythe discrepancy between youth/parent
  reports was significantly greater for those with
  previous counseling experience in 5 domains
• Mental health Accountability
• Family/parenting Health risks
• Education/vocation

55
Who Is Being Referred toUrinalysis?

• Higher scores on Substance Use
• As reported by both youth and parents
• Also, the discrepancy between youth/parent
  reports on substance use was significantly
  greater for those referred to urinalysis

56
Who Is Being Referred toVictim Awareness?

• Youth reports indicated significantly lower risk
  levels in the following domains
• Family
• Traumatic events
• Mental health

57
Who Is Being Referred forVictim Apology?

• Significantly higher scores from youth reports
  on
• Traumatic events
• Also, the discrepancy between youth/parent
  reports on both accountability and prior offenses
  was significantly lower for those asked to make
  an apology to their victim

58
Where We Are Going with the Data

• There are many more issues to be addressed in
  upcoming analyses, most importantly including the
  role that discrepancy scores play in the further
  classification of youth with the greatest
  risks/needs

59
Stay Tuned!
60
Follow-Up Training Agenda

• Part 2 Current Usage and Upgrades
• Special Recognitions
• New Cut-off Scores
• New Interpretation and Recommendation Pages
• New Parent Pages
• Multiple Referrals
• New custom features

61
Special Recognitions

• Special! Special! Special! thanks to
• Dave Vozzella
• Tara Lifland
• For their tireless work in service to norming the
  GRAD
• 51 and43 paired GRADs respectively!!!

62
Special Recognitions

• Other contributors to the GRAD database included
• Nikki Debo
• Deb DePaso
• Shannon Horton

63
New Upgrades to the GRAD Site

• Thank You for your patience and support during
  the initial pilot phase of this project!

64
New Features and Upgradesto the GRAD site

• We are in the process of upgrading and opening
  some basic non-customized features that had been
  suppressed to ensure a complete and timely data
  collection.
• New Cut-Off Scores
• Interpretations of Risk
• Treatment Recommendations
• Parent Pages
• Multiple Referral Feature

65
New Cut-Off Scores

• Your cut-off scores are uniformly going to drop.
  This means that
• It will take more identified concerns to have a
  youth score moderate risk and high risk in
  all of the domains

66
New Interpretation Pages

• 2 components
• 1. What does this domain actually measure?
• 2. Where is this youths risk score in
  comparison to all other youth?

67
This youth has scored HIGH RISK on the
Family/Parenting domain.
What does this mean?
68
Example High Risk Family/Parenting Page

• This means that this youth has scored higher than
  2/3 of the youth who have been assessed with the
  GRAD instrument on questions pertaining to
• Various forms of conflict within the home
• Evidence of verbal and physical aggression in the
  home
• Lack of appropriate supervision of the youth
• Use of inconsistent and/or inappropriate
  discipline methods
• Problematic family interaction and relationships
• Stressors related to basic family needs being met

69
You also have access to interpretations
of Moderate and Low Risk Levels
70
New Recommendation Pages

• What considerations do you make in determining
  referral/disposition?

71
This youth has scored HIGH RISK on the
Family/Parenting domain.
What do I do?
72
Example High Risk Family/Parenting

• The evidence that the family plays a major role
  in delinquency has prompted interest in treatment
  programs that include the family of the offender.
  The different types of family-based intervention
  that the juvenile justice professional should
  consider include the following
• Family therapy
• Home-based family-preservation
• In the event that it is not possible for the
  youth to remain in her or his home while
  treatment is taking place, a foster care option
  may need to be exercised.

73
You also have access to recommendations
for Moderate and Low Risk Levels
74
Parent Pages

• In addition to the interpretations and
  recommendations sections created for the worker
  we have also included three pages for parents
• 1. Talking to parents about their high risk youth
• 2. Talking to parents about moderate risk youth
• 3. Talking to parents about low risk youth

75
Parent Pages

• Includes Bell curve interpretation
• Provides tips on talking to parents
• Monitoring and supervision
• Communication
• Peers
• Leisure
• Seeking professional help
• Further assessment
• Open fields for the worker to mark areas they
  have discussed with the parent

76
Multiple Referrals Function

• Referral Page currently requires one referral to
  be made between GRADs
• Problems with this system of completing
  referrals
• The user conducting the assessment is not always
  the person who is making the referral
• Initial referrals are made to programs where the
  true recommendations will be made
• Result The loss of valuable referral
  information and/or the cluttering of the database
  with false GRADs

77
Multiple Referrals Function

• The new referral function will allow the user to
  enter multiple referrals between GRADs
• Where is this function located?
• User enters a youth ID number on the GRAD
  homepage
• User is asked to verify the information is
  correct
• The user is sent to the GRAD score page that asks
  the user
• Run another assessment
• Complete a referral

78
What do you see?
79
New Custom Features

• Aggregate reports
• Demographics
• Referral disposition

80
When will you see these features?

• Users will see these new features implemented
  over the next few weeks
• Directions will be provided on the website