Ohio Alternative Response Project: Evaluation Findings Presentation to the Ohio Alternative Response Symposium May 13-14, 2010 Tony Loman, Ph.D. and Christine Filonow, MSW Institute of Applied Research St. Louis, Missouri www.iarstl.org Pilot Counties A S H TA B U LA LA K E LU C A S F U L TO N W I LLI A M S O TTA W A GE AUGA C U YA H O G A W OOD HENRY D E FI A N C E ERIE
SA NDUS KY TR U M B U LL LO R A I N PU TN A M PO R TA G E HU RON SE NECA PA U LD I N G ME DINA S U MMI T MA HONIN G HANCOCK VA N W ERT C R A W FO R D W YA N D O T A S H LA N D W A YN E
S TA R K R I C H LA N D A LLE N C O LU M B I A N A HARDIN ME R C E R MA R ION A U G LA I Z E C A R R O LL H O LM E S MOR R OW TU S C A R A W A S KNOX LO G A N S H E LB Y UNION J E FFE R S O N C O S H O C TO N D E LA W A R E
HARRISON C H A M PA I G N DARKE LI C K I N G MI A MI GUERNSEY B E LM O N T FR A N K LI N C LA R K PR E B LE MU S K I N GU M MA DIS ON M O N TG O M E R Y FA I R FI E LD N O B LE PE R R Y MONR OE GREENE PI C K A W A Y
MOR GA N F A YE TTE HOCKING B U TL E R W ARREN W A S H I N G TO N C LI N TO N ROSS A TH E N S V I N TO N H A M I L TO N H I G H LA N D ME IGS PI K E C LE R M O N T JACKSON BROW N ADA MS G A LLI A S C I O TO
LA W R E N C E Time Line for Evaluation 2008 2009 2010 Jn Fb Mr Ap My Jn Ju Au Sp Oc Nv Dc Jn Fb Mr Ap My Jn Ju Au Sp Oc Nv Dc Jn Fb Mr Ap Preparati Data collection Analysis/ on Report Random assignment SACWIS and manual/email methods Monthly case-specific data from workers Family Surveys Cost study site visits Site visits Community & worker surveys Site visits
Cost study Community & worker surveys Site visits Reception/Analysis of pathway tools (c. 10,000) Reception/analysis of service plans Pathway Assignment and Random Assignment Screened out Child Maltreatment Reports A Initial Screening For CPS Inappropriate for AR TR Traditional Investigations The Evaluation Control Group B
Accepted Report AR Pathway Assignment (Screened in) AR-appropriate Estimated 51.7 percent of reports determined to be ARappropriate C Pool of Reports Eligible for AR D Random E Outcome/Impact Assignment Control Cases Analysis Experimental Group Cases offered AR AR Family Assessments
Final study group included 4,529 families (50.5% Exp. and 49.5% control Alternative Response: What is Different? Removing the negative: No formal victims, perpetrators No substantiation or indication of abuse/neglect No entry of adults into a central registry of abusers Enhancing the Positive Establish child safety (create a safety plan) Engage the family Focus on broader family needs Emphasize family participation in decisions Changes under AR: Negative Emotional Response of Families Discouraged Pessimistic (ns)
Positive Emotional Response of Families Optimistic (0.0760) Encouraged (0.0001) Positive (0.0090) Gratef ul (0.0090) Experimental Reassured (0.0010) Control Comf orted (0.0490) Thankf ul (0.0020) Pleased (0.0150) Helped (0.0470) Satisf ied (ns) Hopef ul (0.0001) Relieved (0.0270) 0% 5% 10% 15% 20% 25% 30%
35% Changes under AR: Satisfaction with Worker Experimental Control Very 0% Satsified 10% 58.4% 49.5% Generally Satisfed 20% 30% 40% 33.9% 39.2% Generally Dissatisfied 50% 60% 70%
3.7% 4.0% 6.8% 4.5% Very 80% Dissatisfied 90% 100% Changes under AR: Involvement in Decision Making 17.4% No decisions were made Not at all A little 24.2% 5.1% 5.0% Experimental Control 7.5% 9.0% 15.8% 20.7%
Somewhat 54.3% A great deal 0.0% 41.1% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% Family Characteristics Single Parent Families
Over two-thirds of families reported incomes of $15,000 or less compared to 8% for Ohio generally. Compared with statewide statistics, a higher proportion of family caregivers in the study had less than a high school diploma (31 percent) than in the general population of Ohio (13 percent). The caregiver in over six in ten families was unemployed. Only 12 percent were employed full time. Characteristics: Welfare and Other Support Retirement check Unemployment benefits Housing assistance TANF Social security disability Utilities assistance Child support WIC School breakfast/lunch Food stamps 1.4% 1.1% 1.6% 10.2% 10.4% 10.0% 17.2% 17.6% 16.8% 23.4%
Workers indicated these needs among the families they encountered but also emphasized problems in family functioning, especially parent-child relationships and parenting skills. Perhaps, consonant with this: Nearly half of caregivers indicated problems in childrens behaviorespecially uncontrollable and aggressive behaviorand childhood depression and anxiety. About one-third of family caregivers surveyed reported that their children had problems in school. Factors in Service Provision Family need for services Family engagement Time devoted to cases Funding availability Service/resource availability and worker knowledge Agency patterns of resource use POVERTY-RELATED SERVICES======= Food or clothing for your family** Help paying utilities*** Other financial help*** Car repair or transportation**
Housing Money to pay rent** Appliances or furniture* MEDICAL OR WELFARE SERVICES==== Medical or dental services** Welfare/public assistance EMPLOYMENT SERVICES=========== Help getting into educational classes Job training or vocational training Employment help LEGAL, CHILD RELATED, DIRECT Legal services Child care or day care Assistance in home Respite care Meetings with other parents PARENTING, COUNSELING TREATMENT===== Parenting classes Mental health services* Counseling services*** Alcohol or drug treatment Help for a family member with a disability Family Reports : Level and Types of Services Receive d
Experimental Control * Statistical Trend (p < .10) **Significant (p < .05) *** Significant (p < .01) 0% 2% 4% 6% 8% 10% 12% 14% 16% 18% 20% Family Reports: Level and Type of Services Poverty-related services increased: AR workers more often provided referrals for or helped families to receive food and clothing, help with utilities, other financial help, car repair and transportation, money to pay rent or help in obtaining appliances and furniture. Experimental families under AR also reported receiving more referrals to traditional counseling and mental health services. No difference was found in the number of services or the provision of direct services between Caucasian and African-American families under AR. Family Reports:
Satisfaction with Help Offered or Received 11.1% No Services Offered 20.7% 4.5% 2.6% Very Dissatisfied Experimental (AR Family Assessment) Control (Traditional Investigation) 3.2% 5.6% Generally Dissatisfied 33.7% 36.7% Generally Satisfed 47.5% Very Satsified 34.4%
0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% Worker Reports of Services Provided Workers reported providing more services, support and assistance under AR and more information about where services could be found. Workers indicated that basic poverty-related
services were provided significantly more often to experimental families, such as rent payments, housing services, help with basic household needs, emergency food, and transportation. Worker Reports of Services Provided Under AR, 46.7 percent of AR workers said they were responsible for directly providing or connecting families to resources and services, while only 26.3 percent of TR workers reported this. Correspondingly, AR workers indicated they provided only information & referral for 41.2 percent of the services compared to 59.2 percent for TR workers. AR workers directly assisted with 83.3 percent of services in the category help with rent or house payments compared to 30.0 percent for TR workers. Most Frequent Types of Services Listed in Service Plans County (# plans reviewed) Clark (72) Fairfield (69)
Franklin (214) Greene (104) Guernsey (43) Licking (54) Lucas (107) 1st Frequent Service Mental health/ counseling Mental health/ counseling (in-house worker) Beds, other household items Mental health/ counseling 2nd Frequent Service 35% 55% 26% Appliances, furniture, linens Benefit assistance/ budgeting (in-house worker) Settlement house
23% Baby items/ household items 20% 53% Clothing or food voucher 32% Utility assistance 26% 47% Housing assistance 29% Mental health/ counseling 24% Ross Unable to review
Trumbull (136) Tuscarawas (17) Mental health/ counseling/ therapy Case-management by PCSA Family and Worker Reports of Services Provided Experimental families were also somewhat more likely than control families to indicate that the services received were enough to really help. According to workers, AR families were also more likely to participate in services than control families. Changes under AR: Level of Contacts of Workers with Families 5.0 4.6 4.5
How many face to How many How many other How many contacts How many face to face meetings did telephone contacts contacts did you did you have with face contacts did you have with did you have with have with a family others on behalf of other service members of the members of the member (ns) this family (p = .008) providers have with family (p = .003) family (p = .012) a family member (p = .02) The average number of days until case close was 53.6 for experimental families and 44.7 days for control families. What Changed? Operational Shifts Longer assessment period Alternative Response workers provided post-assessment services
Family Service Plans Flexible funding Service partnerships What Changed? Practice Shifts Removal of barriers to family engagement Approach is less incident-driven Increased communication and trust Allows for more preventative support, services, and follow-up With alternative response, Im not there to gather information to make a decision; Im there to help the family come to a resolution. What Changed? Service Shifts Factors: Family need for services Family engagement Time devoted to cases Funding availability Service/resource availability and worker knowledge Agency patterns of resource use Worker and Supervisor
Responses Workers and supervisors that performed work related to AR reported observable adjustments in their approach and practice, indicating that AR was implemented as intended and produced positive changes within the agency. Workers reported feeling more able to intervene effectively with AR families than with non-AR families. Reactions of AR families to assistance were seen as more positive by workers than the reactions of other families. Worker and Supervisor Responses The majority of staff involved with AR stated that the pilot had affected their approach to families a great deal or in a few important ways. AR-involved staff saw AR as leading to a more friendly approach to families, more family participation in decisions and case planning, and more cooperation
from families in the assessment process. A strong minority (38.9 percent) of county staff involved with the pilot reported that AR had increased the likelihood that they will remain in the field of child welfare. Worker and Supervisor Comments Alternative response has reminded me how blame driven we could be. A lot of the time, it doesnt matter whose fault it is, as long as the family is willing to work to function better. I have always tried to address concerns without labeling people, but the traditional approach makes this difficult. Now with the alternative response approach it is much easier to focus on the problems and solutions with the family. Our alternative response team is outstanding. From the start to the close of an AR case, each worker brings an attitude of success which is passed on to the family. The ability to be creative in how families are assisted has been critical. Community Responses Familiarity with AR among stakeholders had increased
by the end of the pilot, from 45.3 percent in 2008 to 68.3 percent in 2009. Attitudes toward AR were highly positive among those who were familiar, although a little less than half of all survey respondents were unsure of their opinion. Nine out of ten judges or magistrates in the pilot counties reported being at least somewhat familiar with the AR pilot. Those nine also perceived that AR had the potential to lower the number of cases coming to court to some degree. Child Safety Short-term child safety from the time of the original report until final contact with families was examined. Child safety problems were identified by workers in a minority of families, 33.2 percent of control cases and 25.4 percent of experimental cases. When a child safety problem was identified, no statistically significant difference was found between experimental and control families in the extent of improvement or decline in safety. There was no evidence that replacement of traditional investigations by AR family assessments reduced the safety of the children.
New Reports of Child Abuse and Neglect Among families entering the study during the first 360 days, 13.3 percent of control families had a new report compared to 11.2 percent of experimental families. This difference was statistically significant. A proportional hazards analysis that controlled for levels of past reporting on families also confirmed that experimental families that were served through the AR family assessment pathway had fewer new reports than control families that were approached through a TR investigative assessment. New Report s of Child Abuse and Neglec t Experimental Control New Reports of Child Abuse and Neglect
Racial differences in later accepted reports were also examined. Although study families as a whole were largely in poverty, African-American families were significantly and substantially more impoverished than Caucasian families. Race was taken as a proxy measure for poverty. Analyses demonstrated that the major positive effects of AR on new reporting of child maltreatment at this point in tracking families appears to have occurred among African-American families. This was interpreted to mean that AR has its greatest effects among the poorest families in the population. Subsequent Removal and Placement of Children Differences in out-of-home placement were also examined in the evaluation. Within the control group 3.7 percent of children had been removed while 1.8 percent had been removed in the experimental group, a significant difference.
This difference also remained significant in the stronger proportional hazards analysis. AR appeared to reduce the number of child removals and out-of-home placements. Cost Analysis: Indirect Costs Indirect costs were calculated using cost allocation data and average time that workers spent with experimental and control families. AR family assessments averaged $940 per family compared to $732 per family for TR investigations. Reflecting increased worker time with families, AR was more expensive in the immediate term. For subsequent work, experimental families averaged $145 per family compared to $266 for control families. Total costs for control families averaged about $999 per family compared to $1,084 for experimental families. At this point in the follow-up, experimental families were slightly more expensive ($85 per family) overall in indirect costs than control families. Cost Analysis: Mean Indirect Cost per Family Cost Analysis: Mean Direct Cost per Family
Cost Analysis: Total Direct and Indirect Costs, Initial and Subsequent to AR Control Initial direct costs Experimental $ 99 $194 Initial indirect costs $ 732 $ 940 Subsequent direct costs $136 $ 48 Subsequent indirect costs $ 266 $ 143 Total Initial Costs
$ 831 $ 1,134 Total Subsequent Costs $402 $191 $1,233 $1,325 Total Costs Quotes from AR Families The social worker was fantastic. She did not come to our home with predetermined ideas, but came to conclusions based on our family and our home. I was surprised by how much help was offered. I didnt know they offered you all that extra help. It was appreciated. The caseworker treated us with respect and made us feel like we mattered and that we had our own voice to speak. We enjoyed our case worker coming and explaining things to us and made us feel wonderful. My caseworker was awesome. She saw I wasnt a bad mother. I just needed a little help to get back on the right track. And I love her for that.
Issues in Diversity Gender Language Groups Race Ethnicity Rural / Urban Special Needs physical disabilities mental disabilities Ethnic & Racial Diversity in the United States In 1950 minorities made up 11% of the U.S. population. This figure increased to 30%...
Introducing Game Strategies to Primary School Children. Presented by Steve Clogstoun. PE and Health Co-Ordinator Willmott Park Primary School. To download this information please visit. www.acephysed.wordpress.com
Analysis of RT distributions with R Emil Ratko-Dehnert WS 2010/ 2011 Session 08 - 11.01.2011 * * * * * * Kuiper is in package CircStats Do not use ks.test for normality checks unless you know the parameters of the...
"In science, the credit goes to the man who convinces the world, not to the man to whom the idea first occurs." --Sir William Osler Scientific Writing, HRP 214 More on that/which "Careful writers, watchful for small conveniences, go which-hunting,...
Hazardous Waste Management Spill Response Procedures Spill Response Procedures Workers must be familiar will VAMC Integrated Contingency Plan (ICP) Incidental spill V.S. Non-incidental spill Report ALL spills to your supervisor, No matter what size it is Incidental Spills Small quantity...
Objectives OM Strategy Competitive Dimensions Order Qualifiers and Winners Strategy Design Process Productivity Measures What is Operations and Supply Strategy? Operations and supply strategy is concerned with setting broad policies and plan for using the resources of a firm to...