MAKING NUMBERS MAKE SENSE

1
MAKING NUMBERS MAKE SENSEJim Adams-Berger,
Ph.D.OMNI Research and Training, Inc.Summer
1999
Using Data to Identify Community
Problems Explore Problem Relationships, and
Choose Responsive Programming
2
Our Overarching Goals

• Be Helpful
• Be Simple

3
DATA
DATA
DATA
DATA
DATA
DATA
DATA
DATA
DATA
DATA
DATA
DATA
DATA
DATA
DATA
DATA
DATA
4
The Major Analytic Elements
Beliefs Data Research
Values
5
Values As a Filter
Beliefs Data Research
Beliefs Data Research
Values
6
Training Framework
Selecting Programs
7
Surfacing And Identifying Problems

• Use data to surface potential problems
• Use values and discussion to weigh the relative
  importance of identified Problems
• Prioritize a set of problems for further
  investigation

8
Example Problems

• Teen Pregnancy
• Sexually Transmitted Diseases
• Child Neglect
• School Drop Out

9
Modeling Problem Relationships

• Use data to uncover the structure of the
  Problem(s) considering possible causes,
  inter-relationships and changes over time
• Use existing theory and research to further
  explore the structure of the problem
• Use values to explore agreement with formal and
  informal sources of information
• Negotiate and develop a model of the Problems
  showing causes and associations

10
Modeling Tools
11
A Problem Model
12
Selecting Programs

• Use research on promising programs to explore the
  various programs that may be responsive to the
  identified problems and the developed model
• Use values to determine which of the potentially
  responsive programs may work best in the
  community
• Choose one or more promising programs

13
Choosing a Program
14
Surfacing and IdentifyingProblem(s)

• Using Numbers and Rates

15
Surfacing and IdentifyingProblem(s)

• Purpose
• To teach participants basic data interpretation
  skill and to use these skills with a data set to
  derive a prioritized set of problems

16
What Do Numbers Say?
17
What Do Numbers Say?
18
Why Are Numbers Limited?

• Their Meaning is Often Only Understood In
  Relation to Population Figures
• Because of This, Numbers May Distort Our
  Understanding of Problems

19
Instead, Use Rates
A Rate is a Simple Calculation That 1.
Relates a Number to a Relevant Population
Figure 2. And Then, Multiplies it by a Large
Number to Make it More Interpretable
20
Why Do We Do This?
Because Rates Allow Us To Make Numeric
Comparisons Between Places and/or Over Time Even
When Population Figures Are Different!
21
How Do We Calculate a Rate?
We Need Three Pieces of Information The Number
(e.g., Aggravated Assaults) The Population
Figure The Per Number We Want to Use To Make
the Rate Easy To Interpret
22
How Do We Calculate a Rate?
The Simple Formula is Number Per
Number Population
23
Calculating Rates
Choose 10,000 as the Per Figure
24
Calculating Rates
8 Assaults Per 10,000
25
Calculating Rates
26
So, How Do Our Numbers and Rates Look Different?
27
Number
Rate
28
Are Rates All You Need To Know?
Rate Per 10,000
29
Are Rates All You Need To Know?
Rate Per 10,000
30
Modeling ProblemRelationships

• Graphic Models and
• Analytic Tools

31
Modeling ProblemRelationships

• Purpose
• To teach participants more advanced
  interpretation skill and to use these to create
  and model their understanding of one or more
  problems

32
A Modeling Example
33
Modeling Tools
Cause Arrow - X Causes Y
34
Modeling Tools
Modeling Tools
Cause Arrow - X Causes Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
35
Modeling Tools
Modeling Tools
Association Line - X and Y Happen Together
36
Modeling Example
Teen Pregnancy
Sexually Transmitted Diseases
KEY BBelief DData RResearch
37
Modeling Tools
Line Arguments
B Is a BELIEF D Is Implied by the DATA R
Has RESEARCH or Theory to Support
38
Modeling Example
Teen Pregnancy
Relaxed Values
D
B
Sexually Transmitted Diseases
Not Enough To Do
KEY BBelief DData RResearch
39
Modeling Example
R
Relaxed Values
School Drop Out
B
Not Enough To Do
KEY BBelief DData RResearch
40
Modeling Example
R
Relaxed Values
School Drop Out
B
Not Enough To Do
KEY BBelief DData RResearch
41
Choosing AProgram

• Matching a Problem Model
• to
• A Promising Approach

42
Choosing AProgram

• Purpose
• To use the Promising Program literature to choose
  the most effective program choice

43
Choosing a Program
Teen Pregnancy
R
Relaxed Values
School Drop Out
D
B
Sexually Transmitted Diseases
Not Enough To Do
Choose a Leverage Point
Choose an Outcome
44
Choosing a Program
Teen Pregnancy
R
Relaxed Values
School Drop Out
D
B
Sexually Transmitted Diseases
Not Enough To Do
Choose a Leverage Point
Choose an Outcome
45
TheEnd