Appendix B

Table B-1

Effect Size for Various Sample Sizes and p Values

Total of ne + nc	p=0.05	p=0.01	p=0.001
50	0.56	0.75	0.95
100	0.33	0.53	0.68
200	0.23	0.37	0.47
300	0.19	0.30	0.38
400	0.16	0.26	0.33
500	0.17	0.23	0.29
800	0.14	0.18	0.23

Table B-2

Effect Size for Various Sample Sizes and Statistics

Total of ne + nc	t value	F value	x2 value
50	0.28 5 t	0.28 X (SQRT F)	0.28 X (SQRT x2)
100	0.20 5 t	0.20 X (SQRT F)	0.20 X (SQRT x2)
200	0.14 5 t	0.14 X (SQRT F)	0.14 X (SQRT x2)
300	0.12 5 t	0.12 X (SQRT F)	0.12 X (SQRT x2)
400	0.10 5 t	0.10 X (SQRT F)	0.10 X (SQRT x2)
500	0.09 5 t	0.09 X (SQRT F)	0.09 X (SQRT x2)
800	0.07 5 t	0.07 X (SQRT F)	0.07 X (SQRT x2)

Table B-3

Effect Size for Decreased Use Given in Percentages

Before use rate	Decrease of 5%	Decrease of 10%	Decrease of 15%	Decrease of 20%	Decrease of 25%	Decrease of 30%
10	0.36	>1	>1	>1	>1	>1
20	0.19	0.44	0.61	>1	>1	>1
30	0.15	0.32	0.51	0.76	1.12	>1
40	0.13	0.27	0.42	0.59	0.78	1.02
50	0.13	0.25	0.39	0.52	0.67	0.84
60	0.13	0.25	0.38	0.51	0.64	0.78
70	0.14	0.27	0.40	0.52	0.65	0.78
80	0.17	0.29	0.46	0.58	0.72	0.84

Reviewer Criteria for Determining Model Programs

The ratings for dimensions 1 through 9 range from 1 for very low quality to 5 for very high quality.

1. Theory葉he degree to which the project findings are based on a clear and well-articulated theory, clearly stated hypotheses, and clear operational relevance.

2. Fidelity of interventions葉he degree to which there is clear evidence of high-fidelity implementation, which may include dosage data.

3. Sampling strategy and implementation葉he quality of sampling design and implementation, and the strength of evidence concerning sample quality (e.g., data on attrition).

4. Measures葉he operational relevance and psychometric quality of measures used in the evaluation, and the quality of supporting evidence.

5. Data collection葉he quality of implementation of data collection (e.g., amount of missing data).

6. Analysis葉he appropriateness and technical adequacy of techniques of analysis, primarily statistical.

7. Plausible threats to validity葉he degree to which the evaluation design and implementation addresses and eliminates plausible alternative hypotheses concerning program effects. The degree to which the study design and implementation warrants strong causal attributions concerning program effects.

8. Integrity葉he overall level of confidence that the reviewer can place in project findings based on research design and implementation. This is the same integrity rating used by reviewers in the original review of the 37 projects.

9. Utility葉he overall usefulness of project findings for informing prevention theory and practice. This rating is anchored according to the following categories, and combines the strength of findings and the strength of evaluation.
   • The evaluation produced clear findings of null or negative effects for a program with well-articulated theory and program design; the study provides support for rejecting the program as a replication model.

   • The evaluation produced findings that were predominantly null or negative, though not uniform or definitive.

   • The evaluation produced ambiguous findings because of inconsistency in result or methods weaknesses that do not provide a strong basis for programmatic or theoretical contributions.

   • The evaluation produced positive findings that demonstrate the efficacy of the program in some areas, or support the efficacy of some components of the program.

   • The evaluation produced clear findings supporting the efficacy of well-articulated theory and program design; the study provides support for the program as a replication model.

Previous | Table of Contents

---

PRIVACY POLICY | SITE MAP | DISCLAIMER | ACCESSIBILITY

Last Updated: March 4, 2002