Table 3

Percent of studies with specific threats to validity of homeopathic and conventional interventions.

Statistical Conclusion Validity
Homoeopathic
Conventional

Low Statistical Power
6.8
24.0
Violated Assumptions of Statistical Tests
18.6
14.0
Fishing/Error Rate Problem
6.8
24.0
Reliability of Measures
100.0
94.0
Reliability of Treatment Implementation
16.9
0.0
Random Irrelevancies in Experimental Setting
8.5
0.0
Random Heterogeneity of Subjects
57.6
52.0

Internal Validity



History
3.4
2.0
Maturation
6.8
2.0
Testing
1.7
0.0
Instrumentation
0.0
0.0
Statistical Regression
3.4
0.0
Selection
100.0
70.0
Attrition
76.3
0.0
Interaction with Selection
20.3
2.0
Ambiguity of Direction of Causality
0.0
2.0
Diffusion of Treatment
1.7
0.0
Compensatory Equalization of Treatments
0.0
0.0
Compensatory Rivalry
0.0
0.0
Resentful Demoralization
1.7
0.0

Construct Validity



Inadequate Pre-operational Explication of Constructs
22.0
4.0
Mono-Operation Bias
37.3
92.0
Mono-Method Bias
66.1
94.0
Evaluation Apprehension
0.0
0.0
Experimenter Expectancies
3.4
0.0
Confounding Constructs with Level of Constructs
1.7
0.0
Interaction of Testing and Treatments
0.0
0.0
Restricted Generalizability Across Constructs
22.0
0.0

External Validity



Sample Bias
94.9
52.0
Interaction of Selection and Treatment
1.7
0.0
Interaction of Setting and Treatment
74.6
0.0
Interaction of History and Treatment
0.0
0.0

Jonas et al. BMC Complementary and Alternative Medicine 2001 1:12   doi:10.1186/1472-6882-1-12