ABSTRACT
BACKGROUND: A previous study reported a significant statistical interaction between experiment date and treatment effect of Argentum nitricum 14x-30x on the growth rate of duckweed (Lemna gibba L.). The aim of the present study was to investigate the stability of the test system and intra-laboratory reproducibility of the effects found. METHODS: Duckweed was treated with A. nitricum potencies (14x-30x) as well as succussed and unsuccussed water controls. The outcome parameter area-related growth rate for day 0-7 was determined by a computerised image analysis system in two series of independent randomised and blinded experiments. Systematic negative control (SNC) experiments were carried out to investigate test system stability. Statistical analysis was performed with full two-way analysis of variance (ANOVA) and protected Fisher's Least Significant Difference (LSD) test. RESULTS: In the first repetition series we found a significant treatment effect (p = 0.016), while in the second series no effect was observed. The negative control experiments showed that the experimental system was stable. An a posteriori subgroup analysis concerning gibbosity revealed the importance of this growth state of L. gibba for successful reproduction of the statistically significant interaction in the original study; flat: no interaction (p = 0.762); slight gibbosity: no interaction (p = 0.356); medium gibbosity: significant interaction (p = 0.031), high gibbosity: highly significant interaction (p = 0.005). CONCLUSIONS: With the original study design (disregarding gibbosity status of L. gibba) results of the original study could not be reproduced sensu stricto. We conclude that the growth state gibbosity is crucial for successful reproduction of the original study. Different physiological states of the test organisms used for bioassays for homeopathic basic research must carefully be considered.