moon.Had we included admissions for self-injury and accidental poisoning, including smoke injury, the figures would have been 390 and 336.The mean number of selfpoisoning admissions, with standard deviation, on full moon days was 5 9 2 3 (range 2-10) but on new moon days only 5-1 2 6 (0-12).We made comparisons for each full moon and new moon pair, and found t= 1 8487, df= 61.If we merely ask whether there might be a significant difference among the pairs then p = 007, which does not reach standards for significance.However, the report from India specified a direction of difference and allowed a prediction, and if we consequently adopt a one-tailed test then p =0035.The likelihood that the higher rate of admissions for poisoning on full moon days than on new moon days in Edinburgh, as in Patna, could be a matter of chance alone would thus be only about 1 in 28.
No takes yet. Share an insight, caveat, or question.
G T Haneveld (1981) studied this question.