Statistical evaluation of methods using headspace gas chromatography for the determination of ethylene oxide

Two methods for the determination of levels of ethylene oxide, both based on headspace gas chromatography but one involving internal and the other external standardisation, have been evaluated by comparing their performances in determinations by three analysts using previously characterised material. Statistical analysis has shown that the methods give results that are generally consistent. There was a significant interaction between one analyst and the external standard method, associated with intermittent high determinations, but there was little evidence of such interactions involving the other analysts. There was also little evidence of any significant difference between the two methods, both of which may be recommended subject to the use of conditions adequate for the complete desorption of ethylene oxide.