Later, you conduct another experiment where you aim to incubate eggs at several different temperatures (low, medium and high... say 25.C, 28.C and 31.C), to see how this affects sex ratio. Only three or four eggs fit in each incubator at a time, but luckily you have access to lots of incubators! Unfortunately, when you test the incubators with an independent probe that you know if highly accurate, you find that the thermostats in the incubators are not very accurate and thus it is impossible to get each incubator to be exactly the same temperature. For example, an incubator set to be at 28.C may actually reach a temperate of 28.5-C or 27.1.C or 29.2.C. So you put three or four eggs in each incubator, record the actual temperature reached in each incubator using the probe that you know if highly accurate, incubate the eggs for the relevant time at that constant temperature, and then for each egg you record whether it ends up male or female. You find that each incubator was actually at a different temperate, but the range of temperatures achieved in the incubators covered the range of temperatures that you are interested in. When you discuss the analysis with your supervisor, she assumes that you will calculate the proportion of males for each incubator, and then conduct a standard one-way ANOVA with three levels to test whether proportion of males varies with temperature. What do you say to her