It sounds like you're trying to write a report for an experiment you've already done, rather than writing a proposal for an experiment which you intend to do.
You should already have some kind of theoretical background for your experiment (which would be what prompted you to conduct the experiment in the first place). If Bob's Law said that the susceptibility of a ceramic to breaking is proportional to its temperature, your experiment might be to investigate this and see whether you can provide evidence for or against it. So, Bob's Law would be your hypothesis. Then you draw the properties that you're going to measure from that hypothesis. In this case, it would be susceptibility to breakage and temperature.
I understand what you're saying - you want to work backwards, and have the perfect experiment where everything can be done and fits in neatly. Experiments aren't perfect. If you have difficulty measuring the temperature of the ceramic, document it and approximate it as best as you can. The same goes for susceptibility to breakage.
It doesn't matter if you can't get an exact quantitative measurement of temperature, or if you can't compensate for air resistance rotating your ceramics when they're falling. You try to overcome it, of course, but if you can't it's just another thing you add to the discussion list.
Decide on your hypothesis, and measure properties that correspond to it. Fitting your hypothesis to your experiment instead of the other way around completely defeats the purpose of conducting an experiment in the first place, and probably won't go down well with your teacher (or at least it didn't with mine).