- Thread starter
- #71
BlacksheepCardigans
Songster
Quote:
You can culture the failed eggs, which is the instrument you're talking about. However, you need sterile conditions and a hood to do it.
It has to be hundreds because a small n (number of objects in the experiment) creates a large p (probability that the results are due purely to chance or to other variables) unless you control the experiment extremely rigidly. When poultry researchers study what happens when you do something to hatching eggs, they dial the variables down as low as they possibly can - the birds are all the same breed, they're all from the same flock, the flock is only one age (and they mean within a few weeks of each other), they're all in the same incubator, etc. - and they STILL have to use hundreds of eggs to get a p low enough to have a valid result. Hundreds is actually considered a small n.
The reason that all published research starts with a long explanation of where the subjects came from ("Three hundred eggs were received from a 30-week-old commercial broiler flock" and so on) is because that's how you prove that the result can be trusted.
If a normal small breeder wanted to get a reliable result, he or she would have to get an enormous n in order to get p down to a level where the result can be seen to be significant, and they might not ever be able to do it. There would just be too many alternative hypotheses and the calculations would become impossible.
That's why the ag colleges are so valuable and why big broiler farms and so on don't just try stuff and see what happens. The research done by the ag colleges has the ability to generate meaningful results in a way that the farms can't, allowing the farms to develop best practices and change their behavior to get the best results without sacrificing a hundred thousand eggs. Like, for example, this guy, who did the research on using quat in a way that helps hatchability.
You can culture the failed eggs, which is the instrument you're talking about. However, you need sterile conditions and a hood to do it.
It has to be hundreds because a small n (number of objects in the experiment) creates a large p (probability that the results are due purely to chance or to other variables) unless you control the experiment extremely rigidly. When poultry researchers study what happens when you do something to hatching eggs, they dial the variables down as low as they possibly can - the birds are all the same breed, they're all from the same flock, the flock is only one age (and they mean within a few weeks of each other), they're all in the same incubator, etc. - and they STILL have to use hundreds of eggs to get a p low enough to have a valid result. Hundreds is actually considered a small n.
The reason that all published research starts with a long explanation of where the subjects came from ("Three hundred eggs were received from a 30-week-old commercial broiler flock" and so on) is because that's how you prove that the result can be trusted.
If a normal small breeder wanted to get a reliable result, he or she would have to get an enormous n in order to get p down to a level where the result can be seen to be significant, and they might not ever be able to do it. There would just be too many alternative hypotheses and the calculations would become impossible.
That's why the ag colleges are so valuable and why big broiler farms and so on don't just try stuff and see what happens. The research done by the ag colleges has the ability to generate meaningful results in a way that the farms can't, allowing the farms to develop best practices and change their behavior to get the best results without sacrificing a hundred thousand eggs. Like, for example, this guy, who did the research on using quat in a way that helps hatchability.
