Topic: Starting with a conclusion and then searching for evidence

[AngleWyrm]

I'm not sure, but I suspect this expression is not completely sound logical thinking, and here's why: I conclude that my coffee cup will hold one cup of coffee.

If challenged, I can go on to give anecdotal evidence by saying that this morning I poured 8oz of coffee into it, and it was full after that. Maybe I can reference documentation produced by the manufacturer that states this model is designed to hold 8oz of liquid. Maybe I've got historical evidence in the form of pouring 8oz from a demarcated coffee pot into that cup several times a day for months.

But all I've done is gone on to defend my position with various proofs and evidence. It starts to smell funny, because some might argue that I've started with a pre-conceived notion and then gone on to display only supporting evidence.

Here's the problem: I was quite sure my cup would hold 8oz of coffee even before the presented evidence became available. Does that mean any experiments conducted by me on my coffee cup are necessarily flawed by my prejudice? Do I have to prove that I don't actually believe the cup will hold about 8oz in order to prove that it does?

[ting-bu-dong]

Here's the problem: I was quite sure my cup would hold 8oz of coffee even before the presented evidence became available. Does that mean any experiments conducted by me on my coffee cup are necessarily flawed by my prejudice? Do I have to prove that I don't actually believe the cup will hold about 8oz in order to prove that it does?

No, as long as the experiment is set up in such a way that the outcome will accurately reflect the holding capacity of the cup regardless of what you believe about the cup, which makes it an empirical problem rather than one about logic.

If the experiment is to keep pouring until the cup is full and recording how much volume of liquid it took to fill the cup, there are a couple of potential outcomes:

1. The capacity is what it seems and the liquid volume equals the cup's interior volume.
2. The cup has a barely visible leak and no matter how much you pour the cup will never fill.
3. The cup is made of an absorbing material so the poured volume exceeds the capacity of the cup.
4. There is a mirror inside the cup which makes it appear larger or smaller than it really is.
5. There is a dimensional vortex at the cup's bottom which captures some of the liquid.

And so on. Whatever the case, the experiment will accurately determine the actual properties of the cup regardless of your biases, as long as the cup's volume and the volume of poured liquid are recorded accurately.

[superdave]

The real problem is that science is messy and rarely follows the the neat "scientific method" paradigm they teach in grade school.  Obviously that is what scientists shoot for, but because results can be ambiguous, experiments can be hard to physically perform, and new ideas can come up along the way, it rarely unfolds by the book.

Long story short, starting with a conclusion and searching for evidence can be ok, as long as a thorough and competent job is done.

[andrewclunn]

Having a hypothesis before doing research is fine.  Hopefully you can be objective enough to design tests that won't be affected by bias.  If not, that's why we have peer review.