Ah, I had to do this last semester (except we had to write code that did the conversion in assembly languages! X_X) Well, if you've gotta know it by hand for a test, then a calculator probably isn't going to help you. The process of turning a decimal value into a floating point is fairly straightforward, especially if you already know how to find the decimal representation in binary. (honestly, I'd need to brush up on my binary arithmetic to do that stuff again).
But if you can get to that point, then it's not bad. The first bit is just a sign bit, 0 if the number is positive, 1 if it's negative. Now, you just move your decimal point to be right after the 1st '1', this pretty much gives you the scientific notation for it (well... in binary). So, what you do now is count the number of times you moved the decimal point to the left (count negative if you move it to the right), and this will give you the exponent. Now, just add 127 to that, and write that value in for the next 8 bits of the floating point. (the reason you add 127 is to allow you to have negative exponent values. Weird, but you've just gotta remember it). Now, to fill in the rest of the 23 bits left, you write out all the numbers in the decimal number you wrote down right after that first '1' value. With all that information, the floating point has enough information to know exactly (well, with accuracy issues) what decimal value you have.
Yup... pretty straightforward... though the whole process really sucks. If you can remember what the 1, 8, and 23 bits need to hold, you should be in good shape. It's just really easy to get messed up with all the arithmetic. I can assure you that after this class, unless you're going to get really hardcore into hardware design, you with never, ever use this knowledge again. Though it is kinda cool knowing how a computer can decode all this stuff. At any rate, that's about all the help I can give, as complicated as it all is. Good luck with the test!