[CHOOHO]

The empirical and molecular formulas of an unknown compound are determined by combustion analysis and a freezing point depression experiment.
a. When a 0.8425 gram sample of an unknown molecular compound containing only carbon, hydrogen and oxygen is combusted in excess oxygen, 1.8850 grams of carbon dioxide are produced and 0.8996 grams of water are formed.
i. Calculate the mass in grams of oxygen required for the combustion
ii. Calculate the mass in grams of oxygen contained in the unknown compound (I got 0.2273g)
iii. Determine the empirical formula for the unknown compound (for this - do I use the grams of C, H, and O that I already found and convert them into moles? Even then, they still come out as decimals. Very confused)

I figured out the grams of C and H in the CxHxOx molecule but I can't find out the grams of O2 that the CxHxOx combusts in.

[fledarmus]

Start with a balanced equation for the reaction...

[CHOOHO]

Start with a balanced equation for the reaction...

I can't do that since the empirical and molecular formula is unknown. The most you can get out of what you said is:
CxHxOx(aq) + O2(g) ----> CO2(g) + H2O(g/l)

[Borek]

You don't need reaction equation. You are given initial mass, calculate how much carbon and hydrogen was present in the initial sample - if the mass of the sample is higher, the rest must be oxygen.

[fledarmus]

Start with a balanced equation for the reaction...

I can't do that since the empirical and molecular formula is unknown. The most you can get out of what you said is:
CxHxOx(aq) + O2(g) ----> CO2(g) + H2O(g/l)

Yes you can - you can balance it using variables. Try starting with C_xH_yO_z (rather than all x's, which implies identical subscripts). From this, you will be able to work out the relative number of moles of each compound. Then the masses that you have can be used to back-calculate everything else you need.