Element size does not alone determine the melting point of the element. Look at Xenon. It's radius is about 105 pm, yet it's melting point is FAAAAAAAAAAAAAAAAAR below that of Cesium. Cesium has a radius of about 300 pm and is a mighty big element, but that is not why its melting point is so low. (Also, many of the gases and liquids at room temperature have low atomic radii, yet elements which are bigger than them can have much higher melting points).
The real reason why the melting points decrease down Group 1 is due to a combination of electron configuration and element size. The elements down towards the bottom of group one have a great deal of nuclear shielding going on. As a result, the electrons are more spaced out due to their own repulsion. Those outer electrons have a hard time seeing their own nucleus, let alone the nucleus of a surrounding atom. As a result, the bonding between the atoms is fairly weak which results in a lower melting point. Elements like lithium and sodium have a higher melting point because their atoms can still see a part of their nuclei so there's a slightly stronger attraction. Now at this point you might be saying 'yeah, that's exactly what I said before. The bigger elements have a lower melting point'. The funny thing is, if you look at the halogens the complete opposite is true. As you go down the column, the atom sizes increase yet the melting points increase as well. So saying that one element will melt at a lower temperature than another simply because the atomic radius is bigger is not correct.