Relative validation of a pre-coded food diary in a group of Norwegian adults – Comparison of underreporters and acceptable reporters