Good question. The formatting wasn't an issue for me, I just pasted the answers into notepad and started to put line breaks and spaces where I would put them if writing SQL - and realised before I finished that I would be wasting time doing any more formatting as the four answer options were identical apart form a couple of column names, which made it all pretty clear (apart from remembering which of the column ids was the id in the parent table).

I think the explanation is slightly wrong, though, as it seems to suggest that key_ordinal is the same as index_column_id in all entries in sys.index_columns for anything except columns of clustered indexes; but actually this can only be true for key columns, since for non-key columns key_ordinal is always 0.