A field itself couldn't be tracked then it's always part from a table. Load times from tables could be tracked per load-logging (option within the document properties in tab general) or per time measurement like:
I may be totally incorrect Marcus, but isn't the load time for a specific field within a table same as the load time for the table itself?
Lets say there are two tables, Table 1 has 10 columns and Table 2 has 1000 columns, will Table 2 take longer to reload? It make sense that it should, but with my limited experience I have found out that number of columns haven't impacted the reload time for me. What did impact was the number of rows in the table.
A field is always related to the other fields in its table (and isn't a single table to all other linked tables, too) and therefore the other fields may have an impact and often they does.
For example, if you load a sales-table with 50M rows it will be give a difference if you load only three fields with a date, a storeID and a value or if you included much more fields maybe an additionally timestamp which could be result in millions of unique fieldvalues instead of probably a few thousands unique fieldvalues. It won't be the same neither from table/file-size and load-times.
Otherwise if the amount of records aren't really big the performance of qlikview is often too fast to notice bigger differences.
