You should always specify the scale and precision if you can. It does not always matter, but sometimes it does, so it is a good habit to develop. It matters when storing values that have been computed using floating point arithmetic. If so, the result may have a lot more precision than you require, and that precision wastes storage space, which in turn causes extra I/O. Consider the following example. N1 is a NUMBER column; and N2 is a NUMBER(2) column. The row inserted has the same value for both columns, but N1 requires 19 more bytes of storage than N2!

    SQL> create table numbers (n1 number, n2 number(2));

    Table created.

    SQL> insert into numbers values (log(2, 4), log(2, 4));

    1 row created.

    SQL> select * from numbers;

	    N1         N2
    ---------- ----------
	     2          2

    SQL> select vsize(n1), vsize(n2) from numbers;

     VSIZE(N1)  VSIZE(N2)
    ---------- ----------
	    21          2

In schema definition for number columns, defining "Number" or "Number(x,x)" is more efficient. How does space management handle this? Will there be any performance degradation if "Number" datatype is used. Please support your theory.