Since, given data is not continous, so we add 0.5 in the lowe and upper limit of each class
Here, (assumed mean) a=57, (calss width) h=3 (total observations ) N=400 and `sumf_(i)u_(i)`
By step deviation method,
`"Mean"=a(sum f_(i)u_(i))/(N)xxh=57+(25)/(400)xx3`
`=57+0.19=57.19`