✗ ci )
design matrix ,
with ith row removed
Yii) similar to xci)
Bci ) y
Xi ith row of ✗ matrix
Yi ith element of y
Ii ith estimated residual
,
when all observations used to calculate §
Expression for studentised residuals
externally
n
ti point )
}
= Yi -
Xi .
Bci) Memorise our start
Sii ) ( I -1 x :( ✗F) ✗ ;)
- '
a kit
>
Similar to M matrix
'
ie ) M In ✗ ( ✗ Tx) XT
-
= -
v
show that expression can
be simplified to
ti = eli
still hii -