,
,
, Factors influenang Om
If X remains equal and N increases, Om decreases /up until v 1000)
· =
=larger sample size yields smaller estimation errors
IfN is constant, And O of scores increases Om increases
↑
.
↳small Om is good because it indicates that M scores are close , to M ,
so less
Prediction error ,
so magnitude of sampling error is small
Distance from M to M
· location of a
single X score
by computing a z-score
(M
32 =
· location of sample mean Example :
2 N = 130 M = 73 76 . 0 7 062
=
.
population
n= 9 M 76 22
=
.
sample
Among .354
2
z-values above 1 96 .
are considered
-if do not have we O ,
use SD Question : is M far from M ? far fram M
=SEm=
6.
->
for 95% confidence interval , no
characteristics
=same
formula , same
↳ But it does have 2 errors ; SD & M
the t distribution (or when sample is small
so : if O is not known ,
we use
t =
M
-
-
Mo ->
M = No + toSEm
SEM
Tdistributions ↓
degrees of freedom :
n 1-
-> the
larger of ,
the more normal the curve
->
When af >120 ,
the sample error of SD becomes negligible
= close to o = known
note af D =
normal distribution
=Middle 95% with increasing of