See the attached SAS program graders.sas and its output, graders.lst. The analyst
who wrote graders.sas was interested in testing whether or not the graders all
graded to the same standard and in estimating the mean score given by each
grader, but was not sure how to implement the appropriate analyses, so he tried
more than one approach in graders.sas.
b.
(6
pts.) Write down an appropriate model for the analysis of these data
and write down the hypothesis that the analyst is interested in testing in
formal terms (that is, in terms of model quantities). Test this hypothesis
at a
=
0.05 and state the appropriate conclusion.
OJ
'
,/or:
_
/."
}
1
v'
:::
re
J'~/j
5c
(~jrc;",J.
e
C;:'i.ve~
j
".1
ItC;
('
':7rooIC/
iJ~
J
e
'":'J.(
I
'
lI.
.
II·'
0
.
7
//'
U
c,)
i
I)
t
I
c~,?I'lt~
u.
~)
25"
c::!
~o
('.,
C?ft
C'' C
I
/lO
/'Z.
.t
J.

e
X{Z11Vl
')
(':; ,) '.  . )
.J
.j
J
..' )
i
\
(nD!
oj/
DosS
I
.;/e
(r:.l,"Vt~i.tlC;.lJ1J
01'
()
J
c:·.re
<5
6.st/
Ve
cL)
/\/J.
I!L
I
7
.,.'
J
('
2.. )
'_
I
..
LD
~
.
r
,
. ,
t.
l
,I
'~~.
.
~

i
./'
i
'
'.' _
+
J.+
2>'
+
e.
''
1
E':,'
'J
rJ
N
u
[;
2
or
=
0
L.
&
=
0
.
,M
<:71 [
r
J
lJ)
lJ
)')
.
I
).
I
J
ft
LI
c.
,j
///1,. /'7
I
.'
..J
I
/','
,(
OV'0.~"
I
~
/'=
SIl.
c,de/'
.
\..)
e
eXt:"',";'"
/~d
J
I('i)
3.
\,
/0'Jcll1.s
15, ':
H;
d
1;='" ::::
o{z.::C
0
,~~v"
~
(if!.
{/
rU,.1
r/
'J
C
J
Uf)
i//~
0.
J
e
feci
E
t
KL
Lj
.
A
::::
.'
:=
A.A
} <
Dr
1,o~l"'\I
/
£,.
.
~
L(,
'1
f
b A
2
L(
7{;
d.l
LJ
LJi2
',U
=
/llA'f
c>«
~;0:3
({C,Jz/
l
,
F
/

j
D <
..
0
0
01
_~
)
;2e;}2
C
;
lb,
,br
c
1.<.J£.r
s
}
y;//
1/
,
v:v~£:~;eI
eXcc.V'o15
J
(')
/JDJ
';(
f';,.
[)}
e
h
,/'1.(
J&vc>(
Sl:,:V\.d~d"
j
CY
}
"J
)
:'/
~
(C~
01'"\
C.j
If
c.
(4
pts.) Estimate the mean score that grader number 3 assigns when
j,
c;;
v{'
Jc;~Vh€.
grading writing samples of this kind and give a 95% confidence interval for
/J
~
)
this quantity.
/l/[.cL/{
SC,L/e

;:
b:J)(
62
/
L/
C'
JJJ
cL
of
(0,(D)~
C7,95/Y)
>
..~r"
2