Guidelines in Hypothesis Testing with One Population

H0 Value of Test Statistic H1 Rejection Region

x − µ0
µ = µ0 z= √ if σ is known µ < µ0 z < −zα
σ/ n
x − µ0
z= √ if σ is unknown µ > µ0 z > zα
s/ n
n ≥ 30 µ 6= µ0 z < −zα/2 or z > zα/2
x − µ0
µ = µ0 t= √ µ < µ0 t < −tα
s/ n
σ unknown µ > µ0 t > tα

n < 30, df = n − 1 µ 6= µ0 t < −tα/2 or t > tα/2

p̂ − p0
p = p0 z=p p < p0 z < −zα
p0 (1 − p0 )/n
X
p̂: sample proportion = p > p0 z > zα
n
p: population proportion p 6= p0 z < −zα/2 or z > zα/2

Guidelines in Hypothesis Testing with Two Populations

H0 Value of Test Statistic H1 Rejection Region

(x1 − x2 ) − d0
µ1 − µ2 = d0 z=p µ1 − µ2 < d0 z < −zα
(σ12 /n1 ) + (σ22 /n2 )
σ1 and σ2 known µ1 − µ2 > d0 z > zα

µ1 − µ2 6= d0 z < −zα/2 or z > zα/2

(x − x2 ) − d0
µ1 − µ2 = d0 t= p1 µ1 − µ2 < d0 t < −tα
sp (1/n1 ) + (1/n2 )
(for large samples) σ1 = σ2 but unknown, df = n1 + n2 − 2 µ1 − µ2 > d0 t > tα
(n1 − 1)s21
+ (n2 − 1)s22
s2p = µ1 − µ2 6= d0 t < −tα/2 or t > tα/2
n1 + n2 − 2
(x1 − x2 ) − d0
µ1 − µ2 = d0 t= p 2 µ1 − µ2 < d0 t < −tα
(s1 /n1 ) + (s22 /n2 )
σ1 6= σ2 and unknown µ1 − µ2 > d0 t > tα
(s21 /n1 + s22 /n2 )2
(for small samples) df = µ1 − µ2 6= d0 t < −tα/2 or t > tα/2
(s21 /n1 )2 (s2 /n2 )2
+ 2
n1 − 1 n2 − 1
d − d0
µD = d0 t= √ µD < d0 t < −tα
sD / n
µD : mean difference df = n − 1 µD > d0 t > tα

µD 6= d0 t < −tα/2 or t > tα/2

(p̂1 − p̂2 ) − (p1 − p2 )
p1 − p2 = 0 z= s   p1 − p2 < 0 z < −zα
1 1
p̄q̄ +
n1 n2
X1 X2
p̂1 = and p̂2 = p1 − p2 > 0 z > zα
n1 n2
X1 + X2
p̄ = and q̄ = 1 − p̄ p1 − p2 6= 0 z < −zα/2 or z > zα/2
n1 + n2
 Guidelines in Hypothesis Testing the Significance of Correlation

H0 Value of Test Statistic H1 Rejection Region

√
r n−2
ρ=0 t= √ ρ<0 t < −tα
1 − r2
(Pearson Product-Moment df = n − 2 ρ>0 t > tα

Correlation Test) r: correlation coefficient ρ 6= 0 t < −tα/2 or t > tα/2

√
ρ n−2
ρ=0 t= p ρ<0 t < −tα
1 − ρ2
(Spearman Rank df = n − 2 ρ>0 t > tα
2
P
6 D
Correlation Test) ρ=1− ρ 6= 0 t < −tα/2 or t > tα/2
n(n2 − 1)
D: difference in ranks between X and Y

Guidelines in Hypothesis Testing of Means of Three or More Populations

H0 Value of Test Statistic H1 Rejection Region

s21
µ1 = µ2 = . . . = µk f= at least two of the f > fα [k − 1, N − k]
s22
means are not equal (i.e. fα [column,row])

Source of Variation Sum of Squares Degrees Mean Square Computed f

of Freedom
k
X SSC s21
Column Means SSC = ni (x̄i − x̄)2 k−1 s21 =
i=1
k−1 s22
ni
k X
X SSE
Error SSE = (xij − x̄i )2 N −k s22 =
i=1 j=1
N −k

SSE = SST − SSC

ni
k X
X
Total SST = (xij − x̄)2 N −1
i=1 j=1

Guidelines in Hypothesis Testing for Independence of Variables

H0 Value of Test Statistic H1 Rejection Region

Xn
(Oi − Ei )2
i=1
variables are χ2 = , variables are χ2 > χ2α [(r − 1)(c − 1)]
Ei
independent Oi := observed frequencies dependent r:= # of rows

Ei := expected frequencies c := # of cols

Standard Normal Distribution (critical z-values)

t Distribution (critical t-values)

  !"#$%&'()*+,-#%.-/!!("(0(#1
23456746859:;4<56=7>885?345?>@56745?345674685?>5?3457:93?5>A5?345=7:?:=6B5;6BC4D52>5B>>E5C@56<567465><5?345B4A?F58CG?76=?5:?
A7>H5><4F56<I5?34<5B>>E5:?5C@5J:4K5χLD2LMDistribution
5><5?345B4A?5:85LDNM5><(critical
5?3457:93?O χ2 values)
PQ RSTTU RSTT RSTVU RSTU RSTR RSWR RSRU RSRXU RSRW RSRRU
W YYY YYY LDLLZ LDLL[ LDLZ\ ]D^L\ _D̀[Z MDL][ \D\_M ^D̀^N
X LDLZL LDL]L LDLMZ LDZL_ LD]ZZ [D\LM MDNNZ ^D_^` ND]ZL ZLDMN^
a LDL^] LDZZM LD]Z\ LD_M] LDM`[ \D]MZ ^D̀ZM ND_[` ZZD_[M Z]D̀_`
b LD]L^ LD]N^ LD[`[ LD^ZZ ZDL\[ ^D^^N ND[`` ZZDZ[_ Z_D]^^ Z[D̀\L
U LD[Z] LDMM[ LD̀_Z ZDZ[M ZD\ZL ND]_\ ZZDL^L Z]D̀__ ZMDL`\ Z\D^ML
c LD\^\ LD̀^] ZD]_^ ZD\_M ]D]L[ ZLD\[M Z]DMN] Z[D[[N Z\D̀Z] Z`DM[`
V LDN`N ZD]_N ZD\NL ]DZ\^ ]D̀__ Z]DLZ^ Z[DL\^ Z\DLZ_ Z`D[^M ]LD]^`
d ZD_[[ ZD\[\ ]DZ`L ]D^__ _D[NL Z_D_\] ZMDML^ Z^DM_M ]LDLNL ]ZDNMM
T ZD^_M ]DL`` ]D^LL _D_]M [DZ\` Z[D\`[ Z\DNZN ZNDL]_ ]ZD\\\ ]_DM`N
WR ]DZM\ ]DMM` _D][^ _DN[L [D̀\M ZMDN`^ Z`D_L^ ]LD[`_ ]_D]LN ]MDZ``
WW ]D\L_ _DLM_ _D̀Z\ [DM^M MDM^` Z^D]^M ZND\^M ]ZDN]L ][D^]M ]\D^M^
WX _DL^[ _DM^Z [D[L[ MD]]\ \D_L[ Z`DM[N ]ZDL]\ ]_D__^ ]\D]Z^ ]`D_LL
Wa _DM\M [DZL^ MDLLN MD̀N] ^DL[] ZND̀Z] ]]D_\] ][D^_\ ]^D\`` ]ND̀ZN
Wb [DL^M [D\\L MD\]N \DM^Z ^D^NL ]ZDL\[ ]_D\`M ]\DZZN ]NDZ[Z _ZD_ZN
WU [D\LZ MD]]N \D]\] ^D]\Z `DM[^ ]]D_L^ ][DNN\ ]^D[`` _LDM^` _]D̀LZ
Wc MDZ[] MD̀Z] \DNL` ^DN\] ND_Z] ]_DM[] ]\D]N\ ]`D̀[M _]DLLL _[D]\^
WV MD\N^ \D[L` ^DM\[ `D\^] ZLDL`M ][D^\N ]^DM`^ _LDZNZ __D[LN _MD^Z`
Wd \D]\M ^DLZM `D]_Z ND_NL ZLD̀\M ]MDN`N ]`D̀\N _ZDM]\ _[D̀LM _^DZM\
WT \D̀[[ ^D\__ `DNL^ ZLDZZ^ ZZD\MZ ]^D]L[ _LDZ[[ _]D̀M] _\DZNZ _`DM`]
XR ^D[_[ `D]\L NDMNZ ZLD̀MZ Z]D[[_ ]`D[Z] _ZD[ZL _[DZ^L _^DM\\ _NDNN^
XW `DL_[ `D̀N^ ZLD]`_ ZZDMNZ Z_D][L ]ND\ZM _]D\^Z _MD[^N _`DN_] [ZD[LZ
XX `D\[_ NDM[] ZLDN`] Z]D__` Z[DL[Z _LD̀Z_ __DN][ _\D^`Z [LD]`N []D^N\
Xa ND]\L ZLDZN\ ZZD\`N Z_DLNZ Z[D̀[` _]DLL^ _MDZ^] _`DL^\ [ZD\_` [[DZ`Z
Xb ND̀`\ ZLD̀M\ Z]D[LZ Z_D̀[` ZMD\MN __DZN\ _\D[ZM _ND_\[ []DN`L [MDMMN
XU ZLDM]L ZZDM][ Z_DZ]L Z[D\ZZ Z\D[^_ _[D_`] _^D\M] [LD\[\ [[D_Z[ [\DN]`
Xc ZZDZ\L Z]DZN` Z_D̀[[ ZMD_^N Z^D]N] _MDM\_ _`D̀`M [ZDN]_ [MD\[] [`D]NL
XV ZZD̀L` Z]D̀^N Z[DM^_ Z\DZMZ Z`DZZ[ _\D^[Z [LDZZ_ [_DZNM [\DN\_ [ND\[M
Xd Z]D[\Z Z_DM\M ZMD_L` Z\DN]` Z`DN_N _^DNZ\ [ZD__^ [[D[\Z [`D]^` MLDNN_
XT Z_DZ]Z Z[D]M\ Z\DFL[^Distribution Z^D^L` ZND^\`(critical _NDL`^f -values)
[]DMM^ [MD^]] [NDM`` M]D__\
aR Z_D^`^ Z[DNM_ Z\D^NZ Z`D[N_ ]LDMNN [LD]M\ [_D^^_ [\DN^N MLD̀N] M_D\^]
bR ]LD^L^ ]]DZ\[ ][D[__ ]\DMLN ]NDLMZ MZD̀LM MMD^M` MND_[] \_D\NZ \\D^\\
UR ]^DNNZ ]ND^L^ _]D_M^ _[D^\[ _^D\`N \_DZ\^ \^DMLM ^ZD[]L ^\DZM[ ^ND[NL
cR _MDM_[ _^D[`M [LD[`] [_DZ`` [\D[MN ^[D_N^ ^NDL`] `_D]N` ``D_^N NZDNM]
VR [_D]^M [MD[[] [`D^M` MZD^_N MMD_]N `MDM]^ NLDM_Z NMDL]_ ZLLD[]M ZL[D]ZM
dR MZDZ^] M_DM[L M^DZM_ \LD_NZ \[D]^` N\DM^` ZLZD̀^N ZL\D\]N ZZ]D_]N ZZ\D_]Z
TR MNDZN\ \ZD^M[ \MD\[^ \NDZ]\ ^_D]NZ ZL^DM\M ZZ_DZ[M ZZ`DZ_\ Z][DZZ\ Z]`D]NN
WRR \^D_]` ^LDL\M ^[D]]] ^^DN]N `]D_M` ZZ`D[N` Z][D_[] Z]NDM\Z Z_MD̀L^ Z[LDZ\N
26GB45>A5e><?4<?8

22989869269289920412788 020