3/11/2017

I NQIDUNGCOB~
[ MON PHAN TiCH DA BIEN

XEM xET TiNH CHAT DiJ LI:E:U

-
D~ tili:

1'7'\ IcAc Kt TBU.j.T BllTIII

\,!.. . (Grwpllktll . . . ..,.)

(EXAMINING THE DATA)

Q) 1 =-~LW!QUANBiNDO"LJluTIIItu l
Nhom 8 - L{yp Qutin tr/ Klnh doanh 05 - Cao hpc K26

Nguyln Thi Dlim (Nh6m trnemg) @ CW GIA DINK TB6NG :d QUAN TRQNG
Trdn Hoang Linh (fU IIIIJIOriMt ~Aml-,plilJ•)
L,e Thi Thanh Thuyen
Nguyln Qu6c H&ng
rran Thi Thu~

Nhom 8 CHA05NCK26

IcAc Kt TBU.j.T BllTIII cAc Kt TBU.j.T BllTIII

(Grwpldcld »ddlflla) I. (Grtlpllklll ~. .)

11. Bllu dA phla b6 tln suit (lllslogtfllfl) I 11. BJiu dA pbln pb6l tln suit (llislogNIII)
Biiu c16 phan phdi (phan b6)
tAn suit Ia m¢t dt;mg bi~u ao
cua phan phoi ella 1 bi~n dan;
th~ hi~n tan suAt xay ra cua 01
gia tri trong bi~n.
ThU'ong OU'Q'C sl} d1,mg d~ t~o ra
nhCrng $0 sanh tryc quan aoi
v6'i cac phan ph6i thong thU'ong.

Sliu c16 Pie Chart va S.r Chart thong thcr6flg Bt4u cfd phan ph6/ bi4n s6 tu6t cua
ngiJ'dll w dpng xe

Nhom 8 CHAD5fiCK26 NhOm 8 CHA05NCK26

1
 3/11/2017

I
:I tiC Kf THU~T Db Till I(Grtlpllbl1«:1Jif••)
cAc Id THU.6-T 116 TBJ
:I
I
I(GNpltbl Tdlf•a)

~ 1. Bliu ttA pbln pb61 dn suit (Histogr11m) I [t. Bliu ttA pbln pb61 dn suit (lllstogram)
Bi~u cro histogram t1U'Q'C xiy dvng
dU'6'i cac thanh th~ hi~n gia tq dCY
/
P·\ r - Bi~u do dlnh IU'Q'ng nhimt . r-/
~\

~r-
xtv
li~u; t1U'Q'C dwng bing cich «6m
gii tri so nguyln cua tin suit cua
v ~~\ \
d'c diem cua phan ph6i mQt
each dO'n gian va trl)'c quan
["
I
/1 - -
\
\
gii tri dfJI lilu -+ xiy dU'ng bang kA hO'n. ! r~l
'

Bi~u dO histogram dU'Q'C hCYu dvng /

\ - E>oong cong dU'Q'C them vao
d~ dimh gia phan ph6i thf/0 " I
I I\
d~:
1)PhAn nh6m cAc aii tri dii neu cua
cac bi~n s6 th~nh cac khoang.
1 •, , . ..

O'l>fool
/.} l)
ti, v6'i phan ph6i mong dQ'I
,
cua cac nha ngh1en cU'u
/f. .. ... ~·
.\,
(2)Th~ hi~m cac d~ng hinh cua phAn BJ4u t16 phln ph6i bidn
'16 tu61 cua ngf.I'Oti su Ghl CbUi Bltu dO histogram kh6ng dung duyc cac bl4n danh nghia
ph6i nhU' (J6 m6o. (J6 nhon d!Jngxe (A histogram iln1 not used fD representlh• nQI1metrlc velue)

Nhcm a CHAJ)$NCK26 Nhdm a CHAD~CK26

cAc Id THU.6-T II{) TBJ

I. (Grtlpldal »ddlf-a) ~~~TBJ . I
12. Bliu ttA pb&n tan (Sc11~t) I

,,
05 o,ng tU'O'I'IQ quan:
j2.. Bliu ttA pbln tin (SCIIttnplllt)
y y y
y 00
- Bieu do phan tan (bieu do tan o 0 0

'
o oC o
x,) Ia mot ky thu$t do thi d~ nghltln c
clt'u m6i quan hf giiPa hai b(} s6 I 4 00
c~p ~oo
0

~~
0 U ooo
lifu lien hf xay ra theo c,p. 000 c co
o 0 o
- Bi~u do phEm tan trinh bay cac 0
~ 6 0

c@p nhU' mi)t tflm may di~m. - X X X

- Hinh dfl.ng clc cfam may t/6 ~ y
' ' ' I
'
y
iJ 0 0
M6i lien hf d'ng gi. 0 <> o o 0 D o U C
0 oO
.Jo o o C g o o
0
r::: n o - Ht.tang {dl~ctfon)
o 0c o C 0 0 0 0 0 0 Dtng(torm)
c IJ oo
0
co ~ o
t)
CJ
IJ
~CI D DO o
00 o 0 D
eo chot (strength )
0
c Drf Clic cflem bat thl/O'ng
A u u o n {unusual features)

Nhdm 8 CtiAJ»ffCK26 X Nh ' X 8

2
 3/11/2017

IcAc Kt TIIII4T BO TJQ

(GrqldaJI Tdlf•a)
I cAc KfTHU,6.TB0 TIQ
(Grqlllclll Tdipa)

j :~. Bila etA phln tan (ScllttDplot) ( 13. Blla etA hOp (Boxplot) 0

o 1 bi~u 40· .._P hQ'p cac dhim dqc: theo 1 .- B!Au da hQp, hay con gof fA bi~u da h9p-va-ri u (tii5Kand-
doong tt~~Ang ~ 1 m6i qua., ht tuy~n v6 vj tri. 4f+
lllh/sker plot): m9t hinh in,. ttvc quan khac

tinh (tU'O'nQ quan) phan tan, d~ng hinh, ctt 411 ctuoi va cac gia ttl bit
ttu.rimg (outliers) cua phittphai.
0 1 bi~u "0 ~p hQ'p cac di~m 1heo 1
dU'Cmg cong ~ 1 m6i quaB hf phi tuy~n
tfnh
'------X
o 1 bi~u 40 m6 hinh cac ditm ngiu nhien
~ khong c6 tli'O'Ilg quan

eo chat: Df m(lnh, y~u ti.J'O'f1g quan.

0
t
Nhom 8 CMA~K26 Nhom 8 CliA!)-1<26 10

c.Ac rt mu.~T 80 THJ c.Ac rt TBU.tT 80 THJ

(Grqlllclll T~~ellltJIUII) (Grtqllllclll:&elllf,.) I
3. Blla etA hOp (Boxplot) 3. Bliu etA thln vi Ia (Stem-Luj) I --
Cac thanh phan chu )'tu cua biltu db
*
0 - MOi dong cua btiu dO di.I'Q'C QQi
Age cr motorbike user Stem-and-l.eef Plot
Frequency Stem & l,eaf
hOp Ia: -I GIA TR! LON NHAT J Ia m<?t than; va m6t sOii~u th~ 6.00 1. 889999
(1) Hi;>p hinh chQ> nh~ chlia dt,mg 50% hien tren mot thctn gQi Ia m<?t lc:i 18.00 2 . 000 11122222233344
8.00 2. 55617788
cac gia tri diP lliiL - Khi bi~u do than..v&-la dU'Q'c quay
-I
,
13.00 3 . 0012l33334-M4
a
(2) BU'ong thang trung tam hQp Ia Ttl PHAN V! 3
t rai 90 do, n6 se 00 d~ng hlnh tU'cmg

I
~~!"~ 4.0 3. 5556-
Ia tri trun bin . (mean)
(3) Hai I~ cua hQp j,jtn hai gia tri
11w c6tlj
trteim
arong h6p
.... TRUNGVI tv nhU' bi~u do l;lis1ogram .
12.0U
13.00
10.00
4 . 123333334444
4 • 5555$6ti177189
5. 01233444414
tU> phan vi thlf' 1Jv• tht.l 3 (ruong 9.00 5 . 5GG667779
lf'ng voi gia tri thtit 25'1. (25th -I Ttl PHAN VI 1 2.00
"1.00
6- 03
6. 5567
percentile) va gi' "1 th!Jo 75% (75th
percentile) ella dAY 16 l~u. -I GIATR! NH6 NHAT I 1.00
.oo '
7.. 6
~) Cac "rau" keo dii lr I~ phia tr€m va 0 Stem width: 10
phia dU'oi cua hQp~ gla tri tan nhAt Each lellf: 1 case<•>
vi nho nhAt, nlm trong khoimg t6i Bt§u 6ll ThltH.i t:Ua blffl ed Jll6;
da 1 ,5 l~n khoang Clich giQ>a cac tU> cua ngi.IM slir dt~niJ xe m4y
phan vi tinh tU> 16 ci':la h()p.
Nham 8 CH~CK26 11 Nhom 8 CHA04fiCK26 u

3
 3/11/2017

DU LI:t:U THIEU Vi sao 1\li quan tam d~n gia trj tbi@u?
• Du !flu thiiu (Misfing data): la nhllng
thong tin/du li~u cbn thi~u v8 mOt d6i Gia tri. tbi~u hun giam kich thuac m!u va
tu(]tlg (ho~c m¢t tru<mg hqp) tna cac mAtdttli~u.
thong tin/du lieu khac d6i tuQ'llg ve - T~o ra S\1' chenh tech trong du lieu.
(tru<mg hqp) d6 tbi da c6 sfin - N ~u gia tri thi~u h\ khong ngfiu nhien thi
- < i1:1 111 I run~ 11!,!<,111 ,Jill<'ll 1 I'd.\ !< l I)" ll!,!llnl . JII, ,,. ph•>ll!,!
k~t qua s~ khong do lu<mg dung ml)c dich
1:111 khi'll!:! ,·Jw \ 11 1 I·" ,·:111 l11>i lw:-IL' "" ·,:11 ... 1 l·.l11 nh:q) nghien ciru.
ilL' II
< i1:1 In l!\lll~ k!J,•IIc' 11!'.111 1illll'll i 1\·ll',\1< > '-.1:•11•>1 <I IItle'
J'llilll!,! \<.Ill (Ph ,J.' J.. !J.,II~' 11 :1 !111 llillll';lll II I J,q h1 trllll!,!
J;,,;IL' 111<111!,! ti11h 11<'1"' 111 nh:l\ <.::1111

Nhclm 8 CMA~CK26 14

, , , , """
aANH GIA cAc GIA TRI THIEU aANH GIA CAC GIA TRI THIEU
• Gia trj thi(u dU'&i 10% dol vai m{>t trU'ang hQ'p hay quan • Cac tntb'n( hQ'p co gia trj thieu cho cac bien pht,~
sat thll'c:ma c6 the be qual nsoCjli trll' gia trj thifu xuat thu{>c thU'bng dttQ'c bo de
tranh sv gia tang gia ~o
hi~n trona hoim dmh khong ng§u nhien cv thi cac moi lien h~ v&i bien dOt ·~p
• So IU'c;ms eac trU'ang hQ'p kh6ng gia tri thieu phai du cho • Khi bo m(>·t bien, hay dam bela r~ng luon co cac bh!n
ky thu~t ph'n tfch da chon neu
cac gia tri thtiu khong thay the co
moi lien h~ cao de
dc;ti di~n cho n(»i dung
dU'Q'c thay the bien ban diu
• cac bien hay cac trU'Ctng hc;rp v&i 50% hay han gia trj • Hay luon xem xet tien h~Jnh phan tich ca co va khong
thieu nen loii~i be. Cac bien vai 15% gia trj thifu Ia cac
cac trttb'na; hQ'p (cases) hay c:ac bien dltc;YC l~i bo d.1
Ll'ng vien cho vi~c loCjli bol nhU'ng cck mll'c cS~ &ia trj
xck dinh cic khac bi~t
thiE~u cao han {20% den 3~-) thU'ang dU'Q'C thay the

Nhdm 8 C~CK26 15 Nhdm8 C~CK26 1~

4
 3/11/2017

tic vAN Bt utN QUAN BtN otrueu THIEu

(MI.uiag DtiM)
I 04 GIA BJNII THONG d QUAN TRQNG
(IU I-.porttutt ~AMI'IIf/ldll•)

1. Phan ph61 ca.An - Nomallty

------------------.
• PhU'O'ng phap tic~p c~n 41 cai thien vif.c thieu
d cr lieu
-sli dvns each thLtc chi quan sat khi da co
day du dtt lieu
- X6a bat nhCi'ng bien s6 ho~c tn.tang hQ'p rna
ho
z.tcor" -3 -2 -1

Nham !I CMAPMICK26 17

04 GIA BJNII THONG d QUAN TRQNG

(IU IIIIJiol'tlult SMti61blh6ullf/ldllll)
1. Phan phOI cbutn - Nomality
04 GIA BJNII THONG d QUAN TRQNG
1(04 IIIIJIOrtiJllt ~Amlllf/ldllll)
1. Phan phOi chuin - Nonnality
I
(II) Normal 01e lktir) rc> Po;lbwlw •t.Md
YMn
Tlnh ch4tf'ua phan phoi chuin
( 1) Ham m~t d¢ xac suat ala doi xung quanh gfa trj
trung binh.
(2) Xip xi 68% di~n tiott dU'&i dU'ang pdf nc\m trong
khoanom:ts
X Xip xi 91% di~n tich nim dU'&i dU'ang pdf nAm trong
khoang m.t2s
Xip xi 9t,7% di~n ticb n~m dU'&i dU'O'ng pdf n~m
trong khojng m±3s.

1.2 Cac df}ng phan ph611fch trai va l;ch Jlhal so vm

IWNi~..Qinh thflimg 19 Nh6m 8 C~Ck26 20

5
 3/11/2017

1. Phan phOI chuin - Normality

Di) lfch (dewness) do h.r<nlg d<) lech cua phan phoi -De) nhqn (kurtosis) do IU'ang mtic d¢ nhQn hay b~t
v~ m¢t trong hai phia. cua phan ph6t so v&i phan ph6i binh thU'6'ng (c6 CJQ nhQn
-Ph§n ph611fch trai (ne~e skew, left.skewed) khi bang o).
duOi phi a trai dai hO'n, ya phAJ, l&n so lieu t(lp trung a Phan ph6i c6 dfng nhqn khi gill tri kurtosis dli'O'ng va
phi a phai c:Ua ph~m phOi. c6 d4llng bft khi gia tri kurtosll am.
- Phan ph61 lfch phai (posltlvf sknew, right4kewed) khi
duoi phla prnli dai han, va phAn l&n so li~u t~p trvng a phla - V6i phan ph6i binh thU'bng, gla tri cua de) l•ch vi dq
trai cua phan phoi. nhqn bang o
Khllfch phai, gia trj •tmewness durc:mg;

- m.,
- Khllfch trai, gia tri sk.wness am.
~~ cimg /an thl fif trf sknewness c4fl# /fin hem 0.

Nhom a CHA!i!IHCK26

y y
2. PhU'ang sai khong thay d6i -
(B)
Homoscedasticity
Ban cfljt
Xet vi du mo hlnh hOi qui 2 bien trong d6
bien ph\1 thuQc Y Ia ti6t ki~m cua he) gia
dinh va bi~n giai thich X Ia thu nhap kha
dl)ng cua ho
gia dlnh
0 X (J X
~ ~ ~ ~ ~ ~
Hlflb2..1: (a) PhU'O'Og sai all sai s6 khong doi va~b)
PhlJ'O'Ilg sai Cl)a •I s6 thay dol
24

6
 3/11/2017

• Hinh 2.1a: TfM ki•m trung btnll c6 khuynh hU'angtlng theo

thu nh~p.
3. PhU'O'RI sai cua sai so thay tfoi ...
• Tuy nhien mtc ell> dao dQng gi(ia tiet ki~m cua t{mg hQ gia HETEROSKEDASTIOTY
dlnh so v6'1 m6'c tiet ki~m trung blnh kh6ng thay ct61 t~i mQi
mlt'cthu nh$p. Nguyennhan
• E>ay Ia trU'ang hc;rp cua phli'Oh sai sai so (nhllut khOng
dOi, hay phwcmg sai bang nhau. • Do tkh 'OV kinh nghiem ma sai so theo thO'i
B(ulJ =a2 gian ngaycang giam

Hlnh 7.1b: Ml)c dQ dao d¢ng giG's tiet ki~m

• Do ban ch«t cua hi~n tiiQ'ng kinh te
cua tung hQ gfa c.1inh so v&i mU'c tl&t ki~m • Cong cv v~ thu th~p xU' Jy so lieu cai thitn din
trung blnh thay c.16i thea thu nh$p.
E>ay Ia truang hQ'p phll'ong sal cua sai s6 den sai s6 do lliO'ng va tinh toan giam
thay dOi.
E(ulJ = Cf/

Nhdm 8 CMAI)t~J(2 6 25 Nh.im 8 CMAJ;MiCJ(26

'?

Nguyennhan H~u qui cua phmmg sai thay doi

• Trang m~u c6 cac outlier (gia trj rat nho ho~c 1. U6c lu<;rng OLS vdn tuy~n tinh, khong ch~h
rat 16'n so v6'i cac gia trj quan sat khat) nhung khong phai la u6'c lu<mg hieu qua (vi
• Mo hlnh hl>i quy khong dung (d~ng hlim sai, phuang sai khong nbo nhfrt)
thh~u bien quan trc;>ng) 2. U6c h.rc;mg phuong sai cua u6c hrqng OLS,
• Hi~n tU'Q'ng phliO'ng sai thay doi thllt1ng g~p nhin chung, se bi ch~cb.
khi thu th6p so lieu chao (thee khong sian) 3. K~t qua dv bao khong con hi~u qua nil'a khi
sir dl)llg cac u6c Iuong OLS c6 phuung sai
khong nho nhAt. ·

Nhcim 8C)!AI)iii!ICJ(26 27 N hdm 8~CJ(26 21{

7
 3/11/2017

I>Q TUY{N TINH Slf VANG M~T cuA cAc LOI rlfONG QUAN
• Xac djnh moi quan ht pili tuyen tinh:
- Ki!m tra bi6u d6 phan tim : Bi!u dd phan tan c6 tb6 hi8n thi • Xac djnh ~At tai nrong quat• ;
cac duang thing mo ta m6i quan h~ tuy~n tinh, cho phep cac
nha nghien ctiu d8 xac dinh t6t bon bAt kY dac diem phi tuy~n Ddu tien cac nha nghien c(ru phat xac dinh cac nsuren nhan c6
- Phan tich hAl quy don gian va 1d4m tra phfut du
th8 gay ra. Gia tJi cho mot bi~n neu duqc nh6m ~~ h$ duqc d~t
- Mo hinh th.ay 1h~
thu u,r tren cac bi!n bi nghi ngo va sau d6 kiSm tra cho bAt kY mo
hinh
• Bi~n phap khac ph1Jc : • Bi~n phap kbAc phyc :
Chuy$n &\i mot hoac ca bai bi~n dS d~t duetc dQ tuy~n
tinh, ~o ra cac bi~n mm dS d~ di~ cho phfin. phi tuyfn tinh
Cac phuong phap ph6 bi~n n.llit Ia S\1' b6 sung cua m9t bi&l de
phan tich d~ ditn cho y~u t6 luetc b6

Nh6m8CMA~6 29 Nhom 8 CHA~!C26 30

KiEM NGHteM CAC GiA f>INH TH6NG KE

1. Phan bo chiJ'n (normality) b[ anh hV'hng nghiem trQng trong m~u nhQ
1. Phan b6 chuan (normality): histogram, (<50) nhlfn8 tl~ d9ng giam dan khl c& m§u:::: 200
skewness, kurtosis, Kolmogorov-Sminov 2. Phan l&n trU'anc hgp cua phU'O'nS. sel khong dong nhat
(heterocedtstldty) Ia ket qua cua ph.Ot'l bo khong chuan tr~:>ng m9tt; hay
test, Shapiro-Wilk test nhieu bien; clo~y, vi~c kh~c ph!JC pb&n bo chuan Ia kh0n& can
truet do
co m~u, nliU'r!S ~i can cho phlfO'nJ IIi dong nhat
2. PhU'ang sai dong nhft (homocedasticity):
3. Cac quan ht phi tuyen c6 th~ dlfqcaJ'm dang k~ khi dil' li~ d~J'9C bien
scatterplot, Levene test doi thimh ~·118 tuyen tinh hay ca.C tf'inh phan mo hlnh "U'Q'C _$t)o·d~,JnJ d'
th~ h i~n ph8n phi tuyen cua quan nt,
3. Quan h~ tuyen tfnh (linearity): Q~Q plot, F 4. Cac sai so tU'cmg quan gia tang tCr qv:6 trlnh phai dlfQ'c xU' ly nl'i11 gia trj
test thi~u tlic ,. nht nghien cliu phal xat~lnh cac nguyen nhall gl[)'a Gic: biin
nQi hay nge)•f@ t~p dllli~U; neU ch(ing khong dlfQ'C phat hl~n ~ $llrt
4. Khong c6 cac sai so tf/O'ng quan (correlated chlla, cac tl'la• kien nghiem trc;ma o6 th~ xuat hi~n trong cac ~t qua

errors)
Nhom8C~006 31 Nhom8C~006

8
 11/03/2017

PHAN TiCH NO/DUNG

f----··--···-·-··-·<-·-··-·-· -· -~·G-··-----···--........· --.......-·~··-·-
HOI QUY 801 ..-=:., •
1. HOi quy dan
r--······----·····----·,--··-·-·---···-····0;-··-···· -····-········-· -·-········-··-············

2. HOi quy b¢i

GVHD: NH0M4 3. S'f khac nhau giCra hoi quy dO'n
Dllon9 Sfch Hu~ (NT) 7701260637 A
TS. f)o~m Thanh Hai Ph~m Lan Huang 7701260658A va hOi quy b¢i
Trinh Huy Hung 7701260644A
NguyAn Thi Phuong Thao 7701261035A
NguyAn Thi Hoang Yiln 7701261283A 4. Chey s6 lieu tren SPSS

Nh6m 4 CHADSNCK26

MO HiNH HOI QUY MO HiNH HOI QUY

-·------·---«-.·-·--·-·----~)-··-··-:-·--·--·-.,_,_··-···--·-···----
Trang nghitn cuu, chung fi1hU'cmg phai ki6m c:Jinh cac
~~~;;-~~-;~;;~Oi;----G---~-~. -
gia thuyet v6 m6i quan h~ giila hai hay nhi~u bi~n, trong • 01. Chi phf quang cao cho thU'«:mg hi~u c6 lam tang doanh thu
do c6 mot bi&n phl,l thu¢c, mOt hay nhieu bien d¢c l~p. cua thU'ang httu khong?

• Q2. sv th6a man cua khach hang v&i thU'O'ng h~u co tac dQng
vao long trung thanh cua he;> <f6i vai thU'ang hi~u khOng?
rKhi c6 m¢.t b~n do"? r Neu c6 nhi~u bien """'
l~p. m6 hinh dU'O'C doc t~p. m6 hinh
goi Ia mO hlnh h&i duQ'c gQf Ia mo • Q3. Tfnh h6p dAn cua cong v~e va tien IU'ang c61ao <Jf)ng vao
hinh hOI quy boi xu hU'&ng nghl vi~c cua nhan vltn khong?
quy don SLR
(Simple Linear MLR ( Multiple
\..Regression) ~ Linear ~egression) • Q4. Tinh ki&n .,inh trong hQC t.p. aQng CO' hQC ttp va hlnh anh
'- ./ tru>ang d~i hqC c6 tac crc;,ng vto ki~n thl.fc thu nhfln cua sinh
,_Nh6m 4 CHAOSNCK26
vien khong?
Nh6m 'I CHADSNCK26

1
 11/03/2017

MO HiN~tfO~ QUY _ Hai quy

Tra lai:
• H1. Chi phf quang cao c6 quan h$ c(lng chiilu v&i doanh llu mot thU'ong
hieu.
• H2. Sl,l' thea m~rt cua khach hang v& thU'ong hieu tac dOng tac dong cung
chi~u vao long trung thanh cua hQ a6i voi thU'ong hieu.
• H3a. Tfnh hap din cua cong viec t6c dong nghich chitu vao xu hU'&ng
nghi viec cua nhtn vi em.
V&imo V6imo
• H3b. Ti~n IU'ang Ulc dong nghich chitu vao xu hU'&ng nghl viec cua nhan hlnhgia hinhH3,
vil~n . thuy8tlll , H4, tadimg
• H4a. Tinh kiem Omh trong hoc tc;ip tcie dong cung chi~u v&o kiiln thuc thu H2, tadtinJ mo hlnhh6i
nhc;in cua sinh vltn. mo hlnhb6i quybqi
• H4b. Dong co hQC tc;ip tac dong c(lng- chi~u vao kiiln th.:tc thu nhc;in cua quy don MLR
sinh viem. SLR
• H4c. Hlnh anh ella trU'ong Dc;ti hQC t•c d¢ng cung ch i~U vao kiiln thl1c thu
nhan cua sinh vi6n
NhcSm 4 CHAOSNCI<26

HAi~y 1. Hoi quy dO'n

1<--·- .. --;.._----··-·-···--·-----\@y-----------------------------·----- ~-----··-----··-----·· . . . .~------------·--G}·---··---------------~--·----
y i = a + bx j : Gia tii u&lluQ'Ilg ph\l thu~c Phuong trinb hOi quy: ~ = f(X ) +6i
thong qua2 tham s6 a va b 1

yi : Ja gia tri rna ta quan sat dm;rc = f3o +4X; +ei

a=.x-.Yi : 1a phan du, nota hitu s6 gia tri di~m quan
Yi la gia tri. cita Y t<;ri quan sat thir i: 1,2, ... ,N (N 18. kiob thu6c
sat va gia tQ di~m tien lu<mi :

dam dOng)
Tim a va b sao cho tdng binh plwung d la nho nhdtl khi x; : Ia gia tri. c\la x t<;ri quan s4t tbu i
tto gia tr! nhin thdy va gia trf tien lu(mg la thdp nhdt
ei : la sai s61a,i quan sat thir i
Vqy ta u&c tlnh m6i tuong qua1t46 chinh lata tinhR a : Ia h~g s6 h6i qui
Po _ft
A : Ia trona 110 h6i qui
he
M6i quan ~X va Y Ia m6i quan thdng ke, neu ~::;:() thl X,
M.quan he ham s6, va J;=4+4X Ia ph§n cAu trUe ciia mo

- Nh<lm 4 CHADSNCK26 ~--------------------------------------------1

NhOm 4 CHAOSNCK26

2
l A
1. Hoi quy dCYn 1. HOi 91'Y dcm
~-----------------"---··-----·-··-·G···----·---·----··--------·-----
1. x va Y c6 quan he tuy~n tfnh
--------..~~~~~;~-;.~N-;;-;LR ·---~
2. Y Ia bi~n (fjnh luqng va li&n tr,lc Chung ta cAn dlmh gia ve 4am dOng, nhung ta khOng c6
3. Cac quan Ht Y Ia d¢c l$p thOng tin nay, v•y ta dung thOng tin mau d~ uO'c iU'Q'ng, ki&m
4. Cac gia trl X c6 dinh Y; =Pfl + fl1X1 + 6 i dinh thOng tin vt dam dOng.
5. X duqc dO lt.fO'ng khOng sai t6 Tham s6 va trong s6 /3 0 ,/31 ®a dam dOng, ohung ta khOng
a. Sai s6 Ej c6 phAn b6 chuan th~ c6 dU'Q'C chung. Do d6 ta dung tham so tU'ong ling ttl .
7. Cac gia trj X khOng co quan 11$ v&i Ej mau d~ U'6'C luqng tham s6 cua (lam dong.
a. Phuong sal cua cac sai s6 s; IS hang s6 Co hai phU'ang phap U'6'c IU'Q'nQ' tham s6:
9. KhOng c6 t\IO'ng quan chu3i - Binh phU'ong be nhat OLS(or<finary least square)
- PhU'ong phap xu hU'6'ng CI,J'C ~i (maximumlikelihood)
1-
Nhdm 4 CHAD5NCK26

M6 )NH HOI QUI ON MO HiNH HOI QU flON

f----·-·-···~--··----··....,..·-···-·---··--·-·~}-···--··-------·--··-----·--··-·-- ··--------···-·-·~··-·--·-··-----·--~r-----··---·---··---~,-...--··----~·-·
H$ so xac dinh R2 (coefficient o~rmination) cho th'y bao nhieu He sotu'dng quan Pearson tr~SS
% biE3n thien cae Y c6 the giai thrch bOi sl,J' ph1,1 thu¢c oua Yvao x Ana/yze---Corft!late---Bivariate-(tai ct)t Correlation
Coefficient) chpn Pearson---(/:11 Test of Significance) ch(Jfl
R 2 = SS - = I _ SS" two-tailed
ss y ss y

, -
ss R 2 = L (yi - y ) 2
; "' 1

ss E
2
== r e;
n

f;::; l
= L (y
i=l
n
i - yi)
2

DQc ket qua;

~mpg va horse c6 M s<> tudng quan 1:1
H~ so R2 chf"h ta binh phU'O'ng ht so tuO'nQ quao Pearson r giCI'a ~sig
= .000<0.05--chap nh~n
X va y (R2 ~) = ~mpg va horse Q: •.nt
---tu'dng quan nghjch chfeu
1--
Nhdm 4 CHADSNCK26 NhOm 4 CHAD5NCK26
 -~ ..
1tl1:~a . . . . . .I ;--:. ·( i 1

iN I Hbl I
··-··---------------···-·-·--------···-·(@)-···-··--------·--···~-. . . .-~--~~

• Ki~m nghi~m F dU'Q'C Sli dvng nham

ki~m nghlem gia thuytt v~ Slf phu hQ'p
a. Predictors: (COnstant), Horsepower
cua mo h1nh (3, p. 218)
-chi so R Square: y nghia Ia c6 59,5°/o bien thien cua • Ki~m nghi~m t cua gia thuyet v~ yngh1a
X t<;~o bien thien cua Y ( hay dQ lan cua mOi quan cua he
s6 hOi quy J3 1 (3, p. 219-221)
h~ giU'a X va
Y)
-R2 : cang gin 1; mo hlnh phu h<;1p ; cang gan 0, mo
hlnh kern phu h<;tp

NhOm 4 CHADSNCK26

· l>a cqng tuyen

2. HOi qul bi)i MLR ~il\
~--------------···---~·-------------(@J--·-------------···--·------------ !--··---------------··-··-···---------------~ k-·------------------·~...........__,_.,.. ____ ~----~-
Ml}C dich: 8i6u dien mo('®an h~ giCfa hai hay nhieu • Kip. u6c lu~g mo hinh }i ui hoi chimg ta phai
bien dQC 1$p VCYi mQt bien phl,l thUQC djnh lll'Q'ng kiem tra gia dinh cac bien ~ lap khop.g c6 ttr<mg
quan hoan toan v6i nhau th&lg qua kiem tta hi~n
Mo hinh: tu<Jllg da cc)ng tuyfut.
• Da cong tuy&t la hi~n tu<;Yl\g trong d6 cac bien doc
l~p c6 quan h~ v6i nhau, c6 til ba bi6n d6c l~ o6
tua:ng quan nh.au.
Cac gia dlnh: nhU' h6i qui dan nhU'ng xet th~m hi~n • D8 ki8m tra hi~ tu<l,llg da ~ng tuy6~ chi sA
tU'Q'ng da ct}ng tuyen thuemg d-lltl8la h~ so phong tW phua:ng sai VIF
(Variance Inlation Factor)

f- ~
Nh6m 4 CHADSNCK26 Nh6m 4 CHADSNCK26
 He SO xAc DJNH DIEU CHiNH R;dJ
Trong mo hlnh MLR, vi c6 ~hi'eu bien c1¢c ~p, nen ta iE B NH TRQNG H I Ul B - I
-------- -----------·. ·-w-~------·------e--·---------------·. . . . -~~--~--~
phai tfnh h~ s5 Cfieu chinh R;dJ ere thay cho R2 :
1

•Kiem tra nhlltlg mo hinh cO nhieu biin phl:J thUQC • Kiem tra hi•n tuqng Cia cOng tuy~n (multicolinearity)
' nhu'ng thi!C Sl/ trong <16 c6 m¢t s5 bien khong giup (2, tr.496-497) (VtF>1 0)
cho vi~c giai thfch bien thlen cua Y. • Kiem tra Slf phu hqp mo hlnh: F, t, Levene test
• Kich thu6'c maun ~ 50 + Sp
p: s6 luqng bi~n Cl()c l~p
• Phan tich vai SPSS (3, tr.236-284)

Nh6m 4 CHADSNCK26

i~ DINH TR N I · -I
Dl T (~}...
·- c~-c-s·o-n~uaiinFl--"F---allq;·c-~ -b~n9-pFlantTch pl1wn9-sar
ANOVAv&i:
1. 1. Kiem dinh dQ phu hqp cua mo hlnh Mean Square: 4.1¢ 1$ch binh phuong binh quan
(Phan tlch phliong sal) F: aU'gc tinh trvc tiep tti ti s6 giCra dO l~ch quan phoong hOi
qui va a¢ l~eh qutm phU'O'ng ph!n dU' .
• 1he
hi~n a c<)t ap ch6t cua bang phan ttch
Fw:- ESS(k- 1) ~F,___(k -1 n-k)
Gia trj F dU'Q'C

TSS(n -k) '

a
ANOVA v&i m(rc y nghia quan •at Ia Sig c<)t cu6i cung.
V&i:
Sig < o.os: t>ac bo gia thuyet Ho =} R2 1; o: MO hlnh hOi quy
{ OHo: R2 = tuyen tfnh cru~c xay dl,fng pH~ hQ'p v&i tOng tht.
Kiam dinh gia
thuy~t
H1 :R2 '1-
________ 0
NhOm 4 CHADSNCK26 l9 Nh6m 4 CHAOSNCK26

'
''
!.
 11/03/2017

u1 eor
"' iEM f) NH T 0 I U BQI
··-z.--Ki!m-a)im gll.lhliYit&9hia-cua·lit·iflil;rcfur-··--- ·--···------·--·----~11--·-·· ---------------{&·------------·----··_...--"--·-----~-·- ·-
M~c du mO h'lnh hOi qui tuy'n nh mau ta xay dyng Ta ky vQng gia thuyet nay si bi bac bo, vl n&u 131 =o
dU'Q'C c6 gia trt h$ s6 do dOc S1 ¢ o, nhU'ng ta chU'a th~ nghia Ia Y * l~p v6'i X hay X chang c6 anh hoong gl
den Y, nghfa Ia m6i tU'O'ng guan tuy~n tinh ta nh~n
ch~c h$ s6 <SO dOc cua me hlnh tong th~ da khic o.
Viv~y. ta phil lam ki~m dinh 46 c6 th~ ket 1uan v6 131 thay a mau chi xay ra do nglu nhien chll' khOng phai
ban chat, mb hinh hOi qui tuy6n tinh ta dA xay dl,l'ng
Gia thuytt: Ho : J31 =J32 =._,.= J3k =o dU'Q'C dl,l'a tr6fl mot m6i quan ~ "gia" giQ>a hai bi,n.
H1 : 3 J3j :t: 0 (2 ~ j ~ k)
Gia tri th6ng k~ t va mll'c y nghia hai phia quan sat
dU'Q'C cua ki6m d!nh t aoi vO'i gia thuyet v~ cac h~ s6
Ki~m dintt t:
t =
B1 hOi qui thi hi$n trong hai c¢t cu6i cung ciJa bang
s Bl
Coefficients:
V&i mll'c y nghia Sig < 0.05: bac bo gia thuy6t Ho: X
Phan ph6i cua d~i IU'qng th6ng ke nay Ia Student va Y khong c6 quan h$ tuy~n Unh v6'i nhau.
v6'i N-2 c tv do. _ _ _ _ _ _ _ _ _ _ _ _ _,
NhOm "' CHAD5NCK26

3. Kiim a!nh sv pJlp hQ'P cua gia d!nh 4 . Kich thU'&c miu
lo--··----------------····-·~·-··---------{{ijjr--·---------------·•,___··---------------- n: ki~h thu'6c m§u tbi thi~u cAn thi6t

... . .. ..... . n;;::50+8p

~lz, eJ-z;. -
{ P: sO l~ng bil!n di;lc l~p trQng rnO h.lnh

0 .-~.~
.........
• • ., J'.
• .::...--".~.--=-­
• ••
.·.·. :
0 ....______..L..t-:.L.:!!!.·~-·......
• ·-.-::•;...._.- v&i:

'----- - - - y ,
.a. Phu h9p
..___ _____
b. Phi tuyQ'n
•;
;, Cong thU'C tr~n tU'O'ng dOl phU hqp neu p<'7. Khi
p>7, cong ttlll'c tren hO'i qu4 kh~t khe. Tuy nhi6n,
trong dli titu khao sat, chung ta thU'6'ng d.ung !FA

... . ..•
cung v6'i MLR trong mot nghi&n cll'u.
..............
clz.
Kich thlJO'e mAu tinh tU> c6ng thll'c trem thU'Cmg nho
0 t-----.:1..~
••~
...:·
.............~.-
...
. ......
0 t---!!•c.-:·~·~:__.•._·~·­
.... hO'n kich thU'OC mau doi hOi cho EFA, cho n&n
chung taco th~ dung n6 tinh kfch thU'&c mAu cho
.__ Y; MLR va SO sanh l~i kich th\IOc mau doi h6i cho
4. J->ht tuy;., va phU'OI'IS •ai thay dol
EFA.
Quan h6 giO>a y· va ph an d!.i J,Ngu6n: Stenvens)
f-
NhOm "' CHAD5NCK26 Nh6m "' CHAD5NCK26

6
.. ..-:·.
 11/03/2017

3. sv kh8c nhau g/()a hoi quy d0'11 va

PhU'ang phap phan tich hoi q1Jlb9i
f'-·-·-··--·--··--·-·-·----··--···-·{&·-··-·-·----·-...,....._______ ··---··-··- ,__•-·•-•-••·-·---••11-••-··-··•-•-••·-·(~••-··-·-·-··---.-..,___..~..,~-~--~Y-

ph6p dOng thc7i (PP ENTER trong IPSS)

.
1. PhU'O'ng / ....-··-··-....,.~ ' ·..,,\
Phuong phap nay dung d~ ki~m dfnh gia thuy~t suy diln tll' ly thuy~t.
2. PhU'O'ng piJjp t.:rng bl~VO (PP STEPWISE trong SPSS) (\_ r•,/'
\
/
Phuong phap nay dung d~ kham ~a m6i quan he oua cac bi~n -------~---··-·"'-~·~----l
trong thl tr11tmg Nghien cuu sv 1
phl:J thUQc cua 1
3. PhU'O'ng phap thll' b'c (HIERARCHICAL) hai hay nhitu 1

Trong phU'ong phap nay, nM nghJ.tn cliu se chon biAn doc lcflp nao
bi~n phlsJ thugc
vao m¢t bi~n
dua vao mO blnh truoc va bi~n nto dua vao me hlnh sau. Viec tva d¢cl~p
chon nay o61hi Ia do ly thuy~t cQng c6 th~ do danh gla cua nha
nghi€m cliu.
NhOm 'I CHADSNCK26 'Nh6m 4 CHADSNCK26

3. sv khsc nhau gJO. hoi quy tl0'11 va 3. sv

hoi q b9i
2

cau QL Slf tnuan ti~n c6 anh Q2. 51/ thu~n ti~n, Qli ca, d!ch V1,J
h6i hU'ang dfn long trung th~mh kh6ch hang, thlfdng hlf4i hay gia ca c6
cua kh~cln hang khong? ..,.. hu'Ong den long tN'lg thanh cua bi€n d<:)c ~p kbOng c6 quan '
kh6ch htmg khong?
nhau hoan toiul:
(Xi, Xj) # 1, v6i :i#j
Gia Hl: 51/ tttU~ ti~n c6 anh li2a: 51/ thu~n ti~n ¢.,., hliang den l, ....p;j~t, ... ,p)
thuyet hliang din long trung thanh fbng trung thanh cua khich hang
d~t ra cua khcich hang H2b: Gia ca c6 anh ~g den long
Ltung thimh cua kha<::h ~g sA xac dinh di~u chinh R2ad)
H2c: D!ch v~,t khaah ~ng c6 imh
sosanhmb~
hating den long 1:I'UrfO thanh cua
khich hang
H2d: Thlfdng hi~u c6 6nh hliang den ki&n tra men tuq eta
terlg trung thanh cua kh6ch hang tuy~n

Nh6m 4 CHADSNCK26

7
· . ,. J
 11/031201 7

Khaini~m

• Phan tich bift s6 (Discriminaltt Analysis -DA) Ia m(lt Icy thu@t

phan tich dtrqc SU d\lDg d~ tim tfajy S\f khac biit giita ClC nh6m,
d6i tuqng bAn, each phan tich di Ji~u v&i:
- M(}t bi~n p•v thu(}c la bi~n .ifi:trh ttnh, dung da phan lp3i <t6i
tuqng duc;yc do bfutg thang do dinh danh ho~c thu t\f.

- Oic bi~n tiOe IJp la bi~n ainh, lwng, la m9t s6 dlc tinh dung de
PHAN TiCH DIET
•
s6 phan tich $\1' khac bi~t giiia ciao nh6m d6i tuong duvc do bAng
(Discriminant Analysis-DA) thang do kboing hay tY l~.

Nh6m 3 ¢1>1AC';l5NCK26

Khainiem . Ml)C TI:Eu ciJA PHAN TicH nrt:r s6

• Phan - bi~t s6 bai nh6m (two-group 1. Xac dinh Sl! ton t~i khac bi~t c6 y nghia
Discrinunant Analysis): bi~n ph\1 thuQc co 2 giua cac nh6m theo cac bien dOc l$p
bi~u hietl. 2. Xac dinh nhling bitn aoc l~p t~o khac
• Phan ttclll bi~t s6 b()i (Multiple Discriminant bi~t l&n nhat gilia cac nh6m
Analysis - MDA) : bien ph\1 thu()c c6 3 hay 3. Ph An IO(Ii cac quail sat
vao nhom theo
nhi~u bi~u hi~n gia tri c a· ·bi~n: t!Oc l~p
4. Danh gia tlnh chinh xac cua vie.e phan
lo~i ·
Nh6m 3 ClrJA(!l6NCK26 3 Nh6m 3 ¢1HA06NCK26

1
' ·.• .',•, •'•' '· ................. ~ :--.: '· .·: ~·...'.'.'.'.·.·.·~·~·· .< •· .... .•.•.·:.·.<::·.·.·.·. :.·--:·.-.,;~..-.. . o T o ,T, '
 11/03/2017

Cac tham s6 thBng ke trong phan ticb biet sa

M& hinh ph,AD tich bi~t s6
• Canonical 'cu11'1tation: H~ s6 tuo'Qg quan canonical
• Centroid: la trung binh cl.la cac gia tri tli~t s6 trong m6i bh6m.
• Classificat~on matrix: rna tr~ phdl:l J041i ( rna tr~ dv doatl.)
• Discriminant faaction coefficient$: :W s6 hiun pharr hie\; ( ch'IXa chuAn h6a)
• Eigenvalue: ti ~ gitta tbng cac d¢ l~h binh phtrong gi'lla de nh6.m va t6ng
Trongd6: cac do 11$ch bi.nh phuong trong noi bo B.b6m( SSG/SSW).
Zjk: Di~m pbin bi~t Z (bi~n ph\1 thu{>c) cua hiu:n pha.n bi~t j • F value an(! tlldr significane: gia ~ F duqc tinh tir ANOVA m()t y8u t6
cho d6i tuqns k • Group means tnd group standlird deviations: Trung hinh nh6m va dQ l~b
chufut nh6ri1
• a: di~m chan • Pooled withbi..group correlation nt•trix: rna tr~ ttrong quan n6J b~ nh6m
wi : Cac h~.J6 hay tn;mg sA phin bi~t thu i (i ~ 11n) • Standarze(J dberiminant funct'jon coeficients: cac M s6 ham phful bi~t
chufuth6a
• xi : cac bi~n dQc l~p thu i <t6i tuQ'ng k (i = 1,:n)
• Structure correlation: ttrong quan ket cAu
• Total correlatton matrix: rna t.rW.l tu'emg quan toan b(,)
Nh6m 3 ¢1i-!AC',leNCK26 Nh6m 3 CIIIA05NCK26

M6i quan hf giii'a phan tkh hfii quy,rAn tich Quy trinh phan tich biet s6
ANOVA va pbAn tich bi~t s

H81Qul AM ova PTbi~~J6 PT bi~t s6 b(li

l.dl~m glB!!K nb~

sa hrqng bl~n ph, tin. M(lt M9t MQ( Tit 3 bifu! tro len Cac gia dlnh c\J,a naJrien ciru
sa luC}'IIg blin dQc' ltp Nhi&i biSn ~biSn NhiBupi6ft NhiBu biSn l1ac lu(}llg mo 1UM bi~t thirc, danh. gia tin1t doog
dAn cila t6ng th4 •
l.di~m kbac nhau
Gim thich k~ qua
Bin chit bi@n ph9 (h. £>jnh luqng *hrqng Phanl~ Phan Io(li

Bin chit blin d\)c lfp £)jnh luqng ~~0~ £)jnhlu<;i!!g £>jnh hrc;mg Drum gia k~ qua
Nh6m 3 CliiAt(iNCK26 Nh6m 3 C~CK26

2
 11/03/2017

Xac djnh v4n d~ nghien evu 2

~.
Thiit k~ nghifn c.ru
• L\fa ch-.;11 Win ph\1 thu9c Va tlfc lip
- Bi~n phy t~c: . s6 luQ'rtg cac nh6m bi8n ph\l th.u{>c (bi~u
• Danh gitt S\1' khac bi~t nhom tren mot ~p hQ'P hi~n) co thA » hai ho~c nhiau bon, nhung cac nh6tn nay phAi
da bi~rt duqc lol;!i trit lin nhau va diy .du~ khong gi&i h4Ul cac bi~u
• Phan lo~i d6i tu<mg thanh cac nh6111. hi~n trong cac bi~n phl,l thuoc
• Xac dihb kich thu6c ·cac nh6m. - Bi~n d9t ltp : Bi8n doc lip thuOn.g duqc Iua chon tbeo hai
each. Phu<mg phap dftu tien ]$ tir cac nghien c(ru tru6'c d6.
Phucrng phap thu hai Ia tnrc giac.

Nh6m 3 CHAt';!SNCK26 9 Nh6m 3 Cl>tAC)5NCK26 10

ThUt k~ nghifn c.ru 3 C'e gia djnb nghien clfu

• Kich thu6'c mlu
• Tac dQng d6i vOI d\1' toan va pllln Iotti: Cac gia dWh chflili cita chirc
- Ty 1~ hQ'P ly la 20 quan sat cho m3i biin d\l doan nang phfut bi~ la phdn pMi chwin da btin cua cac biin iJ9c l4p va c6 d9
• Kich thucrc t6i ~u la 5 quan sat cbc) m6i biSn d<)c l~p. phan tan cung nlnr cdu true hi?p ph'lltlfig sai (ma trqn) !)ang nhau dOi vai
Luu y ding It t It m)y ap d\Ulg cho • ca cac bi~n tinh toin trong phan cac nhom theo 4inb nghi'a bi~n ph1,1 th..OC
rich, ngay ~ kht ~c bi~n khong ~ nh~p vao cac hiun bi~ thirc. • Nh~ d'ng de l'i ph'm gia djnh: Box's M test nh~m dimh eia tAm qua:n
• Chia mlu : phin chia m~u thilcll bai ph~n, ml;>t dw;tc Bl't d\Ulg d~ ucrc trQng cila S\l kh4c biQt giua rna tr~ hi~ phuong sai giua cac n)l()pt.
luqng ham pQaJt bi~t duQ'C gQi 1<\ mJu OOC }l,IQ'llg hay mlta phan tich va • Tac dQng d6i vlri tiJ(re luQ'Ilg: oo li~u khong dap irng duqc gia dinh phan
mot dimg d~ Jri&n tra tinh dling d6n • ham phfut biQt qc gQi la m~u ph6i chuftn da bih co thS gay ra :®i&. vAn dS trong u:6c lu~ clia cac
ki~m tra. Kbi mlu du l&n ta co tha chia m~u thimh 2 pl\An bfulg nhau ham phan bi~t
Nh6m 3 C(!lfA()INCK26 11 Nh6m 3 CHA().5NCK26

3
.. .. ( .. ~ : •. :~ ••~- ••··' o;- ...•: .• ! ~!! •• ·::
 11/03/2017

~ CAe gia djnh oghien clfu 4 l1ft llrqng mo hiGh bi~t thtrc va danb gia
~
tinb dung din ctia tang th~
• Lva chqn plnrong phap U'hC lll'fllg:
• Tac dqng 461 v6i vi~c phan loti: Neu ca mtiu nhO va ma tnjn - PhU'ong phap aAng thiri (tiVe ~p) (simultan~ous (direct)
hi¢p phuong aai khong d~u, y nghia th6ng ke cua qua trinh u6c method):uoc luQ11g ham khac bjft khi toan bo cite bi~n doc l~p
lu<)'llg se bi 8nb huemg xAu. duqc dua vao d6ng thoi Thich b(rp phan tich toan bo cac bi~n
• Tac dqng d66 v6i vi~c ghU thkh: hi~n tuQng da cong tuyen d(k l~p va khong mu6n thAy Cat ket qua trung gian.
giua cac bi~ doc I~p. ki~tn tra tAt ca cac k~t qua d~ ki~m - PhU'ong pb'p U'6'c IU'qng trroc bll'6'c (Stepwis~ estimation):
ch1mg S\I hitn dien cua tAt ca cac gia tii ngo:;ri lai, nbim lo:;ri bo dua cac biin doc l~p vao ham khac bi~t da bien l)lQt cacn tuful
neu dn th\~t tv, duqc q(mg khi mu6n xem xet m6i quan he rnQt lu<;mg Iem
cac bien d~c l..p d~ dua vao ham.
Nh6m 3 CMA()5NCK26 13 Nh6m 3 CHA05NCK26 14

lf6-e Jtrqng mo hinh bi~t thtrc va d'Dh gia 4 lfac Juvng mo hinh bi~t thvc va c1anb gla
tinb ctung din cita tang th~ tinb dUn.g din cu*' ting th~
• Mlfc y ngbia tAng th~ (Statisdeal Significant~)
Phuong phap U'UC larQUg titog bU"ac thvc hi~ theo cac
~ Mlfc y ngbta tang th~ OVEJl.\LL SIGNIFICANCE: ap
bu6c:
d\Ulg cho ph1-t<mg phap u6c lu<mg d6ng thCri va p~uong pba.p ooc
1. ChQn bi&l don c6 tinh phatl bi~t nhAt. lu<)'llg timg bucn~
2. Ghep ndi cac bi~n ban diu v6i cac bi~n dOc l~p khac Fhm~~~~C phap U'Oc Plnrong phap tr6'c

li va chQn cac ~ t6t nhAt llfl}'lJ ctAng thOi llrqng timg bwc

3. Lva chQll tbem cac bi~n bA sung - Phep 4o IU'irng Wilks' .. ftep do Mahalano,flfs
I 4. Xem xet hoan thitnh qua trhlh lambtla
- Hotellhl&-.s trace
])t
"'Pttep do Rao's V
~ Pill.&'t criterion

Nh6m 3 CHACiliNCK26 15 Nh6m 3CHA()fiNCK26

4
 11/03/2017

uac Juvng mo hiith bi~t thtrc va c1 h gia .-=--

.!. Giii thich kttqua
tinh ung din cua tang thi
• Phep do ~g Wilks' lambda: l!i t5' ~ giil'a t6ng cac dQ l~ blnh phuong
trong n(>i bQ oao ah6m va t6ng d() l~oh binh phuong to!n ~. Gilt tri clia n6 Bi~n doc l~p v&i ~ .& phin bi~t tuang d6il6n hqn d6ng gop
nk giil'a khoAnf ~ d~n 1 v&i Wilks' lambda cang 16'!1 (gftn 'J) cho bi~t cac Discrimimtnt
nhieu hon vao viec - · dai cua ham (biAn ph\¥ tl!J.uQc ). (thuy
trung binh nh6nt duemg nhu khong kMe: nhau (nguqc 1~) weights d6i tY 1~ thu~n/nghich)
• Gia thuyit JiO: Trong t6ng th€ cac trutaJ binh cua cac hat'11 pbAn bW trong tAt
ca cac nh6m lj\ bing nhau: xae djnh M s6 hi'~ qUAil ~tia tirng bi~lQho tUng hlltnlgi:i
Teat of are df Slg. Discrimin•nt sa
trj bi~t tricfl
Wilka'
Do Ju(mg m6i 9uao ~ tuong quan giila tlrllg bien ~ l~p.
Function( Lambda loadings v&i ham bi~t s6
.359 .000 o
Gia tri t6t ta til .4 tr6' ten

~ Wilks' Lambda * 0.359 chuy~n thlWl•i luqng Chi-squate * 26,13 v&i 5 b{lc • Thvc hiQn bAng c~b ki&n tra cac d9 16rl tuyQt 4Bi cl\a gia illi F
tl,Ido F,..value va xep h~Jng cli~Jlg,
Gia n·j F cua hi4n m\o cang cao thi cho tl)ay nlng lye gW
~ Sig::: .OOO<.OS -> dU can cu bile b6 giA thuy8t HO tUc Sl,l pl)An bt~t co y nghi'a thich S1}' khac ~~t "ella bi~n cang )On
th6ngke.

Nh6m 3 CI>IAC6NCK26 17 Nh6m 3 C~~A~NCK26 1S

Vi DV MINH HQA
Daoh gia kit qua PHAN TICH Bil:T SO BAI NHOM :BANG SPSS

Quytrinh
Phuong phap pli6 b~ nltdt danh gia kat 9ua li viec <.tanh gia • Xac dinh vAJt d~ nghien cU..
tY I~ dfutg (hit ~tio) bay ph~n tram s6 quati. ~lit duQ'c phan
xacth\fC lo~P d(mg. Mu6n xac dinh dfe tr-ung nbi belt CUa cac gia dinh 4a d~n mQt
Cac dii Ii~u duQ'c ~hia 11\m hai phAn: m~ phiall tich dung dS
khu nghi maf (resort) trong vo~g OO.m qua z
TBi U'U bOa u&c luQ'Og ham phb ~. va miiu kiam tra dq d8 xay d\ffig
m~u ki~m tra rna trin phan lo~P:

Phuong phap cro~s-\'llfidation - ki~m tr~ ~: Vai tro cua 2

Cross-validaq(Jl mia miiu nay c6 tbt 4UQ'C thay dBi va vi(lo pi)b t1ch dugc )~p

I'
·~·
Tfnh ra mot di&n pha bi~t (thu<'mg Ia tni;tg bhab centroid cua
I Gin trj centroid cac nh6m).
Neu quan sat oo bifit ~ l&n han gia tti ttaY thi xep vao lo~i
centroid Ian va n~ J~

Nh6m 3 C~A()liNCK26 19 Nh6m 3 CIYAQ:$NCK26 20

 11/03/2017

vi DU MINH HOA vi DU MINH HOA

PIIAN Ti CH Bit£T SO HAl NJi6M BANG SPSS PIIAN tiCH B~T so
HAl NHOM BANG SPSS

• Thift kf ~-n cliu phan ~e.. 'l~t scl ~ Kich th'!ac mlu va chia n;.Aa:
~ L\fa chQn biltt d(}c lip va llifa ph\1 thu(}c: Ph4n 1lch bi~t s6 sir -Kich thu6c mlu: 42 ho gia d:inh
d\lllg 5 bitn ~c l~p d~ phan bitt d~c trung giita eac ho gia dinh
baog6m: -Duqc chia ra Jam hai phfut: rnlu phan tich va mlu ki&n tra
tuan.g img Ut30: 12.
L Thai dQ d6i voi du }ich (1haAg 9 difun)
-
-Viec chia mlu nay d\l'a trep bi!n phantich: gia tri o (mdu
2, Gfut b6 vOi gia dinh (thimg 9 di~m) f)~c tn,rng
~-,.. ki~m tra) ~ giatti 1 (mdu ph§q. tich)
3, Quy moho gia dinh (nguOi) ...... hQgia
dlnh
4, Thu ~p trung binh tMng -Cac gia <tbib 4a d8n resort trong 2 nam qua thl tnt la I;
5, Tubi chuho kh6ng d8n: 2 => bi8n nghimat

Nh6m 3 CHA(';l6NCK26 21 Nh6m 3 CHAf:.eNCK26

. vi DV MINH HQA . . vi DV MINH HQA .. .

PHAN TICH B!¥T SO HAl NHOM :BANG SPSS PHAN tiCH BI:¢T SO HAl NHOM :QANG SPSS

9~ta bi'n ph1;1

ttu.~o WI pi~
d~ l•pvao m6
~nh

QIO\tnslon Rt:Uda

I 11 ~·
!JCftlt~ - lllf
12 ~
ll 1l
14
1.'\.l:l:lltRocpoou
1li 1
"' lrl u ._ . .. ,..Uihta &a It-

Nh6m 3 Cti~K26 23 Nh6m 3 CHACifiNCK26 24

 11/03/2017

vi DU MINH HOA vi DU MINH HOA

PHAN Ttcu BitT s6 sAl NaoM aANG spss PBAN Ttcu BI¥T s6 HAl NaoM BANa sPss
X

~Mf·nd
lllra biBn dQc l~p
\'1\o m~c dinh * ®: riii~:~~ii~~:~
Duqc SD trong • !, neJQ;:Ia ned va rl<!nCI
phan tich nay D !!ahalanollls dlslarict
0 Smiillist Frat o
O E <ro); 'l
t{othod se hi~n
wo> chQn ph\lO'ng
phapva tieu
OhiJ.Iai
obttAfl dua biBn
Fllmm~~ nf~l'! !'ill

Nh6m 3 CHAC';eNCK26 25 2$

. vi DU MINH HOA . . vi DU MINH HOA . .

PIIAN TiCH BI$T so RAJ NHOM BANG SPSS PIIAN TiCH BI.eT SO HAl NJioM BANG SPSS
IClick vao Sta~~f~
id
cn 1ooli!ro l{l n~nl.. 'l*lr--
~.~~~~.~.:.!1.:(~-~:~:
[] Wf.l ...... 1uuau t:n~ ... lar 1:tt

n. bi8n chia m§u n lft'!nlli-QIOU,ii)O.... Rilr~t

vlo ~ Selection
U ,..., nM rl ~ml

1 ... muu ph&n tich

:,
I

0• mlu kiSm tra

j;

Nh6m 3 CHACCIN,CK26 27

7
 11/03/2017

vi DU MINH HOA . vi DU MINII HOA

PHANTiCH B!¥T HAl NHOM BANG SPSS so PHAN TiCH BIJ~T SOHAI.NHOM BANG SPSS
IClick vito Classi~
Xi\c suAt f hiin bi~t cac
dUQ'c tinh quan sat bAng rna X
bAng quy t.t&n hi~p phuong
~i n(ii bQcac
tho giua cac
nh6m .ah6mTB hay
"hom rieng bi~t
na hiQn kq
el!l tiSt ctia
'tM'ngquan
v~~udll phan
t(uulb6mTB:
J
.sat ohllnJ oho cac

B!\ngkq
nMm. ri8ng cho
tli'uJ nh6m va ve
II h
(l{
'---.J ,........,_,......,...., ..........,..J ..._ _,
phan bi~t
t61ll t!t
bi4\$d6 vi tri IClick vao Conti~~~ .-> OK
Nh6m 3 CHAC;eNCK26 29 Nh6m 3 CHA(:eNCK26

. vi DU MINH HOA . vi DU MINH HOA

PIIAN ttCH BI$T soHAl NHOM BANG SPSS PIIAN TICH B!¥T so
HAl NBOM BANG SPSS
•UOC LUQNG A XAC DJNR MU'C lr NCliiA THONG KE
lTrung blnh va d.~ J~n chuAn cua hai nh61~: IMa tr~n tuang q~~~:~ng n()i bQ cac nh6~: ::
fiti8 1Jl ~tm ii;( l k:M.

'JC\ II C r~ (11!:\VI CO) \"1/ lt.s'

ll<~r.ihl r.ft Vl93 1l sh:j, De\llillo n l.ll tWW iyt I.I:!U W•l! i llll.l I 'lnlllf1A r til rtf? f. lg
~0 tn ; l ~·ouf., II Q1 b.! ~ U ~ t'.Jrt'J~;! lb l ~JIIU
~;, · \JU •Jiti~ <.II II o.e :oJ lk.: J!)2 15 l~. :JOQ
IM I dO d't ttU II= ~
gan bo wot gla d lnll
!l?r.
.824
J'J77
~.990. ,
1 )1ft

28
14~

.oz·
~,
r. rrt h1 rJ:No• n~ ·~~443 15 t ~::JOO
til~ ~h~P. ~u1?~n1 1!i 1~ :lfln quyml ~ la dinh' .657 ~ 1 28 .00"
ruc1:~u -.o ~ED '8.11JS2 15 1.tJOO lliU I I h ~ JI A53 ~I .,
, 2$ .~oc
~org IM " Of~·ll.ll., U.!~J ~- n Y U 1b fb.::IIJII !UOI c•u hO .!~*" 1 :110 ;>(I ')'f1
<;;a"l'c•sjth,dl·h ~ . 0::67 :t~17t 16 t~:lllO __,_.-
'·"'' n1r QlltJ\ n~ , QM.,1~ t~:»D
·~
lll•hNP
b.l c l :l'l~ 111it
~
;;o: ::JJ
1 :lll~O,jo'fll(o

O', J7101
1!1
15
111::nn
t$,:)110
C6 kha nang phiin b\;t m(lt each c6
y ngllla khac bi~ ~b tbhitng nguoi
Total tri< I 2Q4f.#C1 4,ij:ti/ fv!II~ U ~ ~u :!1/.::ll<lll dii d8n nghi milt v. n!Wtng nguO'i
!;<I' I J II Y ~ ttht,IJ· II 1 .0 =33 1..,.ae~ 30 3(1.:10~
khong.
,_,., "'r ai<la\onr 3.~~e7 l. tPI9 30 3CI. JPP
ttl• htt.p 20.8.0!:07 !!1L"n2 30 3(1.JOD-
lli CI ' hU 't4t 51.9:: :JJ O.S"Tl9S :JO J p,.J11p
l'lnom ~
31 Nh6m 3C~K2 6

,•,
8
__.~ ..... :· :!
 11/03/2017

~ . .

vi DU MINH HOA VI DU MINH HOA

PIIAN TiCH BI:f:T HAl NHOM BANG SPSS so PIIAN TiCB BIE:T HAl NHOM BANG SPSS so
f1Jh6m va t6ng cac do~ blnh ph1I<Yng n{li b(;$
• SSG/SSW
n•
1Y s6 gifra t&ng cac d(i t4c:h binh plnrong giiia • Chitim 100%
phU'an; .;ai giai
thfch dl.fV(; nguyen
• GIAI THiCll UT QUA

nhAh. · Sti!lnd;~r.._.. Canonical

DM:r ~lriM Fwu;tion
(~llhi
Jlt~ tuang quan = Fun:1lo n
Tb quan trQOg ~Ua cac
~~-L~~~~~~~~~._~~~~~a~t 1 bi8n duqc t:he hi~n qua
BW. ph=g = 0.64
•l oo::> Cho thiy 64% Uri:l ..w ;;r ~u Jt;ff .096 dQ 16n tuy~t d6i CUa h~
~g sai biBn ph\1 "'~n to ¥01gliJi4lf h ?:1:1 s6 chufrn hoa cua ~
lbt,Wc (nghl m;it) dugc ·
~ thl~h OOi rna hinh
CIU~ MlO !JI3 Cl!m .459 phan bi~t => cac hi~ c6
~'
tl~u nh.ap
tuo c l1 ~ tto
.i'~3

.2D9
h~ s6 chuAQ h6a cang
16n thl d6ng g6p cang
nbi8u
Tich sd d~i
hrgng lamda
dan biSn cua Nh6m 3 CHAC5NCK26 34'
tirng ham

.
vi DU MINH HOA .. . . vi DV MINH HQA . . . .
PHAN TtcH BI\tT HAI NiioM BANG sPss so PHAN TiCB BIET SO HAl
·: NHOM BANG SPSS . '

Bi~t s6 trungbmh nb6m

(Group Centroids) .em
FunctiOns at Grou~l nh6m nhfrng ngu<:ri d!
Fumtion
1
a~ s6 tuong q~ k8t C11ntr-*l
:
d~n nghi t;;ti resort la
~Au dang x8p theo thu n,r - ~Ft~nction duong, trong khi cua
hu nhap .an giam dAn
quv mo ~ a clnM .541 1
nhfrng ngum chua <:tan b\
=> Thu nh~p Ia bi8n reso11vi sit am.
gan bo vc 1 gla cunn .346
hal jo ch du llch .213 quan trong n4dt di phan (;0 1.291 ~> Cac bi~n dt>c l~p
bi~t cac nh6m, s~m d6 kl1ong -1 .291 dang tac dong thl$1
tJol :hu to .164
.ai quy mo ~ dinh .... I ) ,1 '0:
chi8u (tY 1e tbuin) ddi
v6i bifut phv tb~

Nh6m 3 CHAC$~CK26 35 Nh6m 3 Cj,jAaJNCK26

9
 11/03/2017

vi DV MINH HQA vi DU MINH HOA

PIIAN TiCH BI:f;T SO HAl NHOM BANG SPSS PIIAN TiCB BIJ];T HAl NHOM BANG SPSS so
DANH GI..\ Kit QUA o DANH GL\ J(tT QUA
o

CIZIHIIICIIFI«! ....W.s•"
--
P' l<l ~·d GIC LII~ Mt'llb~No

.,
t U!liJ~{M 'O ~110~_
,,
T.ota.

. ~ vito bang
Ol""n111'1 nl
:::.uc s. toetect
- I
.SI'.Nift-fi<JI•-'•~~~~
- SuM onrln:al Co~ nl 12 3
.,,
.......... J
.,..." Cutewise n
,,..,...
~ 111 ' 11 1"1 l "l
I
...
-·.
Oln r• t·, ~ co 80,0 :o.J 100.0
S~tics ta thAy,
.I
~ ~.- Acl\tt iOa.:" UI-U)
I ,... ,
J'llrtli>-'C¥ C:tfttf..tl
2 1.. ,
. .aD.. ,

. "~
ln0"9 .0 1CD. ) lOO.O
l I
I 2
I
I J
... U .l tt .UIJ
. It')
m.lu phful tich co ;o••s l\ot$ele~ te.d r:>r g ln 1 ~01r.. nr C:>
WlO"G
L

0
.1.
o-
•
D
''""
I I I )

f Ull quari sat bi pban < I

I
I
I
I
I I
I
l
UJ

'" IM1 '" ¢

571
b* sai hi 1, 7, 14
"" 1\,0"9
OG.T
.D
li. J
1CO.J
100.0
10011
,
I I I_ l Dlt
, ..
2~.1~ :OH7
~: ~ u . u-. Ol' QIOG11·0·; IJIAOI IJ f O..ip i CI CUOC CC II.=I\'C~Olllllll ea

.,.
t
t
I
I
I
f
I
I
I
l
l
fU

.....
011

"'
U/1
,.,.
•.liS
I

' ""
))
Vi e<J gia tri <O =>
pbai bi~t vao nh6m
D. 83. 3,_ or Uf"UI• - or 91ncu ' rQ_ paoa caus :; 11•:1 y e1asw 111'<1

ry I~ p~ bi~t dUng hl (12+ 15 )/~0 90% => l,lon s6 nay hoi pMng-,C,i
u ·~•

D
t f ) ; 1J.f )'IJ
2 (nhfrng nguai
..
14Utf
I'
1l
I
f
I
I ' l
Otl
.,:;.
!!I"
:~I) Thl,I'C hifn pp&n tich bi~t s6 tren • kiam tra d<)c l~p t)l.i tY I~ 1)1\)' Ia
cJD d'en resort)
II I
! I
l
...
0'1
,
~ 01 1Ul

.,
'"
83,33~'0 .

.."
II I I I
Do hai nh6m c6 quy mo b~ng nbQU ali:n ngdu nhien tY I~ dung Ia 50%
·- '""
I f j lO'
•t~ • f
'"" => Phlll' ff(~ nay giup cai thi~n «tfqc leba nang pMn bi~ iil'mg 18n
khoang 3:J. 33% > 25% => m6 hlnh phin bi~t kha t6t
Nh6m 3 CMAoeNCK26 37 Nh6m 3 Ct1.A,dei'1CK26

10