多个样本率地卡方检验及两两比较之spss超简单Word文件下载.docx
- 文档编号:17336590
- 上传时间:2022-12-01
- 格式:DOCX
- 页数:24
- 大小:84.99KB
多个样本率地卡方检验及两两比较之spss超简单Word文件下载.docx
《多个样本率地卡方检验及两两比较之spss超简单Word文件下载.docx》由会员分享,可在线阅读,更多相关《多个样本率地卡方检验及两两比较之spss超简单Word文件下载.docx(24页珍藏版)》请在冰豆网上搜索。
OWeightM&
esby
Fll'
!
二I'
-'
311^1:
-
*
CurrentStatus*Weightcafi&
sbyfr&
q
(2)点击Weightcasesby,激活FrequencyVariable窗口
吗WeigihtCases
©
0natweigtitcases.
jIBMIBJMIIMJIMMUBJII■■■■■|
(WeiglTlcasesbyj
FrequencyVa'
rii3QIe:
CurrentStains.Weigh!
casesDyfr&
R&
setCancel
Help
诳)intE“eritiondriskjevel
OK
(3)将freq变量放入FrequencyVariable栏
绘WeightCases
ODonot術eightcases.
@Weightcasesby
■RequancyVariabls-
CurrentStatus:
WeiontcasesDyITeqi
f'
1(—f
[PasteJJ[Cane巳IHelp
(4)点击OK4.2检验假设5
数据加权之后,我们要判断研究数据是否满足样本量要求,如下:
TCrosstabs
(1)在主页面点击AnalyzefDescriptiveStatistics
駅Crosstabs
□Difiplavdusieredparcharts
□Suppresstables
X
匕]Fj兰ResetICancel
(2)将变量intervention和risk_level分别放入Row(s)栏和Column(s)栏
呛Crosstabs
-Layer1of1
□Displayclusteredbarcharts
□Suppressta&
les
ROMS'
):
Exact...
■
XJ-
「isK_lev&
l
Statistics...1
CeIJs.1
1J
Column(sJ<
Format...|
intervention
鼬底“1
C-[lI-J;
lj;
rJl-J|rj_-|■_■_fITJ-I
OK]
J
'
Paste
KesetCancel
*I
(3)点击Statistics,弹出下图:
Crosstabs:
Statist!
cs
□iChi-square
■■-n[IT■-nilin-iniiii—ii—■■■■
□Correlations
rNominal
rOrdinal
nContinig,encycoeUcienit
I'
【PhiandCramer'
sV
二Lambda
CUncertaintycoefFicient
匚IGaTTiirna
LJSomers'
d
□Kendallsiau*b
□Kendall'
stau<
NominalDyInterval
□.Eta
□!
<
appa
□Risk
□McWernar
I'
ICghr■即‘5andMaitel-Haenszelst苹ti了tios
JIliii_-dI,1-1]-
Continue]Canc&
l|[H&
lp
(4)点击Chi-square
Crosstabs:
C5
■■■■■■■■■■H■■■■■H■■■■HB■H■■H
HiChi-square
•r•..■■■■—ni——p■■■i—nil■■■•・o・r'
i□Correlations
rOrdifial
□Contingencycoefficient
LJPhiandCramersV
□Lamnda
□Uncert^inbrcoefficient
?
_.Garnrna
i'
Somers'
■
rKendallslau*b
CKendall'
stau-c
NomiinalbyInterval
□Eta
□Kappa
WcNemar
LICochran'
sandMantel-Haenszelstatistics
II
ill
[Centinue]Cancel]
(5)点击ContinueCells
-Z'
test'
Counts
nComparecolumnproportions
0-C.IJ'
^I|A孙三II-H:
lI1
■11■―■111l-nill^^lllI—II■IIFT1II
閘也tJSEFVEd\
■•••—H■■■H■...—H■■X
D^peeled
□Hidesmallcounts
Lessthan匸
「P&
rcentages
rResiduals
□gciw
□Column
□lolal
□Unstandardi军匸I
□Standard'
ilzed
□Adjustedstandardized
■NonintegerWsights
@RoundcellcountsORoundcaseweightsOTruncatec&
y匚ountsOTruncaleca&
eweightsONoadjustments
Continue
Cancel
HElp
(6)点击Counts栏中的Expected选项
Crosstab!
;
CellDh,plfiiy
Counts
2-test
□Observed
IVlExpectedlj
111■■_■IlliUlBllliihlllBISIBIIIUIIllUdlllUiiUIliaJ
—-
_Hidesmallcounts
□Comparecolumnp「oportions
H'
7匚.2-.,dL!
IlCIJII“:
IC
-Percentages
一Row
□Column
Total
-F?
esidual^
□Unslandiardized
□Standardized
□Adjustedstand3rdiz&
j
-NonintegerWeights
画RoundcellcountsORoundcaseweightsOTruncatecellcountsOTruncatecaseweights©
Noadjustrrients
(£
Qntiriu&
jCancel
⑺点击ContinueOK
经上述操作,SPSS输出预期频数结果如下:
Levelofcholesterolrisk:
"
High"
&
Normal*'
*Typeofintervention:
Drug'
"
■Diet'
*5."
EKGrcisG"
Crosstabulation
EKpectedlCount
Typeofintervenlion:
Drug"
“Di或’&
ExerdsG"
Drug
Dht
Exercise
Tolal
L&
velofcticlest&
rolrisk:
High
34.7
247
24.7
74.0
■High"
£
.'
Nornnal"
Ncrnal
25.3
26.3
76.0
5010
5DIO
150.0
该表显示,本研究最小的预测频数是24.7,大于5,满足假设5,具有足够的样本量。
Chi-SquareTests表格也对该结果做出提示,如下标注部分:
Chi-SquareTests
Asymptotic
Significance
Value
df
C-sided)
PearsonChi-Square
2
.010
LikelihoodRatio
gms
□09
Linea「-t)y-LinearAssociation
7.7Ga
1
.□OS
MofValidCases
150
a.0cells(01.0%)haveexpectedcountlessthan5.Theminimumexpecledcourtis24.67.
即在本研究中,没有小于5的预测频数,可以直接进行卡方检验(2XC)。
那么,如果存在预测频数小于5的情况,我们应该怎么办呢?
一般来说,如果预测频数小于5,就需要进
行Fisher精确检验(2XC),我们将在后面推送的内容中向大家详细介绍。
4.3卡方检验(2XC)的SPSS操作
nDisplaydusteredbarcharts
□Suppressta&
Row(S'
Exact...I
uJ
Statistics...I
Cells.I
Format...
Sty!
殳”
f):
J-1
Paste11^esetCancel
-Nominal
Ordinal
nContinig,encycoelTicient
□gamma
PhiandCramer'
i.Somers^d
__Lambda
□Kendall'
slau-b
CUncertainty^coeifiaent
□Kendallstau-c
■Norr'
iinalbyInterval
O^Chrsquare
■-niIT■-nilim■■■■i—ii—■■■■■・oofi
匚Eta
-|McNemar
口Cochran'
sandMaitel-Haensselstatistics
ContinueJCanceljIH&
nMcMemar
[JCochran'
sandMantel-Haensielstatistics
Ir;
t?
-,i
匚「osHdb畐CellDitSplay
tConi^arecolumnproportions
■ig匚-三iciirnZillell匚
-Percenlages
厂1Row
□Columr
□Total
®
RourdcellcountsQRoundcaseweights
OTnjnc3l5'
c&
llcountsOTrunea1ecasewe(ghts©
Noadjustnents
[Conti「Ime]Cancel
HeIp
(5)点击Percentage栏中的Column选项
桔Crosstabs:
CellDisplay
jiB■■-■■■■mI■inBiiiiiir■・■■■rn■■■■maiiri-ijMiQbserv^j口Expected
二IHidesmallcounts
二"
■IIfl
-Percentages"
1Rowl3Column□Total
Mest
□Comparecoijrnnproportions
.T匚=t■>
[I濟也HIEIIIIIhlK
rResiduals^
IjUnslaidardized
□Adjustedstandardiz&
d
@RoundcellcounlsORoundcaseweighIs©
Truncatecellcounts©
TruncatecaseweightsO'
NoaljustTients
4.4组间比较
⑴
RowCs):
□Displaydust&
redbarcharts
lGE
菱riskjev&
ColumnCs};
gintervention
1Lsyer1of1
Pastejl'
KgsgtJ、
Cancel][Help
Statistics^...
刑乩“I
Format...|
SMe-I
⑵点击Cells,弹出下图:
HI.JL:
=3ll三II-H:
■P&
rcenlages
□Tolal
■NonintsgerWsigKitfi
RoundcellcountsORoundcaseweights
OTruncatec&
H匚ountsOTruncaleca&
eweights
ONoadjustments
Counls
BIQOserved
□Expected
□Hidesmallcounts
11匚
ORow
HColumri
□lotal
(3)点击z-test栏中的Comparecolumnproportions禾RAdjustp-values(Bonferroni
method)选项
鮭Crosstabs:
CellCispidy-z=te£
t
翩,Com£
ar&
columnpropijrtions
H^Adjusip-valjesCBonferroninnetiod|
I■■■■■■■Illl・Illi■“IIWIIIBHBIIIHIIBI^HIIIlli
-Residuals
□Unstandartod
□Adjustedstandardised
-NonintegerWeights
@RouracellcountsORoundcaseweightsOTrtincalecellcountsOTruncatecasaweightsGNoadjustiTients
5、结果解释
5.1统计描述
在进行卡方检验(2XC)的结果分析之前,我们需要先对研究数据有个基本的了解。
SPSS
输出结果如下:
risklevel*iiit^rwiTtionCrosstabiilation
irtefverrtion
Diet
riskjevelHighCount
EKpectsdCount
%withininlervention
1虽
26b
30b
74
32.036
247
56.0%
74.7
60.0%
740
45.3%
NorrmalCount
EjcpectiidCount
34a
22b
20b
76
£
3皿
253
440曲
4D.O^
50I代
TotalCount
ExpectedC(juni
鳴withinirrbswention
50
50.0
100.09t
100.0%
SO
100.C%
160
Eachsubsenptletterdenotesasubsetof[interventioncategorieswhosecolurnrproportionsdonotdiffersignificantlyliromeactiotheralthe.05level.
该表提示,本研究共有150位受试者,根据干预方式均分为3组。
在试验结束时,药物干预组的50位受试者中有16位胆固醇浓度高,饮食干预组的50位受试者中有28位胆固醇浓度高,而运动干预组的50位受试者中有30位胆固醇浓度高,如下标注部分:
rishjewl*iiiterwiitionCrosstabulation
Toial
riskjevel
H<
gh
Count
伽
23b
30bj
ExpectedCount
74J0I
%wHhinintervention
32.0%
eoiio%
49.3%
Mcirmal
如
253
25J
75.0
騙wrthininfervention
ee0%
MO%
4QL[D9t
ECU船
Ejflpect^dCount
50C
IlSO.iO
%withininlerV'
ention
100lO%
EachsubscriptIsHeirdenotesasubsetofinterventioncategorieswhosecolumnproportionsdonotdiffersignificantlyfromeachotheratthe.05level.
由此可见,药物干预比饮食或运动干预的疗效更好。
同时,该表也提示,药物干预组的50位受试者中有34位胆固醇浓度下降,饮食干预组的50位受试者中有22位胆固醇浓度下降,而运动干预组的50位受试者中只有20位胆固醇浓度下降,如下标注部分:
risklevel*iiTterventionCrosstabulation
T<
j1al
risfcjevel
Courrt
1头
2Gh
3Db
740
%wittiinintervention
56g
Normal
:
5.3
25,3
ze.o
%witlilnIntenention
69.0%
44g
40.0%
50l7%
Counrt
EO'
.O
EO.O
GOO
%wittiininteoention
1000^
100.0^
匚act!
subscriptletterdenotesasubsetofInierv&
rtioncatsgarleswhosecolumnproportionsdorotdiffersignilicantlyfromeachothsratthe.35level.
因
- 配套讲稿:
如PPT文件的首页显示word图标,表示该PPT已包含配套word讲稿。双击word图标可打开word文档。
- 特殊限制:
部分文档作品中含有的国旗、国徽等图片,仅作为作品整体效果示例展示,禁止商用。设计者仅对作品中独创性部分享有著作权。
- 关 键 词:
- 样本 检验 比较 spss 简单
![提示](https://static.bdocx.com/images/bang_tan.gif)