latihan ujian
DESCRIPTION
latihanTRANSCRIPT
Nama:
NRP
Interpretasi Ujian Labkom CRP
Uji Korelasi dan Regresi
Senin, 2 Februari 2015
A. Uji Pearson
- Jenis uji : Uji korelasi Pearson
- Tujuan : Untuk melihat hubungan
- Syarat :
a. Data variabel independet dan dependent harus numerik (interval/rasio)
b. Data harus berdistribusi normal
- Variabel :
a. Variabel X=Independent= Umur ibu =rasio
b. Variabel Y=Dependet= BB Bayi= rasio
- Asumsi :
1. Hipotesis normalitas
H0 : Data berdistribusi normal
H1 : Data Tidak berdistribusi normal
2. Hipotesis penelitian
H0 : Tidak adanya hubungan antara umur ibu dengan BB Bayi
H1 : Adanya hubungan antara umur ibu dengan BB bayi
- Hasil output SPSS
1. Uji normalitas :
2.
One-Sample Kolmogorov-Smirnov Test
age of mother
birthweight
(grams)
N 189 189
Normal Parametersa,,b Mean 23.24 2944.29
Std. Deviation 5.299 729.016
Most Extreme Differences Absolute .095 .043
Positive .095 .043
Negative -.054 -.043
Kolmogorov-Smirnov Z 1.299 .598
Asymp. Sig. (2-tailed) .068 .867
a. Test distribution is Normal.
b. Calculated from data.
P Value Age mother : 0,068>0,05 : Data berdistribusi normal
P Value BB Bayi : 0,867 >0,05 : Data berdistribusi normal
Kesimpulan:
Data numerik dengan variabel Independent dan dependent berdistribusi
normal dan memenuhi syarat uji korelasi Pearson
3. Uji hipotesis :
Correlations
age of mother
birthweight
(grams)
age of mother Pearson Correlation 1 .090
Sig. (2-tailed) .221
N 189 189
birthweight (grams) Pearson Correlation .090 1
Sig. (2-tailed) .221
N 189 189
P value : 0,221 >0,05 : Tidak ada hubungan antara umur ibu dengan BB bayi
o Jika ada hubungan maka lihat pearson correlation
Arah hubungan= positif berarti searah
Kekuatan hubungan = hasil mendekati 0 maka hubungan antara
umur ibu dengan BBLR lemah.
- Kesimpulan :
Tidak ada hubungan antara umur ibu dengan BB bayi
B. Uji Spearman
- Jenis uji : Uji Korelasi Spearman
- Tujuan : Untuk melihat hubungan antara umur ibu dengan bb anak
- Syarat :
a. Data variabel independent dan dependent keduanya harus ordinal atau
salah satunya berskala ordinal
b. Data tidak harus berdistribusi normal
- Variabel :
Variabel X =Independent = Umur ibu = numerik= rasio
Variabel Y=Dependet =BB bayi = Kategorik = Ordinal
- Asumsi :
1. Hipotesis normalitas
H0 : Data berdistribusi normal
H1 : Data Tidak berdistribusi normal
2. Hipotesis penelitian
H0 : Ada hubungan antara Umur ibu dengan BB bayi
H1 : Tidak Adanya hubungan antara umur ibu dengan BB bayi
- Hasil output SPSS
1. Uji normalitas :
2.
One-Sample Kolmogorov-Smirnov Test
age of mother
birthweight
(grams)
N 189 189
Normal Parametersa,,b Mean 23.24 2944.29
Std. Deviation 5.299 729.016
Most Extreme Differences Absolute .095 .043
Positive .095 .043
Negative -.054 -.043
Kolmogorov-Smirnov Z 1.299 .598
Asymp. Sig. (2-tailed) .068 .867
a. Test distribution is Normal.
b. Calculated from data.
P value umur ibu :0,068 >0,05 : Data berdistribusi normal
P Value BB Bayi : 0,867 >0,05 : Data berdistribusi normal
3. Uji hipotesis :
Correlations
age of mother
birthweight
(grams)
Spearman's rho age of mother Correlation Coefficient 1.000 .061
Sig. (2-tailed) . .405
N 189 189
birthweight (grams) Correlation Coefficient .061 1.000
Sig. (2-tailed) .405 .
N 189 189
P Value 0,405 >0,05 : Tidak ada hubungan antara umur ibu dengan BB Bayi
C. Regresi Linier
- Jenis uji : Regresi linier sederhana
- Tujuan :
a. Untuk mengetahui pengaruh satu variabel independent (variabel X)
terhadap 1 variabel dependent (variabel Y) dan seberapa besar
pengaruhnya.
- Syarat :
o Variabel independent =rasio, distribusi normal
o Variabel dependent =rasio, distribusi normal
- Variabel :
a. Variabel Independent (X) : Rasio = BB ibu
b. Variabel Dependent (Y) : Rasio = BB bayi
- Asumsi :
1. Hipotesis normalitas
H0 : Data berdistribusi normal
H1 : Data tidak berdistribusi normal
2. Hipotesis penelitian
H0 : Tidak ada hubungan antara BB ibu dengan BB bayi
H1 : Ada hubungan antara BB ibu dengan BB bayi
- Hasil output SPSS
1. Uji normalitas :
2.
P Value BB ibu : 0,000 <0,05 : Data tidak terdistribusi normal
P Value BB Bayi : 0,867 >0,05 : Data berdistribusi normal
One-Sample Kolmogorov-Smirnov Test
birthweight
(grams)
weight at last
menstrual period
N 189 189
Normal Parametersa,,b Mean 2944.29 129.82
Std. Deviation 729.016 30.575
Most Extreme Differences Absolute .043 .152
Positive .043 .152
Negative -.043 -.075
Kolmogorov-Smirnov Z .598 2.094
Asymp. Sig. (2-tailed) .867 .000
a. Test distribution is Normal.
a. Calculated from data.
b.
Kesimpulan:
Variabel independent X ( BB ibu) tidak berdistribusi normal dan
variabel dependent Y (BB bayi) berdistribusi normal, dan untuk
memenuhi regresi linier sederhana maka variabel X yang tidak
berdistribusi normal harus di kategorikan dulu baru dapat
menggunakan uji regresi linier sederhana
3. Uji hipotesis :
Coefficientsa
Model
Unstandardized Coefficients
Standardized
Coefficients
t Sig.
95,0% Confidence Interval for
B
B Std. Error Beta Lower Bound Upper Bound
1 (Constant) 2835.979 73.744 38.457 .000 2690.502 2981.457
Kategori LWT 220.107 105.128 .151 2.094 .038 12.718 427.496
a. Dependent Variable: birthweight (grams)
P Value :0,038 < 0,05 : Ada pengaruh antara BB ibu dengan BB Anak
a. (Kuat/lemah) dan menentukan seberapa besar pengaruhnya.
Model Summary
Model R R Square
Adjusted R
Square
Std. Error of the
Estimate
1 .151a .023 .018 722.543
a. Predictors: (Constant), Kategori LWT
R : 0,151 ( lemah (Mendekati 0)
Jenis kekuatan Positif
R Square : 0,023 X 100%= 2,3% maka kekuatan regresinya lemah (0-25%)
Nilai R square menunjukan berapa persen pengaruh variabel x
terhadap Y, semakin mendekati 1 maka semakin mendekati 100%.
Karena nilai R square kita 0,023 maka nilai pengaruh variabel BB
ibu terhadap BB bayi sebesar 2,3% sementara 97,7% nya lagi
merupakan nilai variabel lain yang tidak diteliti
D. Regresi Logistik
- Jenis uji : Regresi Logistik multivariat
- Tujuan :
Untuk melihat pengaruh lebih dari satu variabel independent (X)
terhadap satu variabel dependent (Y) dan seberapa besar pengaruhnya
- Syarat :
a. Variabel Independent : (nominal,ordinal,interval,rasio)
b. Variabel dependent :Nominal
c. Tidak dipengaruhi uji normalitas
- Variabel :
a. Variabel independent (X) :
Umur ibu=rasio
LWT =ordinal
Moderate=interval
Race = nominal
b. Variabel dependent (Y) : BB bayi =nominal
- Asumsi :
1. Hipotesis penelitian
H0 : tidak ada hubungan variabel independent (X) dengan variabel dependent
(Y)
H1 : Ada hubungan variabel independent (X) dengan variabel dependent (Y)
- Hasil output SPSS
Uji hipotesis :
Variables in the Equation
B S.E. Wald df Sig. Exp(B)
95% C.I.for EXP(B)
Lower Upper
Step 1a race 7.843 2 .020
race(1) 1.231 .517 5.669 1 .017 3.425 1.243 9.438
race(2) .944 .416 5.140 1 .023 2.569 1.136 5.809
smoke(1) -1.054 .380 7.701 1 .006 .348 .165 .734
age -.023 .034 .434 1 .510 .978 .914 1.045
lwt -.013 .006 3.835 1 .050 .988 .975 1.000
Constant 1.385 1.063 1.695 1 .193 3.993
a. Variable(s) entered on step 1: race, smoke, age, lwt.
Variables in the Equation
B S.E. Wald df Sig. Exp(B)
95% C.I.for EXP(B)
Lower Upper
Step 1a Race 7.843 2 .020
race(1) 1.231 .517 5.669 1 .017 3.425 1.243 9.438
race(2) .944 .416 5.140 1 .023 2.569 1.136 5.809
smoke(1) -1.054 .380 7.701 1 .006 .348 .165 .734
Lwt -.013 .006 3.835 1 .050 .988 .975 1.000
Age -.023 .034 .434 1 .510 .978 .914 1.045
Constant 1.385 1.063 1.695 1 .193 3.993
a. Variable(s) entered on step 1: race, smoke, lwt, age.
a. Lihat sig
Race =0,020 < 0,05 =ada hubungan
Race (1) =0,017 <0,05 =ada hubungan
Race (2) = 0,023 <0,05 = Ada hubungan
Smoke (1) =0,006 <0,05 =ada hubungan
Lwt =0,050 <0,05 = ada hubungan
Age =0,510>0,05 = Tidak ada hubungan
b. Jika P<0,05 (ada hubungan x dan Y)
Lihat nilai Exp (B)/Odd ratio
Jika < 1 berisiko lebih rendah
Jika =1 berisiko sama
Jika >1 berisiko tinggi
ODDS Rasio:
Lwt =o,988
Race (1) =3,425
Race (2) =2,569
Smoke (1) = 0,348
Age = 0,978
Cara nulisnya
1. Lwt =0,988 (OR <1)
Lwt berisiko 0,988 kali lebih rendah untuk melahirkan bayi dengan BBLR
2. Race (1)=3,425 (OR>1)
Race (1) berisiko 3,425 kali lebih tinggi untuk melahirkan bayi dengan BBLR
3. Race (2) =2,569 (OR >1)
Race (2) berisiko 2,569 kali lebih tinggi untuk melahirkan bayi dengan BBLR
4. Smoke (1) = 0,348 (OR<1)
Smoke (1) berisiko 0,348 kali lebih rendah untuk melahirkan bayi dengan BBLR
5. Age =0,978 (OR <1)
Age berisiko 0,978 kali lebih rendah untuk melahirkan bayi dengan BBLR
- :