You are on page 1of 17

Descriptive Statistics &

Korelasi
Dr. Muhammad Ikhsan Sulaiman
Penyajian data
Rata-Rata

Standard Deviasi (S)

Z-Skor
equation isalso
the knownmean ofasa the sample.sample size. You pronounce n as
“sample size” to avoid confusion with the symbol N that
ted oard The •WORKED-OUT
A subscripted uppercase italic X (forthe example, ofX1a) sam-
that
interbl
esac kb represents (and
representstimes.
ple of get-ready
PROBLEMS
one ofYou
is pronounced
theneed
calculate
data three
valuessymbols
as) themeanpopulation
being summed.
to write the Because
size.
in ?
math(optional) Usingthe
equation problem
for
these contains
calculating
symbols ten the
the mean:
creates datafollowing
values, there are ten X val-
equation:

Rata-Rata (Mean)
ues, the first one labeled X1, the last one labeled X10.
• An uppercase
+ X 2 + X italic + XX +with +a horizontal
+ X 7 +line above+it, X ,
+
X 1
X •= pronounced
A lowercaseasitalic 3 4 X
n, which
“X bar,” 5 X 6
represents the
that represents
X 8
thenumberX 9 that
number Xof10 ptg
isdata
the values
mean of that were summed
a sample. n in this sample, a concept
By•usingAalso known
an asuppercase
ellipsis the
(…), sample
you size.
can Youexample,
X abbreviatepronounce as
) nthat
theX1equation as
n terested subscripted
“sample size” to avoid
italic
confusion
(for
with the symbol N that
i represents one of the data values being summed. Because
in ? X +
represents X + ! + Xis10pronounced
X = the 1
problem(and
2
contains ten data values, as) the population
there are ten Xsize.val-
math n
ues, the first one labeled X1, the last one labeled X10.
Using these symbols creates the following equation:
Using
• A the lowercaseinsightitalic
thatn,the whichvalue of the last
represents thesubscript
number ofwill always
X + values + Xthat
3+ X + n, + Xcan6 +inX 7+X 8 + Xthe + concept
Rata-rata adalah penjumlahan dari seluruh data
be equal
X= 1
data
also known…
X
X 2the
to
+ X2 + + Xn
value 4of
were X 5you
summed generalize
this
as the samplen size. You pronounce n as
sample, 9a Xformula
10 as ptg

X=
By
1
“sample size” to avoid confusion with the symbol N that
using an ellipsis (…), you can abbreviate the equation as
dibagi dengan jumlah data tersebut
By
X =
representsn (and is pronounced as) the population size.
X1 + X
using the 2 +uppercase
! + X 10 Greek letter sigma, 8, a standard sym-
Using these symbols creates the following equation:
bol that is used n in mathematics to represent the summing of
X1 +you
values,
Using
X = the 2+
Xinsight
can + X 4 + Xsimplify
X 3further
that 5 + X of
the value 6 + the
X 7 +last
the + X 9as
X 8subscript
formula + X 10
will always ptg
be equal n
¨ Xto the value of n, you can generalize the formula as
= an ellipsis
ByXusing
X 1n + X 2 + … + (…), X
you can abbreviate the equation as
X= n
X1 + X
or=more
X 2+n ! + Xas
explicitly 10
n
By using n
the uppercase Greek letter sigma, 8, a standard sym-
Using theX isinsight that the value oftothe last subscript will always
bol that ∑ i
used in mathematics represent the summing of
be equal to the value of n, you can generalize the formula as
values,
X = i =1you can further simplify the formula as
+n +… + Xn
XX==X¨1 X X 2
n n
in which i represents a placeholder for 8,
a subscript and the i =
Problem: Nilai yang ekstrim besar atau kecil
By
or
bol
using the
moren explicitly
1 and
that isbelow
subscripts
n
uppercase
used inand
used.
Greek letter sigma, a
as above the sigma represent the range of the
mathematics
standard
to represent the summing of
sym-

values, you can further simplify the formula as


∑ Xi
akan mengganggu keakuratan dalam menduga
X
X==¨ i =1X
nn

nilai
in which i rata-rata
represents a placeholdersuatu
or more explicitly as
n
for a subscript populasi
and the i = From the Library of Gayle M. Noll
1 and nX below and above the sigma represent the range of the
∑ i
subscripts
i =1 used.
X=
n

in which i represents a placeholder for a subscript and the i =


1 and n below and above the sigma represent the range of the
Rata-Rata (Mean)
Hitung nilai rata-rata dari temperatur udara
selama periode waktu 10 hari sbb:

Hari 1 2 3 4 5 6 7 8 9 10

Temp (C) 27 32 26 28 26 30 29 31 33 29
Standar Deviasi
(Keragaman)
• Standar Deviasi mengukur variasi data terhadap
nilai rata-rata

• Semakin besar STD maka semakin besar sebaran


data

• Pada penelitian laboratorium, STD yang besar


tidak diinginkan karena menunjukkan ketidak
akuratan pengukuran atau ketelitian yang
rendah
Standar Deviasi
1. Standar deviasi dari sample

2. Standar deviasi dari populasi data


Standar Deviasi
Table 1

Hari
1
Temp (℃)
27
Temp - Rata
-2.1
(Temp - Rata)^2
4.4
Kesimpulan:
2
3
32
26
2.9
-3.1
8.4
9.6
Temperatur berkisar
4
5
28
26
-1.1
-3.1
1.2
9.6
pada nilai
6
7
30
29
0.9
-0.1
0.8
0.0
29.1 ± 2.4 ℃
8 31 1.9 3.6
9 33 3.9 15.2
10 29 -0.1 0.0
Rata-Rata 29.1 Jumlah 52.9
Z-Skor
Z-Skor berfungsi untuk melihat data yang
memiliki nilai ekstrim yang nilai tersebut dapat
membuat nilai rata-rata kurang akurat

Z-skor > 2 atau Z-skor < -2 menunjukkan bahwa


data merupakan nilai ekstrim
Z-skor
Hari Temp Temp - Rata Z-skor
1 27 -3.0 -0.8
2 32 2.0 0.5
3 26 -4.0 -1.1
4 28 -2.0 -0.5
5 26 -4.0 -1.1
6 30 0.0 0.0
7 38 8.0 2.2
8 31 1.0 0.3
9 33 3.0 0.8
10 29 -1.0 -0.3
Rata-Rata 30
STD 3.7
Regresi & Korelasi
Regresi Sederhana
Regresi menunjukkan hubungan matematika antara 2
variabel yaitu antara variabel bebas x (independent variable)
dengan variabel terikat y (dependent variable)

Korelasi menjelaskan keeratan hubungan antara variabel


bebas dan variabel terikat

Dalam pembahasan dan interpretasi data perlu dicari


hubungan antara 2 variabel atau lebih dan bagaimana kedua
variabel tersebut saling berinteraksi dan mempengaruhi

Contoh misalnya hubungan antara nilai kalori makanan


dengan tinggi badan
INTERPRETATION Simple linear regression attempts to discover whether
the values of the dependent Y (such as store sales) and the independent X
variable (such as the size of the store), when graphed on a scatter plot (see
Section 2.2), would suggest a straight-line relationship of the values. The fol-
Regresi antara 2 variabel
lowing figure shows the different types of patterns that you could discover
when plotting the values of the X and Y variables.

A B C

Y Y Y

X X X
Linier positif Linier negatif Logaritmik

D E F

Y Y Y

X X X

Polinomial Eksponensial no relationship


Model matematis

Linier y = a + bx

Logaritmik y = a + b ln x

Polinomial y = a + bx + cx2 + dx3 + …

Eksponensial y = a ebx
Korelasi Antara 2
Variabel
2
Koefisien determinasi (R )
2
0<R <1
2
Semakin mendekati 1 nilai R maka semakin erat hubungan antara variabel y dan x
2
R = 0.89 berarti bahwa sekitar 89% dari titik-titik pada kurva dapat diterangkan
oleh persamaan regresi tersebut

Koefisien korelasi (r)

-1 < r < 1

Korelasi antara y dan x bersifat positif jika r mendekati nilai 1 atau bersifat negatif
jika r mendekati nilai -1

Tidak terdapat korelasi antara y dan x jika nilai r mendekati nol

Koefisien determinasi dan korelasi digunakan untuk memilih persamaan atau hubungan
antara variabel y dan x apakah hubungannya linear, polinomial, eksponensial, dst.
10.2 DETERMINING THE SIMPLE LINEAR REGRESSION EQUATION 211
Regresi Linier
WORKED-OUT PROBLEM 1 You want to assist a moving company owner
to develop a more accurate method of predicting the labor hours needed for a
moving job by using the volume of goods (in cubic feet) that is being moved.
The manager has collected the following data for 36 moves and has elimi-
nated the travel-time portion of the time needed for the move.
Contoh:
Hours Feet Hours Feet
hubungan antara waktu dengan
24.00 545 25.00 557
Moving
volume barang yang dapat
13.50 400 45.00 1,028
dipindahkan oleh pekerja dalam
26.25 562 29.00 793
25.00 540 21.00 523
feet cubic
9.00 220 22.00 564
20.00 344 16.50 312

22.00 569 37.00 757 y = a + bx


11.25 340 32.00 600
50.00 900 34.00 796
ptg
12.00 285 25.00 577

38.75 865 31.00 500


40.00 831 24.00 695
19.50 344 40.00 1,054

18.00 360 27.00 486


28.00 750 18.00 442

27.00 650 62.50 1,249


21.00 415 53.75 995
15.00 275 79.50 1,397
Hasil Regresi & Korelasi

1800
y = 20.815x
1600 R² = 0.85608
y = -0.1007x2 + 25.749x - 13.919
1400
Volume Barang (Ft^3)

R² = 0.90299
1200
1000 y = 542.56ln(x) - 1139.2
R² = 0.85948
800
600
400
200
0
- 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00 90.00
Waktu (jam)
Terima kasih

You might also like