Tuesday, January 22, 2013

Calculating Correlation Regression by Calculator Casio fx-350MS

Here's how to calculate a single linear regression using the calculator Casio fx-350MS. Of the following is an input method on a calculator.
Just on example. 

n
1
2
3
4
5
6
7
8
9
10
11
12
x
20
22
24
26
28
30
32
34
36
38
40
42
y
8.4
9.5
11.8
10.4
13.3
14.8
13.2
14.7
16.4
16.5
18.9
18.5

Compute:
  1. Plot x and y
  2. Σxi, Σyi, Σxiyi, Σxi ², Sxx, Sxy
  3. Estimators β0 and β1
Formerly must know the buttons on the calculator that play a role in this function is as follows: [MODE], [,], [M +], [S-SUM], [S-VAR] and of course the [SHIFT] key numbers and the other buttons. 

Its first do is clean up the data on the calculator from the previous mode functions by pressing the following buttons in sequence.
[SHIFT] [CLR] [3] [=] 


Next is to make a calculator in linear regression mode by pressing the following buttons in sequence.
[MODE] [3] [1] 

The next stage is enter the data one by one. The data input is data in pairs, meaning that should be paired with y1 x1, x2 paired with y2 and so on. Steps to enter the data into the calculator is as follows.
[20] [,] [8] [.] [4] [M +] [AC]
[22] [,] [9] [.] [5] [M +] [AC] .... And so on up to n = 12 

To determine the value of xi, yi and others can be done by following steps
 [SHIFT] [1] will show the option Σx1 à ², Σx1 and others. To choose we can push the numbers being subordinated. If the coefficients are not visible, can be seen on the next screen by pressing the [REPLAY] to the right.
[SHIFT] [2] will show the option à x, y, a, b and others by pressing the [REPLAY] to the right and to the left. 


The result of the calculation is the following calculators:
  • Σxi = 372
  • Σyi = 166.4
  • Σxiyi = 5419.6
  • Σxi ² = 12 104
  • Sxy = Σxiyi-(ΣxiΣyi / n) = 5419.6-{(372) x (166.4) / 12} = 261.2
  • S ² x = Σxi ² - (Σxi ² / n) = 12 104 - {(372) ² / 12} = 572
β1 = b1 = Sxy / S ² x = 261.2/572 = 0456 -- A
β0 = b0 = ȳ-B1X = 13 867 - (0.456x31) = -0289 -- B
Regression estimators equation: y = 0.456x + -0289 

Sources: Statistical Analysis of lecture notes