You are on page 1of 6

Statistics and Probability

Governor Pack Road, Baguio City, Philippines 2600


Tel. Nos.: (+6374) 442-3316, 442-8220; 444-2786;
442-2564; 442-8219; 442-8256; Fax No.: 442-6268 Grade Level/Section: Grade 11
Email: email@uc-bcf.edu.ph; Website: www.uc-bcf.edu.ph

MODULE 10 – Stat Subject Teacher: JUWAN ROSS B. TAGUDAR

REGRESSION ANALYSIS
- A standard procedure used to describe the nature of the relationship between two or more
variables

 Linear Regression Analysis


Regression is a descriptive statistical technique for finding the best fitting straight line between
two variables. It is the line drawn through the points on a scatter plot to summarize the relationship
between the variables being studied.

Calculation of a Regression Line


In statistics, we can calculate a regression line for two variables if their correlation is very
strong, and their scatter plot shows linear pattern. A regression line is a single that best fits the data (in
terms of having smallest overall distance from the line to the points). This technique for finding the
best fitting line is called simple linear regression analysis using the least square method.

Least square regression equation: y = a + bx


where:
y = dependent variable a = slope
x = independent variable b = y intercept, when x = 0

Regression equation (regression coefficients)


𝑛 (∑ 𝑥𝑦) − (∑ 𝑥)(∑ 𝑦)
𝑏=
𝑛(∑ 𝑥 2 ) − (∑ 𝑥)2

∑ 𝑦 − 𝑏 (∑ 𝑥)
𝑎=
𝑛
Example 1.

Number of Hours Studied (X) Examination Grade (Y)


8 56 The following data were obtained from a group of
11 79
students regarding the number of hours that they
devoted for studying and the grades that they
10 70 obtained in their examination. Compute for the
5 54 regression analysis.
18 94
n=7
15 85
2 33

∑ 𝑥 = 8 + 11 + 10 + 5 + 18 + 15 + 2 = 69
∑ 𝑦 = 56 + 79 + 70 + 54 + 94 + 85 + 33 = 471
∑ 𝑥𝑦 = 8(56) + 11(79) + 10(70) + 5(54) + 18(94) + 15(85) + 2(33) = 5,320
∑ 𝑥 2 = 82 + 112 + 102 + 52 + 182 + 152 + 22 = 863

Substituting the values in the regression formula to obtain the values for a and b, we have:

7 (5,320)−(69)(471)
𝑏= 7(863)−(69)2
= 3.7039 ≈ 𝟑. 𝟕𝟎

471−3.7(69)
𝑎= 7
= 30.8143 ≈ 𝟑𝟎. 𝟖𝟏

Page 1 of 6
Statistics and Probability
Governor Pack Road, Baguio City, Philippines 2600
Tel. Nos.: (+6374) 442-3316, 442-8220; 444-2786;
442-2564; 442-8219; 442-8256; Fax No.: 442-6268 Grade Level/Section: Grade 11
Email: email@uc-bcf.edu.ph; Website: www.uc-bcf.edu.ph

MODULE 10 – Stat Subject Teacher: JUWAN ROSS B. TAGUDAR

Regression analysis yielded a regression equation of y = 30.81 + 3.70x. The regression coefficient
means that for every single hour increase in studying there is a corresponding 30.81 unit of increase in
their examination grade.

The annual consumer expenditures and annual net incomes of a sample of 10 families in a
Metropolitan Area in a particular year are shown in the following table below. Prepare a regression
and a correlation analysis of their expenditures and net incomes for the 10 families for the year.

Family Net Income(X) (in hundred thousand) Expenditure (Y) (in thousand)
A 10 23
B 2 7
C 4 15
D 6 17
E 8 23
F 7 22
G 4 10
H 6 14
I 7 20
J 6 19
Solutions:
Family x y xy x2
A 10 23 230 100
B 2 7 14 4
C 4 15 60 16
D 6 17 102 36
E 8 23 184 64
F 7 22 154 49
G 4 10 40 16
H 6 14 84 36
I 7 20 140 49
J 6 19 114 36
n = 10 ∑x = 60 ∑y = 170 ∑xy = 1,122 ∑x2 = 406

Substituting the values in the regression formula to obtain the values for a and b, we have:
𝑛 (∑ 𝑥𝑦) − (∑ 𝑥)(∑ 𝑦) 10 (1,122) − (60)(170)
𝑏= 2 2
= = 2.2174 ≈ 𝟐. 𝟐𝟐
𝑛(∑ 𝑥 ) − (∑ 𝑥) 10(406) − (60)2

∑ 𝑦 − 𝑏 (∑ 𝑥) 170 − 2.22(60)
𝑎= = = 𝟑. 𝟔𝟖
𝑛 10

Regression analysis yielded a regression equation of y = 3.68 + 2.22x. The regression coefficient means
that for every single unit increase in the net income there is a corresponding 3.68 unit of increase in
expenditure.
For example, let us say the net income is 500,000, using the formula in getting the expenditures
Y = 3.68 + 2.22(5) = 14.78 (we use 5 instead of 500,000 because of the case in the given problem).
Page 2 of 6
Statistics and Probability
Governor Pack Road, Baguio City, Philippines 2600
Tel. Nos.: (+6374) 442-3316, 442-8220; 444-2786;
442-2564; 442-8219; 442-8256; Fax No.: 442-6268 Grade Level/Section: Grade 11
Email: email@uc-bcf.edu.ph; Website: www.uc-bcf.edu.ph

MODULE 10 – Stat Subject Teacher: JUWAN ROSS B. TAGUDAR

Reference:
Downie, N. M. (n.d.). Basic Statistical Methods (5th ed.). Mandaluyong Metro Manila: Harper & Row.
Wahab, D. M. M. A. (n.d.). Correlation & Regression. University of Alexandria.
Zaid, D. M. A. (2015). Correlation and Regression Analysis TEXTBOOK . The Statistical, Economic and
Social Research and Training Centre for Islamic Countries (SESRIC) , 1–39.

Page 3 of 6
Statistics and Probability
Governor Pack Road, Baguio City, Philippines 2600
Tel. Nos.: (+6374) 442-3316, 442-8220; 444-2786;
442-2564; 442-8219; 442-8256; Fax No.: 442-6268 Grade Level/Section: Grade 11
Email: email@uc-bcf.edu.ph; Website: www.uc-bcf.edu.ph

MODULE 10 – Stat Subject Teacher: JUWAN ROSS B. TAGUDAR

Performance Task

INSTRUCTION: Short or long bond paper (scrap bond paper will


do as long as the other side of the paper is clean for your
answers). Follow the format shown on the right. HAND WRITTEN,
please make sure your hand writing is readable.

Introduction

In the study of Tingganay (2003), students stay incompetent in mathematics for they lack
competency on the fundamental concepts comprised in the study of this discipline. In the 2012
Program for International Student Assessment (PISA), United States recorded low scores in
mathematics and ranked 36th out of the 65 participants worldwide. PISA is a system of international
assessments that measures 15-year-olds’ capabilities in basic mathematics literacy, including science
and reading (Willms, 2003). According to Wilson (2000), Mathematics is not exactly known as
academe’s most progressive discipline when it comes to curricular reform. Particularly, in statistics
and probability, students are still taught to plug in numbers and chug through a formula, and some
undergraduates never learn how this course relates to other disciplines, much less the real world. In
addition, Husch (2001), approximately 40% of the students enrolled in the first semester statistics and
probability class at The University of Tennessee in fall semester, 1999, made grades of D, F, or W
(Withdrawn). As cited by Husch (2001), repeatedly, professors have reported similar situations with
high failure rates. This entails that academic performance of students in Mathematics is lower than
what is desired.

Goal

You are tasked to determine the equation that will predict students’ grades in Statistics and
Probability based on student’s grades in General Mathematics. The following are reported grades of
15 randomly selected students under the House Keeping Program enrolled in the Academic Year
2017 – 2018.

General Mathematics (x):


88 85 87 81 87 80 92 89 86 87 83 89 90 85 91
Statistics and Probability (y):
90 89 85 87 87 82 89 87 92 80 84 86 88 90 85

1. Prepare a table then compute for the coefficient of correlation. (Use Pearson Product
Correlation)

2. If a correlation exists, determine the regression equation that will predict students’ grades in
Statistics and Probability based on the students’ General Mathematics grades.

Page 4 of 6
Statistics and Probability
Governor Pack Road, Baguio City, Philippines 2600
Tel. Nos.: (+6374) 442-3316, 442-8220; 444-2786;
442-2564; 442-8219; 442-8256; Fax No.: 442-6268 Grade Level/Section: Grade 11
Email: email@uc-bcf.edu.ph; Website: www.uc-bcf.edu.ph

MODULE 10 – Stat Subject Teacher: JUWAN ROSS B. TAGUDAR

3. Summarize your results in the following table.

Variables Mean r Regression Equation


General Mathematics
Statistics and Probability

4. Interpret the results in the table under Task Number 3.


5. Using the regression equation, recall your grade in General Mathematics then substitute it on
the regression equation to predict your Statistics and Probability grade. Do you believe that
you will obtain this grade? Justify

SAMPLE OUTPUT

1. Correlation

General General
XY 𝐗𝟐 𝐘𝟐
Chemistry (X) Mathematics (Y)

𝚺𝐗 = 𝚺𝐘 = 𝚺𝐗𝐘 = 𝚺𝐗 𝟐 = 𝚺𝐘 𝟐 =

(𝚺𝐗)𝟐 = (𝚺𝐘)𝟐 = (𝚺𝐗 ⦁ 𝚺𝐘) n=

𝑛(∑ XY) − (∑ X ⦁ ∑ Y)
𝑟= = 0.80
2 2
√[𝑛(∑ X 2 ) − (∑ X) ][𝑛(∑ Y 2 ) − (∑ Y) ]
2. Regression Equation
n (∑ XY) − (∑ X)(∑ Y)
𝑏= = 21.08
n(∑ X 2 ) − (∑ X)2

∑ Y − b (∑ X)
𝑎= = 0.57
n

𝑦 = 0.57 + 21.08x
3. Summary

Variables Mean r Regression Equation


General Chemistry 89.53
0.80 y = 0.57 + 21.08x
General Mathematics 88.73

Page 5 of 6
Statistics and Probability
Governor Pack Road, Baguio City, Philippines 2600
Tel. Nos.: (+6374) 442-3316, 442-8220; 444-2786;
442-2564; 442-8219; 442-8256; Fax No.: 442-6268 Grade Level/Section: Grade 11
Email: email@uc-bcf.edu.ph; Website: www.uc-bcf.edu.ph

MODULE 10 – Stat Subject Teacher: JUWAN ROSS B. TAGUDAR

4. Interpretation

The correlation coefficient of 0.80 indicates that students’ grade in General Mathematics has
a high correlation with General Chemistry. The correlation is positive which suggests a direct
relationship between the variables. This means that students tend to have a high grade in General
Chemistry if they obtained high grades in their General Mathematics subject. This result can be
attributed to the nature of General Chemistry as a subject that involves computational and
analytical skills which are evident in the competencies of General Mathematics.

In addition, the equation that predicts students’ grades in General Chemistry is given by y =
0.57 + 21.08x. The regression equation in predicting General Chemistry grade simply means that for
every unit increase in General Mathematics grade, there is 0.57 increase in General Chemistry grade.

5. Recall your grade in General Mathematics then substitute it on the regression equation to predict
your Statistics and Probability grade. Do you believe that you will obtain this grade? Justify

Page 6 of 6

You might also like