tiểu luận kinh tế lượng THE DETERMINANTS IMPACTING ON THE WEIGHT OF k57 VJCC STUDENTS

FOREIGN TRADE UNIVERSITY
FACULTY OF INTERNATIONAL ECONOMICS
…………..o0o…………..

ECONOMETRICS FINAL EXAM
TOPIC: THE DETERMINANTS IMPACTING ON THE
WEIGHT OF K57 VJCC STUDENTS

Class

: 57 JIB

Lecturer

: Dr. Tu Thuy Anh
Dr. Chu Thi Mai Phuong

Group

: Đặng Thu Trang - 1815520227
Chu Bình Minh - 1815520203
Ngô Phương Linh - 1810520190
Ha Noi - 2019

Contents


Abstract………………………………………………………....…………………......1
Introduction…………………………………………………………………………...3
1

Literature review……………………………………………..……………….3

2

Theoretical background………………………………………………………6
2.1 Calorie…………………………………………………………………….6
2.2 Exercising………………………………………………………………....6
2.3 Gender…………………………………………………………………….6
2.4 Sleeping hours…………………………………………………………….6
2.5 Meal frequency…………………………………………………………....7

3

Descriptive statistic of data…………………………………………………...8

4

Econometric model…………………………………………………………..10

5

Robust check…………………………………………………………………12
5.1 Multicollinearity………………………………………………………....12
5.2 Normality………………………………………………………………...13
5.3 Heteroscedasticity………………………………………………………..15
5.4 F test of the overall significance of model……………………………….14
5.5 t test of the hypothesis relating to a regression coefficient…………........16
5.6 Adjusted regression model……………………………………………….20

6

Finding and Discussion………………………………………………………21
6.1 Finding and supplement………………………………………………….21
6.1.1 Follow mentalhelp.net………………………………………………..21
6.1.2 Follow medbroadcast.com…………………………………………...22
6.2 Discussion………………………………………………………………..22

7

Conclusion……………………………………………………………………24

8

Reference……………………………………………………………………..26

9

Appendix……………………………………………………………………...28

Abstract

This analysis was performed to assess the effect of exercising, sleeping hours, gender,
1


meal frequency and calories on body weight of students at university in general and of
VJCC institute in particular. We collected information about these factors from K57
VJCC students and use the Excel and Gretl software to calculate and analyze them.
Our results reveals that: the weight will increase by 0,0115 kilogram whenever a
VJCC student has 1 calorie more in his or her meal; with each meal more, a K57

student will increase their weight by 2,08948 kilogram; and being a K57 boy will have
4.28871 kilograms more than from a K57 girl. From the healthy perspective, it
suggests that K57 students should take into account the impacts of calories taken in
everyday and their number of meals per day, besides, have suitable sleeping time each
day and hours of exercising per week to get the appropriate number of weight.

Introduction

The most important issues that people always pay attention to is their health. In
particular, weight is one of the main factors showing human health. Rarely does
anyone care about their weight, choose a salad over a cake or run for hours on a

2


treadmill instead of lying on a warm bed. Most research has shown a consistent
association between weight gain and health problems. Moreover, overweight and
obesity may increase the risk of many health problems, including diabetes, heart
disease, and certain cancers. For health problems, we always want a very effective
drug that can solve the problem right away. Unfortunately, no immediate dose can help
you lose weight, as weight is influenced by many factors. Who are they? How do they
affect weight? These questions have been answered in many studies. In this article, we
will rely on the knowledge gained to conduct a more in-depth analysis of the weight
and factors affecting weight data collected from 40 K57 VJCC students to get a
specific look on the impact of those factors on their weight.

1

Literature review

In this article, we have selected only a few common factors that affect weight to do a
search for related studies. The factors are: exercising, sleeping hours, gender, meal
frequency and calories.

3


Regarding the exercising factor, studies searched have shown that exercising has a
direct effect on weight. Ballor DL and Kessey RE (1991) had performed a metaanalysis to assess the effects of type, duration and frequency of exercise training on
changes in body mass (BM), fat mass (FM), fat-free mass (FFM) and percent body fat
(percent fat) both for adult males and females. They found that in females, weeks of
training and duration of exercise per session were significant predictors that account
for most of the variance associated with changes in BM, FM and percent fat associated
with aerobic-type exercise training, beside energy expended during exercise and initial
body fat levels (or body mass) in both males and females. These findings confirm
earlier research in males concerning exercise training effects on body mass and body
composition and extend them both to females and to a broader range of exercise types.
Both weight training exercise and aerobic exercise facilitate body fat loss, and also
preserve or increase fat-free mass. In other research, J. S. Garrow and C. D.
Summerbell (1995) had performed a meta – analysis to determine if physical training
converses fat – free mass (FFM) in overweight men or women during weight loss. The
relation between loss of weight and loss of FFM was examined by linear regression
analysis among exercising and non – exercising groups of men or women. Through 28
publications reported results on 226 sedentary men in 13 groups, 233 exercising men
in 14 groups, 199 sedentary women in 23 groups, and 258 exercising women in 28
groups, they concluded that aerobic exercise causes a modest loss in weight without
dieting and exercise provides some conservation of FFM during weight loss by
dieting, probably in part by maintaining glycogen and water.
The second factor that plays an important role in maintaining weight is sleeping
hours. Francesco P. Cappuccio, MD, FRCP, Frances M. Taggart, PhD, NgiangaBakwin Kandala, PhD, Andrew Currie, MB ChB, Ed Peile, FRCP,Saverio Stranges,

MD, PhD, Michelle A. Miller, PhD (2008) had performed a meta – analysis to assess
whether the evidence supports the presence of a relationship between short sleep
duration and obesity at different ages, and to obtain an estimate of the risk. Through
many searches of publications, they concluded that cross-sectional studies from
around the world show a consistent increased risk of obesity among short sleepers in
children and adults and causal inference is difficult due to lack of control for important

4


confounders and inconsistent evidence of temporal sequence in prospective studies.
Thus, sleep enough everyday can help people maintain their weight.
In addition, one factor that affect weight is meal frequency. Brad Jon Schoenfeld,
Alan Albert Aragon, James W. Krieger (2015) had performed a meta – analysis based
on a hypothesis that eating small, frequent meals enhances fat loss and helps to
achieve better weight maintenance. A number of observational studies lend support to
this hypothesis, with an inverse relationship noted between the frequency of eating and
adiposity. Through 15 studies that was identified, they realized that feeding frequency
was positively associated with reductions in fat mass and body fat percentage as well
as an increase in fat – free mass. And although the initial results of this meta – analysis
suggest a potential benefit of increased feeding frequencies for enhancing body
composition, these findings need to be interpreted with circumspection.
Plus, the harmony between meal frequency and energy balance is also very
important. France Bellisle, Regina McDevitt and Andrew M. Prentice (1997)
concluded that there was robust evidence from several independent laboratories to
refute the hypothesis that feeding frequency is a significant determinant of energetic
efficiency in human subject when assessed over 24 hours or longer. They argued that
feeding frequency has no significant impact on the rate of weight loss during energy
restriction and any putative effects on regulation of body weight must be mediated
through effects on the intake side of the energy balance equation.

All of these studies had shown the important of exercising, sleeping hours, gender,
meal frequency and calories in weight maintaining, but none of them had aggregated
statistics about the effect of all these factors on weight of university student. Our
report below will perform this task: The effect of exercising, sleeping hours, gender,
meal frequency and calories on weight of students of JIB K57 VJCC.

2.

Theoretical background

2.1. Calorie
A calorie is a measure of energy expenditure and stored energy. In general thought,
taking less than the amount of calories burned will help to lose weight, taking more

5


than the amount of calories burned will help to gain weight and taking the same
amount of calories will help to maintaining weight. It also depends on many other
factors such as: gender, age… Your body uses the calories you get from food to fuel
your basal metabolic rate (BMR), digestion, and physical activity.
2.2. Exercising
More physical activity increases the number of calories your body uses for energy or
“burns off.” The burning of calories through physical activity, combined with reducing
the number of calories you eat, creates a “calorie deficit” that results in weight loss.
Exercising plays an important role in losing, maintaining, even raising weight.
Therefore, university students who do not have the exercise habits in comparison with
the ones who practice physical exercise everyday might have a higher number of
weight or body fat.
2.3. Gender

The findings revealed that a greater proportion of men were overweight and obese
compared to women. Both genders show equally unhealthy dietary habits, marked by
low mean HES scores. Gender-specific dietary patterns were noted, with more men
reporting eating >6 oz (168 g) of animal protein per day as compared to women. A
higher proportion of female students reported regular consumption of potato chips and
fatty salty snacks and that they ate sweets more than twice a day.
2.4. Sleeping hours
Not getting enough sleep on a regular basis won’t just make you tired—it will also
likely cause your weight to rise. In fact, people who sleep for less than five hours a
night are almost a third likelier to gain weight than those who get seven hours of shuteye a night.
When your body doesn't respond properly to insulin, your body has trouble processing
fats from your bloodstream, so it ends up storing them as fat. So it’s not so much that
if you sleep, you’ll lose weight, but that too little sleep hampers your metabolism and
contributes to weight gain.

2.5. Meal frequency

6


Preliminary research shows that high eating frequency (8x/day) results in more hunger
and desire to eat, and less fullness than low eating frequency (3x/day). On the other
hand, eating only once or twice per day is supposed not only not able to lose weight
but also affect to the health due to the shortage of nutrition. University students
usually tend to have the number of meal below the recommended one result from
many causes like the losing weight in the wrong way, or just simply that they are too
lazy to eat enough meals.

3.

Descriptive statistic of data

From the correlation matrix above, we can make some thoughts about if it has the
correlation relationship among variables or not, but we could not give out the

7


conclusion.

Description:
● Y weight Mean: The average weight of 40 surveyed VJCC K57 students is
54,788 kilograms
● Y weight Minimum: The minimum weight among 40 surveyed VJCC K57
students is 44 kilograms
● Y weight Median: fitted value of dependent variable Y weight 53.5 kilograms
● Y weight Maximum: The maximum weight among 41 surveyed VJCC K57
students is 75 kilograms
● Std. Dev. (Standard Deviation): is a measure of how spread the numbers are,
equals to the square root of sample variance. The Std. Dev. of Y weight here is
7,7789
● C.V. (Coefficient of Variation): is simply the standard deviation divided by the
sample mean. Large values of the C.V. indicate that the mean is not very
precisely measured. The C.V. of Y weight here is 0,14198

8


➢ From the summary statistic chart, we can see that it might be the
representative sample for Weight depends on the five variables

4.

Econometric model

9


● Setup dependent and independent variables:
Dependent variable:
Y : The weight of VJCC K57 students
Independent variables:
: The amount of calories a VJCC K57 student takes in a day
: The number of hours per week a VJCC K57 student spends on doing exercises
: The number of hours per day a VJCC K57 student spends on sleeping
: The number of meals a VJCC K57 student has per day
: The gender of a VJCC K57 student (1 if male, 0 if female)
● The estimate OLS regression
= 23,3895 + 0,0115 - 0,213 + 0,025 + 2,119 + 4,247
= 0,879901 : It means that the five regressors explain 84,6% of the variance of Weight
SER = 2.887228 : It estimates standard deviation of error . A relative low spread of
scatter plot means that prediction of Weight base on these variables might be reliable.

10


5.

Robust check

5.1. Multicollinearity

11


We can see from the chart that all the VIF values of the variables are less than 10.
➢ All the variables do not show the collinearity problem.

5.2 Normality

12


From the Histogram, it seems to be quite normal distribution with the 2 relatively
equal tails.

13


From Kurtosis and Skewness data above, we see that it might not be normal
distribution as when looking with the naked eyes.

However, by the Jarque-Bera test, we can identify whether the assumption of normal
distribution is right or not.
: ui follows normal distribution
: ui does not follow normal distribution

14

It can be seen from the figures that p-value = 0,56734 >> = 0,05
● With the common = 5% for the 2-tail test, we are able to give the conclusion
not to reject the assumption .
● Therefore, we accept the assumption that follows the normal distribution.

5.3. Heteroscedasticity:
White test :

Ho : var ( ) = σ2 for all i
The data table above shows that p-value = 0,362666 > = 0,05
➢ For the common = 5% for the 2-tail test, we are able to give the conclusion
not to reject hypothesis Ho : var ( ) = σ2 for all i.

15


5..4 F test of the overall significance of model
Hypothesis : : = 0 : not suitable model
: > 0 : suitable model

We can check F test by looking at the P-value ( F) from the chart above
P-value (F) = 1.10e-14 << alpha = 0.05
➢ Conclusion : At 5% level of significance, we have suitable model.

5.5. T-test of the hypothesis relating to a regression coefficient
2-tail testing :

:=*
:≠*

16


Our data has :
● The number of observations : n = 40
● The number of variables : k = 5
➢ Degree of freedom = n - k = 35
➢ Level of significance : / 2 = 0.025
Searching in the Significance level table of t-student distribution, we have:

tc( 2.00 ; 2.04 )
Verification of statistical significance of parameters :
Testing 1 : :=0 – no linear relationship between and Y
: ≠ 0 – a linear relationship is recorded between and Y
● From the chart above, t-ratio of = 5.354 > 2.04
p-value = 5.97e-06 < 0.05 =
● Moreover ,*** means that the statistical significance of equals to 1%
➢ At 1% level of significance, we have enough evidence to reject , that a
linear relationship is recorded between and Y.
Testing 2 : : =0 – no linear relationship between and Y

17


: ≠ 0 – a linear relationship is recorded between and Y
● From the chart above, t-ratio of =-0.06104 ( -2.00;2.00 )
p-value = 0.3565 > 0.05 =
● Moreover, no * means that does not have the statistical significance
➢ At 5% level of significance, we have enough evidence not to reject , that
no linear relationship is recorded between and Y.

Testing 3 ::=0 – no linear relationship between and Y
: ≠ 0 – a linear relationship is recorded between and Y
● From the chart above, t-ratio of = -0,06536 ( -2,00;2,00 )
p-value = 0.9517 > 0.05 =
● Moreover ,no * means that does not have the statistical significance
➢ At 5% level of significance, we have enough evidence not to reject , that
no linear relationship is recorded between and Y.
Testing 4 : : =0 – no linear relationship between and Y
: ≠ 0 – a linear relationship is recorded between and Y
● From the chart above, t-ratio of = 3,451 > 2,4
p-value = 0.0015 < 0.05 =
● However ,*** means that the statistical significance of equals to 1%
➢ At 1% level of significance, we have enough evidence to reject , that a
linear relationship is recorded between and Y.
Testing 5 : :=0 – no linear relationship between and Y
: ≠ 0 – a linear relationship is recorded between and Y
● From the chart above, t-ratio of = 3.439 > 2.04
p-value = 0.0016 < 0.05 =
➢ At 1% level of significance, we have enough evidence to reject , that a
linear relationship is recorded between and Y.
➔ According to the results above, the variables , , have the statistical significance
in the model. Therefore, Weight of VJCC K57 students is affected by the
following factors : the number of Calories taken per day, the number of Meal

18


per day and Gender.

5.6. Adjusted regression model

Our research uses data on gender, meal number, amount of calories, time to sleep, time
to do exercise, and weight from 40 VJCC K57 students. We reject (hours to do
exercise per day) and (the number of hours of sleep per day), keep (amount of
calories taken in per day), (number of meals per day) and (gender), we have the
adjusted OLS regression:

The estimated OLS regression is:
= 22,7204 + 0,0115 C + 2,08948 M+ 4,28871 G
For the purpose of the research, let:
C = Amount of calories one takes in 1 day
M = Number of meals per day one has
G = Gender ( 0 if female, 1 if male)
➢ It can be shown from the chart that:
● The OLS estimate of the intercept is 22.7204

19


● The OLS estimate of the coefficient on the amount of calories one takes
in a day is 0,0115. The number let us know the weight will increase by
0,0115 kilogram when a VJCC K57 student has 1 calory more in his or
her meal.
● The OLS estimate of the coefficient on the number of meals per day one
has is 2.08948. Through this, we can see with each meal more, a VJCC
K57 student will increase their weight by 2.08948 kilograms.
● The OLS estimate of the coefficient on the gender is 4.28871. This is the
biggest factor that influences the weight of VJCC K57 student. Being a
VJCC K57 boy will capture 4.28871 kilograms more than from a VJCC
K57 girl.
●

= 0.876686 : It means that the three regressors explain 87.66% of the
variance of Weight

● SER = 2.843184 : It estimates standard deviation of error . A relative
low spread of scatter plot means that prediction of Weight base on these
variables might be reliable.

6.

Finding and Discussion

6.1. Finding and supplement
6.1.1. Follow Harvard Heath Publishing : People's weights are determined by
multiple factors including personal factors, community factors, societal influences,
social norms,..
● Personal factors: Your age, gender, race or ethnicity, as well as how you react
emotionally to other people and stress affect your diet & exercise

20


● Community factors: Refer to the various physical location you inhabit - home,
school, and workplace - as well as where you receive medical care, exercise,
and obtain food
● Societal influences: Include government, public health, agriculture, marketing,
community design and a range of industries food, beverage, physical activity,
and entertainment.
● Social norms: Such as culture definition of an “ideal” body, or family pressure
to clean your plate, overarching layer of influence.

6.1.2. Follow medbroadcast.com
There are some ways to improve the weight without too much long-term effort :
● Medical intervention: Medications may be part of a weight management
program.Medications aren't "magic cures" leading to permanent weight loss.
They're generally used in combination with a proper diet and exercise program.
They are only for people who are classified as obese (i.e., those with a BMI
over 30), or people with a BMI of 27 and extra heart disease risk factors such as
high cholesterol or diabetes. Some medications are approved for short-term use
only. One example of a weight-loss medication available in Canada is orlistat*,
which blocks the absorption of fat from the bowel.
● Surgery is only considered when other weight management options have not
been successful. There are many forms of obesity surgery, but often surgery
reduces the size of the stomach so that only a small amount of food can be
eaten comfortably. Some of the terms used to describe the surgeries used to
treat obesity include: gastric surgery, gastric bypass surgery, laparoscopic band
surgery, Roux-en-Y gastric bypass, stomach "stapling"

6.2. Discussions
For sleeping, research suggests that there’s a link between how much time people
sleep and how much they weigh. In general, children and adults who get too little
sleep tend to weigh more than those who get enough sleep.

21


For example, in the Nurses’ Health Study, researchers followed roughly 60,000
women for 16 years. At the start of the study, all of the women were healthy, and none
were obese; 16 years later, women who slept 5 hours or less per night had a 15 percent
higher risk of becoming obese, compared to women who slept 7 hours per night. Short

sleepers also had 30 percent higher risk of gaining 30 pounds over the course of the
study, compared to women who got 7 hours of sleep per night.
There are several possible ways that sleep deprivation could increase the chances of
becoming obese. Sleep-deprived people may be too tired to exercise, decreasing the
“calories burned” side of the weight-change equation. People who don’t get enough
sleep may take in more calories than those who do, simply because they are awake
longer and have more opportunities to eat. Lack of sleep also disrupts the balance of
key hormones that control appetite, so sleep-deprived people may be hungrier than
those who get enough rest each night.
Regular physical activity is an important part of weight management. In addition to
managing weight, exercise also improves overall health and can help reduce the
risk of diseases such as certain cancers, heart disease, and osteoporosis. Regular
physical activity doesn't mean you have to join the nearest gym. It can be as simple as
climbing the stairs instead of taking the elevator, walking or cycling to work and
leaving the car at home (if at all possible), or going for a walk at lunchtime with
coworkers. What's important is to add exercise to your daily routine, and to work
towards a higher activity level. Choose activities and exercises you enjoy.
Although how much time one spends on doing exercises per week or how many hours
one sleeps per day seems not to have an effect on VJCC K57 students’ weight, it could
be that these factors became their habits and their body has formed self-adaptive
mechanism so their weight might not be affected by these factors. But in fact having
a suitable period of time for them have an important effect on not only weight
but also health followed by some above researches.

22


7.

Conclusion

After doing survey about effects on weight of K57 VJCC and using Gretl to solve data
given, we have drawn to some conclusions below:
● Gender has an effect on a VJCC K57 student’s weight. As we know, each
gender has different body composition. Therefore, overall, the weight of
male is different (higher) from female’s.
● The weight of VJCC K57 students is also affected by the number of meals
per day.The different number of meals one has per day influences
differently to his or her weight. A limited research suggests that eating three
or more structured meals per day compared with fewer than three meals per
day can help control appetite and lead to feelings of fullness. (According to
www.health.harvard.edu).

23


● Last but not least, changing the amount of calories taken in is definitely the
effective way of influencing weight of VJCC K57. Because a person on
average needs about 2000 calories per day to maintain energy for normal
activities. If one would like to gain or lose weight, he or she can increase or
decrease the amount of calories taken in (suitable for each body’s condition
and status).
Recommendations:
● It is recommended that VJCC K57 Students should take those below factors
into consideration: Amount of calories they consume per day and the number of
meals they have per day. On the other hand, although other factors like physical
exercise or sleeping hours per day do not have the statistic significance in that
data but K57 students of VJCC should sleep enough at least 6-7 hours per day
and take to physical activities to have a fit body as well as healthy lifestyle.
● For the overweight ones, harsh diet do not have the same meaning with

your weight loss, instead, try to take in the as low as possible number of
calories to lose kilograms. They could think about an enough-nutrition diet
with not too low energy, as long as calo-in less than calo-out, do some
physical exercise to increase metabolic ability, drink enough water and
sleep enough everyday.
● For the too thin ones, they should consider to add healthy calories to gain
weight but not to drastically change the diet. You can consider increasing
calories by adding nut or seed toppings, cheese, and healthy side dishes.
Try almonds, sunflower seeds, fruit, or whole-grain, wheat toast. Include
high calorie fruits in your diet like ripe mangoes and bananas to gain
weight. Avoid the unhealthy high-calorie food which only increase the
number of calories without good digestion, leading to an unhealthy body
and obesity.

24