econometrics project factors affecting fmt students sleep duration

HANOI UNIVERSITY

FACULTY OF MANAGEMENT AND TOURISM

ECONOMETRICS PROJECT

FACTORS AFFECTING FMT STUDENTS’ SLEEP DURATION

Tutor: Ms Le Thanh Binh

Group 7 — Tut 1

Ngo Vu Thu Ngan - 2104050001

Khuong Hanh Nguyen - 2104050036

Vu Hong Nhung - 2004050041 Nguyen Thị Ha Trang - 2004040109

Hanoi, 5" May 2023

Copyright © 2023 by Thu Ngan, Hanh Nguyen, Hong Nhung, Ha Trang All rights reserved

No part of this project may be reproduced, stored or transmitted in any form or by any

means, photocopying, scanning or otherwise without written permission from the authors

It is legal to copy or distribute it without the approval.

TV Data analysis and resuÌt - 1 221211211121 121151 11211211111 1110111112 1111111 0111 TH1 H1 HH 4

2 Model analysis and resuÌfs - + 112021 1121111121 1121151121 1101211 01110111101 11 11 H1 1112111 7

2.1 Testing for signiicance (sIgmificantf †€SẨ) ác 0.1 1 1111 111111111 0111111 H1 re, 7 a) T-test of 1ndrvidual partial eoefÏTc1enIs -¡ c HH n1 1012111112121 He trà 7 b) Testing the overall signiRcance of all eoefficients (F-test) cài 8 2.2 Test for the functional form of the mod€lL, St 211 11321111222 11121 101 tre ng 9 2.3 Dummy vartable reeresslon mod€ÌÏ: - c1 1211111131111 111111111111 111110111111 He, 9

2.4 Error checking test ốc 10 a) MulticolÏIi€ATIEV ác n1 11 01111111211111 111111111 t1 HH k HH1 He, 10

b) Heteroscedasticity P8 - 12 Pa\11091409631-1010(0/GHaadđaaaadđdaiảảii 16

VỊ Summary and ConeÌusion - c1 2212222211211 11 11211511211 121 1111211181111 1 11 8111111111111 111kg 17 REFERENCES L1 0n n1 n1 1111111111151 H1 HH tk tk kh 1E kk H1 x1 v 9111165 19 1aà0I9)8 01 .4 G 20

Table 1: Table 2: Table 3: Table 4: Table 5: Table 6: Table 7: Table 8:

ACKNOWLEDGEMENT

Sleep is one of the most basic human functions, it plays an important role in helping the body rest, restore and regenerate energy for the activities of organs in the body, especially the brain However, many people pay very little attention to it Especially for students, the decline in sleep quality is becoming more and more common, which not only causes damage to physical and mental health but also greatly affects students’ learning and research results Therefore, we conducted this research paper to examine the sleep duration and factors affecting the sleep quality of students The research sample was collected from 133 FMT students of Hanoi University To get the best view on the status of students' sleep duration, and make

recommendations to help students improve their sleep, we study based on income, screen time,

GPA and sleep environmental factors.

I- Introduction

Sleep is a vital component of our daily routine and has a profound impact on our physical and mental well-being It is crucial for students, especially those in higher education, to maintain a healthy sleep schedule as it can impact their academic performance, overall health, and productivity Yet university students often face challenges in getting adequate rest The increasing prevalence of technology and social media has led to many students spending long hours on screens, which can interfere with their sleep patterns Additionally, the demands of academic work and extracurricular activities can also contribute to irregular sleep schedules Studies showed that university students tend to have erratic sleep patterns and do not get the recommended 7-9 hours of sleep per night Specifically, Lund et al (2010) found that university students had a mean sleep duration of 6.8 hours per night and that over 60% were categorized as poor-quality sleepers Moreover, Gaultney (2010) found that university students had a high tendency of suffering from sleeping disorders

To better understand the factors influencing the sleep patterns of students in the Faculty of Management and Tourism (FMT), a survey was conducted to investigate 4 factors: screen time, GPA, income, and environment With the increasing importance of achieving academic success, it is crucial to examine the role of these factors in promoting or hindering students' sleep health

This information can be used by FMT students, educators, and health practitioners to create

strategies and interventions to improve the sleeping habits of FMT students

II- Literature Review

Several studies have investigated the relationship between screen time and sleep duration among university students, primarily in developed countries For example, a study by Hale and Guan (2015) found that increased technology use, including screen time, was associated with shorter sleep duration among university students in the United States Another study by Exelmans and Van den Bulck (2017) found similar results among Belgian university students However, there is limited research on this topic in developing countries, particularly in Southeast Asia Given the growing use of technology among university students in Vietnam, it is important to investigate the impact of screen time on sleep duration in this population Therefore, the present study aims to fill this gap by examining the relationship between screen time and sleep duration among

FMT students at Hanoi University

One of the factors that have been proposed to affect sleep duration among university students is income Few studies have specifically examined the relationship between income and sleep duration among university students However, some studies have explored the relationship between financial stress and sleep duration, which may provide insight into the potential role of

income in sleep duration For example, it was found that financial stress was negatively associated with sleep duration among university students in the United States (Du et al, 2021), Similarly, in a study in 2010, financial stress was found to be a significant predictor of poor sleep quality among university students in Norway (Lund et al, 2010) These findings suggest that financial strain, which may be related to income, may negatively affect sleep duration among university students Other studies have also found that factors such as academic workload, stress, and lifestyle factors can influence sleep duration among university students For example, it was found that academic workload and stress were negatively associated with sleep duration among university students in Hong Kong (Suen et al, 2017) In summary, the relationship between income and sleep duration among university students is not well-established, and few studies have specifically examined this relationship However, studies on financial stress and other factors such as academic workload and lifestyle factors suggest that income may be one of several factors that influence sleep duration among university students

According to research by scientists at Harvard Medical School (Division of Sleep Medicine, Harvard Medical School, 2008), the bedroom environment can have a significant effect on sleep quality, including light, noise, and temperature Light affects the human circadian clock through the "light-sensitive" cells in the retina of the eye As for noise, although background sounds can be relaxing, the volume should be low Too much noise can make it difficult for people to fall asleep, or cause a state of startled awakening many times, then unable to fall back to sleep again Research shows that the ideal temperature range for sleep varies from person to person, but being too hot or too cold can also cause sleep disruption Thus, it can be said that factors belonging to the sleeping environment such as light, noise, and temperature have a specific impact on the students’ sleep duration

There have been several studies on the relationship between sleep duration and academic performance While the effect of sleeplessness on how well students perform academically has been proven many times, the influence of academic achievement on students’ sleep duration has not been a frequently mentioned topic in research We analyse academic performance as an independent variable in the current study, because students subjected to high standards may sacrifice sleep for a chance of improving their GPA In fact, a few papers have touched on this topic, and found that sleep duration does not strongly affect one’s academic performance However, the population of the research does not really correlate with ours, for they only sampled medical students in clinical years (Alsaggaf et al., 2016) We want to check this assumption with the population of FMT students in Hanoi University Students’ academic

performance will be measured by their GPA in the previous semester, i.e semester I, course

2022-2023

III- Methodology

1 Survey

This research sample is derived from the population of students studying at the Faculty of Management and Tourism, Hanoi University in the second semester of the academic year 2022- 2023 The study sample was 133 students who participated in a survey about their sleep quality in the last six months

2 Instrument

The survey is conducted in the form of a questionnaire, including 11 questions to indicate the student's referral information and factors affecting sleep (gender, academic year, income, usage habits, internet usage, environment, semester subjects, GPA) Collected data are presented through Eviews and Excel software

3 Analytic Framework

This study is conducted based on a multivariable linear regression model, to determine what factors have an impact on the average sleep time of FMT faculty students in the second semester

WoW

of 2022-2023 This project examines the correlation between the factors "income", "screen time",

"GPA", "environment" and sleep duration

The research model is presented as follows:

Sleeping duration = B, + B.*Screen time + B;*Income + Bu*GPA + Bs*Sleep environment + u

Therein:

- Sleep duration: The average sleep time of students in the last six months - Screen time: Student's internet usage time in the last six months - Income: Student's average income in the last six months - GPA: GPA of the 2nd semester of the academic year 2022-2023 - Environment: Factors that most influence students’ slee 4 Procedures of testing

O = Individual partial coefficient test (t-test): Test whether the importance of each coefficient has an impact on the dependent variables or not by T-test

O Testing the stability of the estimated regression model: Test the significance or impact of

all the coefficients on our dependent variable using the F test, in which, HO: B2 =0, B3 =

0, B4 =0, B5 =0 (all variables are unaffected) and H1: B2 40, B3 4 0, B4 £0, B5 40 (at least one variable has an effect)

Testing functional form of regression model: Select the most suitable model R2 (coefficient of determination) and C.V (coefficient of variance) are applied when testing with lin-lin, log-log, and semi-log R2- The highest “Goodness of Fit” where an equation can be constructed is preferred

Testing the stability of the estimated regression model: Use Chow test and Dummy to

exercise the structural change Error — checking test:

O Multicollinearity: Inaccuracy in functional interactions that exist between independent variables

O Heteroscedasticity: Since heteroscedasticity causes errors and inaccuracies in the

process of developing a regression model, there are approaches to determine whether or not this phenomena exists

Level of significance: We chose the level of significance of 5% (0.05) since scientists discovered that an alpha level of 5% is a reasonable choice for balancing those two forms of error

IV- Data analysis and result 1 Descriptive statistic

First, we have to import the dataset into Eviews, and then we put factors affecting Sleep Duration to get the multiple regression function:

Sleeping duration=p: + B2*Screen time + B3*Income + B4*Sleep environment + B;*GPA + u From the Eviews, we will have results in the table below:

Duration 1.104839166

-1.619594098

Screen time

Income -872489.9655 Sleep

environment GPA

Correlation

Sleeping

Duration Sleeping

Duration 1 Screen time Income Sleep environment

GPA

10.20687 0.006076 760.5 900.1767 133

Screen time

-1.619594098 -872489.9655 6.768245802

6.039414 0.048816 2.17E+08 2.02E+14 133

Income

1401643.677

22.17087 9.629287 0.000015 0.00811 72 407.13 33.02256 9.740131

133 133

Sleep

environment GPA -0.3133868506 -0.07823590932

0.5285770818 1401643.677 1519955633445 235626.0953

Screen time

-0.5922684433 -0.6732795981 -0.5983450666 0.4370026848 0.4077469704

03835556689 -0.5922684433 1

-0.6732795981 0.4370026848 1

-0.3133868506 0.528577/0818 235626.0953 -0.07823590932 0.1179495732 56652.82379

Income

0.2482898977 0.0096 15014981

Sleep environment

-0.5983450666 0.4077469704 0.3835556689 1

-0.2750426409 0.1675335294 0.1698046314 0.07130399546

0.1179495732 56652.82379 0.009615014981 0.07323406637

GPA -0.2750426409 0.1675335294 0.1698046314 0.07130399546 1

The scatter graphs illustrate the relationship between Sleeping duration and other variables are provided below:

Sleep duration Sleep duraion

As can be seen, there is a quite strong relationship between Sleeping Duration and Screen time,

Income, Sleep environment, and GPA

Dependent Variable: SLEEP_DURATION

Method: Least Squares

Date: 05/05/23 Time: 00:31 Sample: 1 133 Included observations: 133

Variable Coefficient Std Error t-Statistic Prob

c 9.356179 0.614948 15.21458 0.0000

SCREEN_TIME -0.103304 0.024231 -4.263226 0.0000 INCOME -3.52E-07 5.06E-08 -6.962267 0.0000

ENVIRONMENT -0.687210 0.122535 -5.608254 0.0000 GPA -0.539295 0203542 -2.649554 0.0091

R-squared 0.662623 Mean dependent var 6.169173 Adjusted R-squared 0.652080 S.D dependentvar 1.055087 S.E of regression 0.622340 Akaike info criterion 1.926210 Sum squared resid 4957535 S criterion 2.034869 Log likelihood -123.0929 Hannan-Quinn criter 1.970365 F-statistic 62.84947 Durbin-Watson stat 1.982128

Prob(F-statistic) 0.000000

2 Model analysis and results

2.1 Testing for significance (significant test) We have the function:

SLEEP DURATION = 935617928315 - 0.103303589501*SCREEN TIME - 3.52127416677e-07*INCOME - 0.53929480864*GPA - 0.687209614225*ENVIRONMENT

a, T-test of individual partial coefficients

First of all, we will conduct the T test to check if Income, Screen time and GPA and

Environment affect Sleep Duration

O Hypothesis

Ho: B2 = B3 = B4 = B5 = 0 (all variables are unaffected) H.: B2 #0, B3 40, B4 40, BS £0 (the variables has an effect) Test statistic: t* = Bi1-OSe(Bi) (= 2, 3, 4, 5)

¡ñ Level of significance: o = 5% = 0.05

T° «20.025, n-k = to.025, 128 = 1.979

O Decision rule: Reject Ho if t* > te The result table according to Eviews is:

Conclusion: There is enough evidence to conclude that Income, Screen time and GPA and Method: Least Squares

Date: 05/04/23 Time: 14:22

Sample: 1133 Included observations: 133

Variable Coefficient Std Error t-Statistic Prob c 9.356179 0.614948 15.21458 0.0000 SCREEN_TIME -0.103304 0.024231 -4.263226 0.0000 INCOME -3.52E-07 §.06E-08 -6.962267 0.0000

-ñ.538285 203542 -2.649554 0.0091 ENVIRONMENT -0.687210 0.122535 -5.608254 0.0000 R-squared 0.662623 Mean dependentvar 6.169173

Adjusted R-squared 0.652080 $.0 dependent var 1.055087 S.E, of regression 0.622340 Akaike info criterion 1.926210 Sum squared resid 49.57535 Schwarz criterion 2.034869 Log likelihood -123.0929 Hannan-Quinn criter 1.970365 F-statistic 62.84947 Durbin-VWatson stat 1.982128

Prob(F- statistic) 0.000000

Environment have an influence on Sleep Duration

b Testing the overall significance of all coefficients (F-test)

2.2 Test for the functional form of the model:

The functional forms, which combine the lin-lin, and log-lin models should be taken into

account to identify the best model Whichever contains the highest R-square is considered the most suitable one

Model lin-lin log-lin

However, given that the difference is only around 1%, it is still reasonable to use the lin-lin

model because of its other advantages, such as simpler interpretation, and it also better fits the underlying theory

2.3 Dummy variable regression model: - Step 1: Introduce the dummy

In the equation for Sleep Duration, the possibility of FMT students affected by the sleep environment will get a shorter sleep duration holding other factors equal In this case, the specified model is:

Sleeping duration = Bi + B2*Screen time + B3*Income + B4* D, + Bs*GPA + u

Where D, = 1 if the student is affected by noises during sleep, = 0 if otherwise (not affected by

noises)

- Step 2: Estimate the equation and find the difference in the relationship between 2 groups: The following regressions are obtained

- Not affected by the sleeping environment (noises):

Sleeping duration-hat = B: + B2*Screen time + B3*Income + By*GPA

= 935617928315 - 0.103303589501*Screen time - 3.52127416677e-07* Income - 0.53929480864*GPA