Study on the Path Relationship Between Residents' Health

Awareness, Health Behavior and Health Satisfaction

Hongmei Li

College of Science, North China University of Technology, Beijing, China

Keywords: Health Awareness, Health Behavior, Health Satisfaction, Structural Equation Model, Path Analysis.

Abstract: In the past over two years, COVID-19 has greatly changed people's health awareness and health behavior,

which further affects people's subjective perception and satisfaction with personal health. Health status has

increasingly become one of the most important indicators to reflect people's yearning for a better life. Based

on the questionnaire survey data, using Amos software to construct a structural equation model, this paper

makes an in-depth study on the impact and path relationship between residents' health awareness, health

behavior and health satisfaction, and draws some meaningful conclusions. The result shows that the direct

effect of health awareness on personal health satisfaction is relatively weak, but the indirect effect is very

significant through the intermediary effect of the variable "health behavior". At the same time, it shows that

people's health awareness has a strong direct effect on health behavior, and health behavior has a strong direct

effect on personal health satisfaction perception. At last, this paper puts forward some reasonable suggestions

for the government and relevant management and service departments in the aspects of "health science

publicity", "nutrition catering", "improving the medical and health system" and "leisure and fitness services".

1 INTRODUCTION

Since the beginning of 2020, the sudden COVID-19

has brought an unprecedented disaster to mankind,

greatly reducing people's sense of happiness and

security. However, under the difficult situation, the

epidemic has its positive side. It has greatly awakened

human awareness of public health and safety. The

most prominent performance is that people's health

awareness has generally improved. They take the

initiative to understand the development of epidemic

situation, promote scientific epidemic prevention, and

pay attention to health preservation, environmental

quality and waste classification. Their awareness of

food safety has also been significantly strengthened.

Reflected in action, people tend to care more about

their own health, increase the amount of fitness

exercise significantly, and pay more attention to

nutrition in daily diet and routine physical

examination.

In order to study these changing characteristics of

the people in a more detailed way, further study what

Li Hongmei, College of science, North China University

of Technology, No. 5, Jinyuanzhuang Road, Shijingshan

District, Beijing, 100144.

(Tel: 18610988259. E-mail: lhmei6280@163.com)

aspects of the people's health awareness and health

behavior are reflected, what factors people's health

satisfaction perception mainly come from, and what is

the action path between these variables, firstly, we

specially organized a questionnaire survey. Secondly,

using the survey data, we constructed a structural

equation model by means of Amos software, and

conducted an in-depth study on the impact and path

relationship between variables such as "health

awareness", "health behavior" and "health

satisfaction". Thirdly, based on the research results,

we have drawn some meaningful conclusions. Finally,

in the aspects of "health science publicity", "nutrition

catering", "improving the medical and health system",

and "leisure and fitness services", we have put forward

some reasonable suggestions for the government and

relevant management and service departments.

Li, H.

Study on the Path Relationship Between Residents’ Health Awareness, Health Behavior and Health Satisfaction.

DOI: 10.5220/0012041800003620

In Proceedings of the 4th International Conference on Economic Management and Model Engineering (ICEMME 2022), pages 677-682

ISBN: 978-989-758-636-1

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

677

2 VARIABLE SELECTION AND

DATA DESCRIPTION

In order to study the relationship between people's

health awareness, health behavior and health

satisfaction perception, we set up a group of attitude

scales related to health with the 5-level Likert scales

in the questionnaire. The options of the scales are

"very agree", "agree", "general", "disagree" and "very

disagree", and the values are 5, 4, 3, 2 and 1

respectively. The items and corresponding variables

are shown in Table 1.

Table 1: Correspondence between items and variables.

Ite

VariableIte

Vari ab le

1.I am very concerned about health

preservation and health knowledge.

(V1) Regimen

6. I have a regular physical

examination at least once a

year.

(V6)

Regular physical

examination

2.I am very concerned about the epidemic

development and scientific epidemic

revention.

(V2) Epidemic

prevention

7.I can achieve a reasonable

diet every day.

(V7)

Reasonable diet

3.I actively advocate garbage classification.

(V3) Garbage

classification

8.I do fitness exercises at least

three times a week.

(V8)

Regular fitness

4.I am very concerned about environmental

ualit

and adhere to the

reen conce

(V4) Quality of

environment

9.I personally have a good

sense of health.

(V9) Perceived

satisfaction

5.When buying food, I will carefully check

the production place, shelf life, ingredients

and other information.

(V5)

Food safety

10.I personally have good

health behavior.

(V10)

Behavioral

satisfaction

We distributed questionnaires in Beijing and

finally collected 561 valid questionnaires. First, the

reliability of the data is tested, Cronbach's

alpha=0.871, indicating that the data has a high level

of reliability. At the same time, KMO and Bartlett's

test is carried out on the data, and the KMO

coefficient is 0.920>0.5, sig.= 0.000, which meets the

conditions of factor analysis. And the common factor

dimensions are highly consistent with the preset

latent variables, indicating that the data has high

factor structure validity, and the scale design and data

quality meet the basic requirements.

3 SETTING AND

MODIFICATION OF

STRUCTURAL EQUATION

MODEL

In order to study the path relationship between

residents' health awareness, health behavior and

health satisfaction, we choose three latent variables,

which are "health awareness", "health behavior" and

"health satisfaction". "Health awareness" is reflected

by five observable variables, "regimen", "epidemic

prevention", "garbage classification", "quality of

environment" and "food safety". "Healthy behavior"

is reflected by three observable variables, "regular

physical examination", "reasonable diet" and "regular

fitness". "Health satisfaction" is reflected by two

observable variables, "perceived satisfaction" and

"behavioral satisfaction". In this way, the study

involves a total of 25 variables, including 10

observable variables and 15 unobservable variables

(including 3 latent variables and 12 residual terms of

endogenous variables). The hypothetical model

setting is shown in Figure 1.

Figure 1: Hypothetical model of structural equation.

By fitting the hypothetical model with the sample

data, the results show that the hypothetical model can

converge and be identified. However, the fitting is

poor, the chi square value is too large, and the p-value

is too small, so the model needs to be further

modified. According to the modification indices, the

variance of e12 is found to be -0.018, so it is limited

to 0 first. Further, it is found that there is a large

covariant correlation between the six groups of

residuals, such as e5 and e9, e2 and e4, e3 and e5, e8

ICEMME 2022 - The International Conference on Economic Management and Model Engineering

678

and e10, e4 and e5, e4 and e9, so a two-way

connection is established in sequence, and then the

model is refitted and estimated. Finally, the model

shows convergence, and the overall model fitness test

statistic 𝜒



=37.398, and the significance probability

level, p-value=0.088>0.05, so the original hypothesis

is accepted. In addition, the variance of each residual

term in the model diagram of non standardized

estimated value is positive. In the model diagram of

standardized estimation, each of the absolute value of

standardized regression coefficient and correlation

coefficient is not greater than 1. And the symbols are

consistent with the theoretical expectation, indicating

that the modified model can be adapted to the sample

data. The modified model is shown in Figure 2.

Figure 2: Modified model of structural equation.

Table 2 shows the summary table of the non

standardized regression coefficient and its

significance test. The first column is the non

standardized regression coefficients, that is, the non

standardized direct path coefficients. The second

column is the standard error of the estimated

parameters, and the third column is the test statistic

"critical ratio"(C.R.). If C.R.>1.96, it means that the

regression coefficient reaches the significance level

of 0.05. It can be seen from table 2 that only the

second group is not very significant, and the p-values

of other coefficients are all less than 0.001, indicating

that the path coefficients of the model basically meet

the significance level requirements.

Table 2: Summary of non standardized regression coefficient and its significance test.

Estimate S.E. C.R. P Label

HB <--- HA .558 .041 13.669 *** Pa

-10

HS <--- HA .079 .104 .762 .146 Pa

-8

HS <--- HB .644 .173 3.727 *** Pa

-9

V1 <--- HA 1.000 ----- ----- ----- -----

V2 <--- HA .513 .037 13.783 *** Pa

-1

V3 <--- HA .517 .042 12.265 *** Pa

-2

V4 <--- HA .477 .039 12.324 *** Pa

-3

V5 <--- HA .531 .042 12.592 *** Pa

-4

V8 <--- HB 1.000 ----- ----- ----- -----

V7 <--- HB .934 .063 14.809 *** Pa

-5

V6 <--- HB .910 .068 13.321 *** Pa

-6

V10 <--- HS 1.262 .109 11.600 *** Pa

-7

V9 <--- HS 1.000 ----- ----- ----- -----

In the process of modification, covariance

relationships between the six groups of residuals of e5

and e9, e2 and e4, e3 and e5, e8 and e10, e4 and e5,

e4 and e9 are added in sequence. Table 3 shows the

summary table of covariance and significance test

between the above variables, and all correlation

relationships are significant at the probability level of

0.05.

Study on the Path Relationship Between Residents’ Health Awareness, Health Behavior and Health Satisfaction

679

Table 3: Summary of residual covariance and its significance test.

Estimate S.E. C.R. P Label

e5 <--- e9 .088 .019 4.614 *** Par-10

e3 <--- e5 .082 .020 4.040 *** Par-11

e2 <--- e4 .073 .016 4.547 *** Par-13

e4 <--- e5 .057 .017 3.292 *** Par-14

e4 <--- e9 .060 .017 3.543 *** Par-15

e8 <--- e10 .050 .018 2.836 .005 Par-16

Further, we evaluate the fitness indicators of the

modified model. Table 4 and Table 5 show the output

results of the model fitting indicators.

Table 4: Summary of model fitting indicators (1).

Indicator Value Critical value Indicator Value Critical value

CMIN

37.398

（p=0.088）

p>0.05 RFI 0.969 >0.90

CMIN/DF 1.385 1-3 IFI 0.995 >0.90

RMR 0.015 <0.05

TLI

（NNFI）

0.991 >0.90

RMSEA 0.026 <0.05 CFI 0.995 >0.90

GFI 0.987 >0.90 PNFI 0.589 >0.50

AGFI 0.973 >0.90 PCFI 0.597 >0.50

NFI 0.982 >0.90 ----- ----- -----

Table 5: Summary of model fitting indicators (2).

Model

NCP

ECVI AIC BIC CAIC

NCP LO90 HI90

Default model 10.398 .000 30.609 .167 93.398 214.631 242.631

Saturated model .000 .000 .000 .196 110.000 348.135 403.135

Independence model 1984.329 1840.588 2135.425 3.660 2049.329 2092.626 2102.626

It can be seen from table 4 and table 5.

(1) The chi square degree of freedom ratio

(CMIN/DF) of the model is between 1-3, and the

fitting residual (RMR), progressive fitting residual

(RMSEA), fitness index (GFI), adjusted fitness index

(AGFI) and other fitting indicators meet the

requirements.

(2) The non centralized parameter (NCP) is

relatively small, and the 90% confidence interval

contains 0.

(3) NFI, RFI, IFI, TLI, CFI and other fitting

indicators are all above 0.90, and the two simplified

adjustment fitting indicators (PNFI and PCFI) all

reach 0.50, meeting the requirements.

(4) The expected cross validity Index (ECVI) of

the modified model is smaller than that of the

saturated model and the independent model.

(5) The values of the three information standards,

Akaike information criterion (AI), Bayesian

information criterion (BIC) and the adjusted Akaike

information criterion (CAIC) are smaller than those

of the saturation model and the independent model.

In conclusion, all important evaluation indicators

meet the requirements of the critical value reference

standard, and it is judged that the modified model is

well adapted, so the path analysis can be further

carried out.

ICEMME 2022 - The International Conference on Economic Management and Model Engineering

680

4 PATH ANALYSIS OF

STRUCTURAL EQUATION

MODEL

In order to compare the size of path coefficients

between different variables, we use standardized path

coefficients. Figure 3 shows the modified model

marked with standardized direct path coefficients.

Figure 3: The modified structural equation model and its path coefficients.

Table 6 shows the comparison of standardized

direct effect coefficients, indirect effect coefficients

and total effect coefficients.

Table 6: Standardized direct coefficients, indirect coefficients and total effect coefficients.

Vari ab le

Direct effect coefficient Indirect effect coefficient Total effect coefficient

HA HB HS HA HB HS HA HB HS

HB .908 .000 .000 .000 .000 .000 .908 .000 .000

HS .169 .844 .000 .766 .000 .000 .935 .844 .000

V9 .000 .000 .563 .527 .475 .000 .527 .475 .563

V10 .000 .000 .681 .637 .575 .000 .637 .575 .681

V6 .000 .628 .000 .570 .000 .000 .570 .628 .000

V7 .000 .707 .000 .642 .000 .000 .642 .707 .000

V8 .000 .732 .000 .665 .000 .000 .665 .732 .000

V5 .600 .000 .000 .000 .000 .000 .600 .000 .000

V4 .591 .000 .000 .000 .000 .000 .591 .000 .000

V3 .580 .000 .000 .000 .000 .000 .580 .000 .000

V2 .656 .000 .000 .000 .000 .000 .656 .000 .000

V1 .731 .000 .000 .000 .000 .000 .731 .000 .000

It can be seen from Figure 3 and Table 6 that

among the five observable variables of the latent

variable "health awareness"(HS), the most influential

factor is "regimen"(V1), and the standardized path

coefficient reaches 0.731, followed by "epidemic

prevention"(V2), "food safety"(V5), "quality of

environment"(V4) and "garbage classification"(V3),

whose coefficients are all above 0.5. It shows that

good health awareness are first reflected in people's

attention to regimen, and then in people's attention to

epidemic prevention and control, food safety,

environmental quality and garbage classification.

Among the three observable variables of the latent

variable "health behavior"(HB), the path coefficients

from large to small are "regular fitness"(V8),

"reasonable diet"(V7) and "regular physical

examination"(V6), indicating that the health behavior

of residents is first reflected in regular physical

exercise, then in reasonable diet matching, and thirdly

in regular physical examination every year.

Among the two observable variables of the latent

variable "health satisfaction"(HS), the path

coefficient of "behavior satisfaction"(V10) is

significantly greater than that of "perceived

Study on the Path Relationship Between Residents’ Health Awareness, Health Behavior and Health Satisfaction

681

satisfaction"(V9), indicating that residents' cognition

of health satisfaction mainly comes from the

evaluation of their own health behavior, followed by

the perception of personal health awareness.

Among the path relationships of the three latent

variables, the direct effect of "health awareness"

(HA) on "health behavior"(HB) is of decisive

significance, with a coefficient as high as 0.908.

"Health behavior"(HB) has a large direct effect on

"health satisfaction"(HS), with a coefficient of 0.844.

The direct effect of "health awareness"(HA) on

"health satisfaction"(HS) is not high, the coefficient

is only 0.169, and the statistical significance is not

significant. However, through the intermediary

variable "health behavior"(HB), the indirect path

coefficient of "health awareness"(HA) on "health

satisfaction"(HS) is 0.766, making the total effect as

high as 0.935. So the influence is very obvious, which

shows that people's perception of health satisfaction

mainly comes from the evaluation of their own health

behavior. Good health awareness must first act on

people’s own health behavior, and finally can truly

improve personal health satisfaction.

5 CONCLUSIONS AND

SUGGESTIONS

According to the path analysis of structural equation

model, we draw the following conclusions.

(1) People's attention to regimen, epidemic

prevention, garbage classification, environmental

quality and food safety can be used as an effective

observation of their health awareness.

(2) Regular fitness, reasonable diet and regular

physical examination can reflect people's healthy

behavior to a great extent.

(3) Health behavior has a strong direct effect on

personal health satisfaction perception.

(4) The direct effect of health awareness on health

satisfaction is relatively weak, but the indirect effect

is very significant through the intermediary effect of

the variable "health behavior".

(5) People's health awareness has a strong direct

effect on health behavior.

Based on the above conclusions, we propose the

following suggestions for the government and

relevant management and service departments.

(1) Strengthen health publicity, regularly hold

popular science lectures on health preservation,

epidemic prevention and environmental protection

through various channels, in order to improve

people's health awareness.

(2) Further improve the medical and health

service system, make it convenient for people to see

a doctor, guide residents to pay attention to physical

examination, and achieve "prevention before illness

and prevention of change after illness".

(3) Strengthen the supervision of nutritional

catering services, especially increase and improve the

service quality of community canteens, further

unblock the "vegetable through train" in the

community, and provide convenience for people to

have balanced nutrition and reasonable meals.

(4) Reasonably arrange sports and fitness venues,

and add convenient sports facilities and places in

communities and "pocket parks" to create a

comfortable and convenient leisure and fitness

environment for citizens.

FUND PROJECT

2020 research on the construction of first-class

undergraduate major in Statistics (Project No.:

108051360022XN522)

REFERENCES

Huang Huiqun. Research on influencing factors of online

open courses based on structural equation model[J].

Journal of Guangxi normal university of science and

technology, 2016,31 (5): 121-125.

Liu Ying, Hao Yanhua, Sun Hong. Analysis on the current

situation and influencing factors of college students'

emergency knowledge level in Heilongjiang Province

[J]. Chinese school health, 2015,36 (3): 330-332.

Wu Hao, Qin Bo. Physical condition, safety awareness and

residents' health consumption[J]. Consumer economy,

2015,31 (5): 41-7.

Zhang Xiaoyan, Xiang Chen, Wang Qianyan. Analysis of

incentive factors and countermeasures of family

doctors based on structural equation [J]. China Health

Administration, 2021,38 (6): 417-422.

Zhang Mengqian, Wang Yancui. Improvement of family

doctor system in China under the basic medical service

system[J]. Medicine and society, 2018,31 (5): 4-7.

ICEMME 2022 - The International Conference on Economic Management and Model Engineering

682