KEY FACTORS AFFECTING HOUSE PURCHASE DECISION OF CUSTOMERS IN VIETNAM

Table 3.1: Main factors affecting customers’ housing purchase decision .... RESEARCH PROBLEMS & RESEARCH QUESTIONS In general, the real estate in Vietnam has got many difficulties in mak

International School of Business

ACKNOWLEGEMENTS First of all, I would like to express my deep appreciation to my supervisor, Dr Dinh Thai Hoang who instructed and helped me enthusiastically during period of the thesis

I also would like to thank you all my colleagues and friends of Hoa Binh Corporation and Sacomreal for their valuable contributions to give comments and suggestion to revise the questionnaire survey

I am grateful to the supervisory board for providing me with their available advices and patient supports when I need

I will never forget the friendly postgraduate students for helping me during studying and doing thesis

The most special thanks go to my parents, my brothers and sisters who always create the most convenient conditions for me as well as support me all time

The main purpose of the study is to investigate the effecting of key factors on housing purchase decision of customers in Vietnam First, a model which is proposed based

on analyzing of previous literature Then the model is tested on a pilot test which is conducted of a small real estate professional group and another group of 15 respondents, and on a larger survey of 263 samples The study finds out a strong positive relationship between top two factors, including “living space”, “distance” and customers’ housing purchase decision The three weaker positive relationship factors are “feature”, “finance” and “environment” to housing decision makers It is also found that there is no difference in decision making of customers according to different demographics consisting of gender, age, marital status, monthly income and education level

Key works: real estate, purchase factors, Vietnam

ACKNOWLEGEMENTS ii

ABSTRACT iii

LIST OF TABLES vii

LIST OF FIGURES viii

ABBREVATIONS viii

CHAPTER 1 INTRODUCTION 1

1.1 BACKGROUND 1

1.2 RESEARCH PROBLEMS & RESEARCH QUESTIONS 1

1.3 RESEARCH PURPOSE 3

1.4 SCOPE OF THE RESEARCH 3

1.5 RESEARCH STRUCTURES 3

CHAPTER 2 LITERATURE REVIEW 4

2.1 LITERATURE REVIEW 4

2.1.1 Feature 4

2.1.2 Living space 4

2.1.3 Finance 5

2.1.4 Distance 5

2.1.5 Environment 5

2.1.6 Purchase decision 6

2.1.7 Demography 6

2.2 CONCEPTUAL FRAMEWORK 7

CHAPTER 3 RESEARCH METHODOLOGY 8

3.1 RESEARCH PROCESS 8

3.2 SAMPLE SIZE 11

3.3 MEASUREMENT SCALE 11

3.3.1 Measurement scale 11

3.3.2 Pilot test 11

3.4 MAIN SURVEY 15

3.5 DATA ANALYSIS METHOD 15

3.5.1 Reliability measure 15

3.5.2 Validity measure by EFA (Exploratory Factor Analysis) 16

3.5.3 Multiple regression analysis 16

CHAPTER 4 DATA ANALYSIS & RESULTS 18

4.1 PREPARATION DATA 18

4.1.1 Editing 18

4.1.2 Coding 18

4.2 DESCRIPTIVE DATA 21

4.3 ASSESSMENT MEASUREMENT SCALE 23

4.3.1 Cronbach’s Alpha 23

4.3.2 Exploratory Factor Analysis (EFA) 26

4.3.2.1 Assessment of data 26

4.3.2.2 Defining number of extracted factors 27

4.4 HYPOTHESES TESTING BY MULTIPLE REGRESSION 30

4.4.1 Checking assumption of Multiple Regression 30

4.4.1.1 Sample size 30

4.4.1.2 Assessment multicollinearity of independent variables 30

4.4.1.3 Normality, linearity, homoscedasticity & outliers 30

4.4.2 Evaluating the model 31

4.4.3 Evaluating the independent of variables 31

4.4.4 Checking hypotheses of model 32

4.4.5 Analysis effect of control variables by Multiple Regression 34

CHAPTER 5 CONCLUSIONS AND IMPLICATIONS 35

5.1 RESEARCH OVERVIEW 35

5.2 RESEACH FINDINGS 35

5.3 MANAGERIAL IMPLICATIONS 36

5.4 RESEARCH LIMITATIONS & DIRECTIONS FOR FUTURE RESEARCH 37

Appendix 1: The first draft of the questionnaire 42 Appendix 2: The English questionnaire 45 Appendix 3: The Vietnamese questionnaire 49

Table 3.1: Main factors affecting customers’ housing purchase decision 13

Table 4.1: Codebook of questionnaire items 18

Table 4.2: Characteristics of respondents 22

Table 4.3: Cronbach’s Alpha test results 25

Table 4.4: EFA results 28

Table 4.5: Correlations among variables 29

Table 4.6: Coefficient table of MLR 32

Table 4.7: Hypotheses results 33

Table 4.8: Descriptive statistics 54

Table 4.9: Cronbach’s Alpha with full items for each constructs 54

Table 4.10: KMO and Bartlett’s test 56

Table 4.11: Total variance explained 56

Table 4.12: Correlation among variables (Partial only) 57

Table 4.13: Factor Matrix 59

Table 4.14: Factor Correlation Matrix 60

Table 4.15: Model summary 60

Table 4.16: Anova 60

Table 4.17: Casewise diagnostics 60

Table 4.18: Residuals statistics 61

Table 4.19: Cofficients of MLR including Sex_Render 63

Table 4.20: Cofficients of MLR including Marital_Render 64

Table 4.21: Cofficients of MLR including Education_Render 64

Table 4.22: Cofficients of MLR including Age_Render 64

Table 4.23: Cofficients of MLR including Career_Render 65

Table 4.24: Cofficients of MLR including Income_Render 65

Figure 2.1: Conceptual framework 7

Figure 3.1: Research process 10

Figure 4.1: Scree plot 58

Figure 4.2: Regression standardized residual 62

Figure 4.3: Normal P-P plot 62

Figure 4.4: Scatterplot 63

ABBREVATIONS

EFA : Exploratory Factor Analysis

GSO : Vietnam Government Statistics Office

HCMC : Ho Chi Minh City

Mil : Million

MLR : Multiple Linear Regression

UEH : University of Economic

CHAPTER 1 INTRODUCTION 1.1 BACKGROUND

As universal population levels continue to rise, the housing shortage in many developing countries has reached critical levels (Morel, 2001, p 1119) Real estate

is one of the most important things to citizens, so “the house purchase decision of them can change their life” (Wells, 1993) The house purchase decisions are different from other business decisions due to “the innate, durable and long-term characteristics of real estate” It is a highly differentiated product with “each specific site unique and fixed in location” (Kinnard, 1968)

In Vietnam, it is known as the third largest population in South East Asia and ranked the 14th largest in the world in terms of total population Its population estimated of

89 million in 2010 (GSO, 2011) The annual average growth population of Vietnam from 2000 to 2010 was approximately 1.03 million people per year or 1.2% annual growth Particularly, one of the top economic centers of Vietnam is Ho Chi Minh City which has around 7.2 million people as in April 2009, but its actual population

is likely to be significantly higher because of unrecorded migration from rural areas The real estate market in Vietnam has significantly changed during from the 1990s

to now and it might be seen as three times fever and declining prices in the last 20 years Up to the end of 2012, the large real estate outstanding loans and a big number of inventories created a serious crisis However, according to the Deputy Minister of Construction Nguyen Tran Nam, he emphasized that “people’s housing demand is very large and solvency is high, but the real estate market lacked of information”

1.2 RESEARCH PROBLEMS & RESEARCH QUESTIONS

In general, the real estate in Vietnam has got many difficulties in making effort to satisfy customer demands According to incomplete statistics of the Ministry of Construction surveyed in 44 provinces up to August 30th, 2012, the country now

had 16,469 unsold apartments, in which HCMC was 10,108 unsold apartments and total number of inventories of low buildings was 4,116, in which HCMC was 1,131 ones (Anh, 2012)

Therefore, the Prime Minister stressed that the solution to rescue real estate market should be included in the Resolution of the Government The main reasons of the crisis were the real estate market supply did not meet customer demands, the investors lacked of exact information of customer and real estate market conditions “There are two main fields of customer research are how customers go about making decisions and how decisions should be made In addition, “creating true value for customer and customer notion focused approach” is confirmed (Edwards

& Fasolo, 2001) It is found that “customer decision making is one of the most important areas of customer behavior and it requires gathering a lot of regarding

information” (Bettman et al., 1998 & Simonson et al., 2001)

There have been many published academic research about customer house purchase with variety of both developed and developing countries However, “the national and cultural characteristics play a very significant role in house purchase decision, that mean finding which is applied in specific context may not extend to another context” (Opoku & Abdul-Muhmin, 2010)

The real estate in Vietnam has got specific characteristics to which connected customer demands closely In recent years, researchers, domestic and foreign companies attracted to real estate field in Vietnam with a number of research works However, there has been not enough research into the way customers making decision to buy real estate as well as which major factors have got relationship with customer decision

Consequently, in the term of real estate purchase decision of customers, the research questions of the thesis are raised as two following questions:

 What are the key factors affecting the house purchase decision of customers in Vietnam?

 How is impact of these factors on house purchase decision of customers evaluated in Vietnamese context?

Understanding relationship between main factors affecting customer house purchase decision is an important role for both real estate developers and enterprises to satisfy customers’ demand and to have available strategies in the real estate field 1.3 RESEARCH PURPOSE

Based on the research questions, the main purpose of this thesis is to identify what factors have impact on house purchase dicision of customers and examine how these factors influence their decision of buying house in Vietnam

1.4 SCOPE OF THE RESEARCH

The research is conducted in Ho Chi Minh City with the respondents who are the postgraduates and students of UEH with various careers, as well as customers of a small book-coffee The timeframe of research lasts from the middle of September to the end of October in 2012

1.5 RESEARCH STRUCTURES

The research is divided into five chapters The first chapter introduces about background, research problems, research questions, research purpose, scope of research and research structures The second chapter covers literature review of the previous research and shows hypotheses, as well as the conceptual framework of the research The third chapter presents the research process, sampling size, measurement scale, main survey, and data analysis method The fourth chapter concentrates on preparation data, descriptive data, assessment measurement scale and hypotheses testing Finally, the fifth chapter points out research overview, research findings, managerial implications, research limitations and directions for future research

CHAPTER 2 LITERATURE REVIEW This chapter presents overview of previous literatures relating to housing purchase decision making of customers Also, a conceptual framework is built up and relative hypotheses of research are raised

2.1.1 Feature

Firstly, “features” of the building structure itself is an important determinant of a household choice of residence (Quigley, as cited in Haddad, 2011, p 234) Also, it

is confirmed that “feature” has significant effects on customers’ house purchase

decision making (Sengul et al., 2010, p 214) The “feature” of house includes

“design”, “house size” and “quality of building” determinants relating to decision

making to buy a house of an individual (Adair et al., 1996; Daly et al., 2003; Sengul

et al., 2010, p.218; Opoku & Abdul-Muhmin, 2010) As a result,

H1 There is a positive impact of house features on customers’ house purchase decision

2.1.2 Living space

Secondly, “private living space” is one of most important factors affecting to

“consumer housing decision” Living space consists of “size of living room”, “size

of kitchen”, “quantity of bathrooms” and “quantity of bedrooms” (Opoku & Muhmin, 2010, p.219) In addition, it is accepted that there is relationship between the “space customer” and customers’ purchase making process (Graaskamp, 1981) Accordingly,

Abdul-H2 There is a positive impact of living space on customers’ house purchase decision

2.1.3 Finance

Thirdly, “financial” status is much significant to customer house choice (Hinkle and

Combs, 1987, p.375; Kaynak & Stevenson, as cited in Sengul et al., 2010, p.220)

The “financial” element of real estate requires access to a relative large amount of

“capital” and as well as “borrowing costs” (Xiao & Tan, 2007, p 865) In addition,

“financial” status bases on combination of “house price”, “mortgage loans”,

“income” and “payment term” (Opoku & Abdul-Muhmin, 2010; Yongzhou, 2009,

p.17) Haddad et al (2011) finds out the “economic” factor which is consisted of

five variables, such as “income”, “interest rate”, “area”, “conversion” and “taxes”

Moreover, Adair et al.(1996, p.24) and Daly et al (2003, p.306) group “interest

rate”, “maximum mortgage”, “maximum monthly payment”, and “length of time payment” into “financial” factor Consequently,

H3 There is a positive impact of financial status on customers’ house purchase decision

2.1.4 Distance

Fourthly, one of the most important factors affecting individual “decision” making

to buy a house is “location” factor (Kaynak & Stevenson, as cited in Sengul et al.,

2010, p.219) The “residential location” has an influence on “people’s housing choice” (Zabel & Kiel, as cited in Opoku & Abdul-Muhmin, 2010, p.220) Distance

to choose house can be affected by “width of adjacent” and “location to school” (Opoku & Abdul-Muhmin, 2010) Moreover, “distance to central business”,

“distance to school” and “distance to work” are considered (Adair et al., 1996, p.23) In addition, “access to recreational facilities” and “access to main roads” are proposed (Iman et al., 2012, p.30) Hence,

H4 There is a positive impact of distance on customers’ house purchase decision 2.1.5 Environment

Fifthly, “environment” including “neighborhood”, “area attractiveness”, “view”,

“noise from around districts” and “general security” is stated as one of the determinants of a household’s residential decision (Adair, 1996, p.23) It is confirmed that “environment” has a big influent to housing buyer (Tajima, as cited

in Opoku & Abdul-Muhmin, 2010, p.224) and it is agreed by Morel et al (2001,

p.1119) Particluarly, “neighbourhood” quality is paid intention highly to house purchase decision making of customer (Gabriel & Rosenthal, 1989, p.240) Therefore,

H5 There is a positive impact of local environment on customers’ house purchase decision

2.1.6 Purchase decision

Customer behavior is an important research topic for recent decades “There is also

a clear shift from rational factors to psychological factors and to social decision factors” (Bargh, 2002) Beside, there is a link between the “intention to purchase” to

“decision to purchase” of customers, especially the decision related to purchase real estate (Ajzen, 1991, p 179; Han & Kim, 2010, p 659; Kunshapn & Yiman, 2011, p.7579)

2.1.7 Demography

“Demographic” characteristics of customers are internal factors related to decision making (Mateja & Irena, 2009) “Demographic” characteristics consist of the individuals in term of “gender, age, educational status, marital status, career, the quantity of family members and children, as well as the residence property”

“Demographic” characteristics consist of age (Yalch & Spangenberg, 1990),

education (Gattiker et al., 2000), income level (Dawson et al., 1990), gender (Zhang

et al., 2007) which are factors influenced on the “purchase intention” of customer

Particularly, “gender” has significantly influence on the financial feature of the

house (Sengul et al., 2010, p.214) It is also confirmed that there is a significant

difference in real estate buying decisions to “age” and “gender”, and not to

“educational levels” and “marital status” (Haddad et al., 2011) Correspondingly, in

this study, “gender” and “age” characteristics are considered as control variables so that investigate whether effect of those demography variables on housing purchase decision making of customers or not

CHAPTER 3 RESEARCH METHODOLOGY This chapter showed all steps of the research process, the minimum sample size, measurement scale, main survey and data analysis method

The research process was summarized as following steps

Step 1: Define the research problems, research questions and research purposes Step 2: Review the literature background from the previous research, then a conceptual model was set up and hypotheses were proposed

Step 3: Made and revise the draft questionnaire

A draft questionnaire with the measurement scales based on the previous research was set up Next, the draft questionnaire was delivered to 02 real estate professionals, 03 management officers to respond, and a discussion about the draft questionnaire was carried out later The aim of the pilot phase was to modify and clear the measure scale

After that, the revised questionnaires were delivered to another small group of 15 persons to test about clear understanding of the questionnaire Finally, a main survey was conducted with 263 receivers

Step 4: Conduct the main survey and collect data within 4 weeks

The questionnaires were directly sent to 263 persons The main respondents were postgraduates of master programs or students who have been studying to get the second business certification in the University of Economic Besides, a small group about 24 persons with a wide variety of careers was also delivered questionnaires at

a book coffee in Ho Chi Minh City Finally, there were 239 respondents giving their feedbacks, but 230 cases were available only

Step 5: Edit, code and adjust missing data before testing reliable and validity of data

In order to prepare the data to analysis, data were edited, coded and adjusted for missing data Next, reliability of measuring instrument was analyzed by calculation

Cronbach’s alpha which was required above 7 (Hair et al., 2010) In addition,

validity of measuring instrument was evaluated due to define the number extracted factors based on the Eigenvalue value over than 1 and changing of the slope in the

Scree plot (Hair et al., 1998; Tabachnick & Fidell, 2001)

Step 6: Test the hypotheses of research and define relationship of factors in model

through the Multiple linear regression analysis

The Multiple linear regression analysis was applied to evaluate the relationship between five independent variables, including “feature”, “living space”, “finance”,

“distance” and “environment” and one dependent variable, namely “decision” Moreover, defining whether there was any significant contributory of control variables consisting of “gender”, “age”, “marital”, “income”, “education” and

“career” on customers’ housing purchase decision was also analyzed by the multiple linear regression All steps were illustrated by the following Figure 3.1

Figure 3.1: Research process

Sampling DesignType, purpose, time frame, scope, environment

Instrument Development

3.2 SAMPLE SIZE

The reliable and validity of variables were tested by using Cronbach’s Alpha and EFA, after that the multiple regression was applied to test model and hypotheses First of all, the sample size was required to have enough quantity for the analysis

The minimum sample size was 100 and not less than five times of items (Hair et al

2010), thus:

n > 100 and n = 5k (where k is the number of items)

Thus, the minimum sample size was 5x34 = 170 samples

In addition, based on five independent factors of the conceptual model, the multiple regression analysis required sample size at least (Tabachnick & Fidell, 2007):

50 + (8xm) = 50 + (80x5) = 90 samples

Where m: is the number of independent factors of the model

Consequently, the minimum sample size should be 170 Based on the actual collection data, the quantity of available respondents from the questionnaire survey estimated 230, so that samples met the requirements above

3.3.1 Measurement scale

In order to operate concepts, it was necessary to measure them in some manners, so different variables were required to choose an appropriate scale The independent variables were applied interval scale with five - point of Likert scale consisting of totally unimportant (1), unimportant (2), neutral (3), important (4), very important (5); beside, the dependent variable was applied the same measure consisting of strongly disagree (1), disagree (2), neutral (3), agree (4) and strongly agree (5) 3.3.2 Pilot test

In order to test logistics of the questionnaires prior collection data on large cover, a pilot test was carried out with a small group consisting of two real estate

professionals of Sacomreal and three management officers of Hoa Binh Corporation All of them had much knowledge and many experience years in the real estate field

Firstly, the aim of the pilot test was explained to all of them; moreover, the questionnaires and relative documents were also sent to them After that, a discussion with them was conducted to define which parts would be deleted or which parts would be added The results were presented in Appendix 01

For items of the “house feature” factor, the item “type of finishing” and “quality of finishing” should be deleted because their content was inside the content of

“construction quality”

While all items of “private living size” factor were agreed, the item “tax” of

“Finance” factor should be changed into “the registration fee”

For “distance” factors, the “house on a main bus route” item should be deleted because this item was not paid attention by customers The “distance from the house

to shopping centre” item was also proposed to delete because it was too specific and related to female only In addition, the group recommended that customers had got tendency to ignore the “location away from industrial areas” item so this item should be removed

For “environment” factor, its “the attractiveness of the area” item had got the same meaning of “view” item, so “the attractiveness of the area” should be deleted

The last “decision” factor, it should change “I will want to buy a new house” into “I will make my effort to buy a new house”

Finally, after adjusting the first questionnaire table, a small sample size of fifteen convenient colleagues was delivered the questionnaires to recognize whether any parts of its unclear to understand or misunderstand However all of them understood meaning of questionnaires quite well and knew the way to answer, so the

questionnaire was the last version to carry out in the massive areas After that, a main survey was conducted

From above discussion above, a summary table of main factors affecting customer’ housing decision making is presented as following Table 3.1

Table 3.1: Main factors affecting customers’ housing purchase decision

Adair et al (1996), Daly et

al (2003), Kaynak &

Tevenson (1982), Haddad

et al (2011), Opoku &

Abdul-Muhmin (2010), Ratchatakulpat (2009),

Sengul et al (2010), Xiao

3

House price X3.1 Adair et al (1996), Daly et

al (2003), Kaynak &

Income X3.5 (2009), Sengul et al

(2010), Xiao & Tan (2007)

The registration fee X3.7

4

Width of adjacent street X4.1

Adair et al (1996), Daly et

al (2003), Haddad et al

(2011), Opoku & Muhmin (2010), Ratchatakulpat (2009),

Abdul-Sengul et al (2010), Xiao

& Tan (2007)

Distance to market X4.3Distance to school X4.4Distance to recreation centre X4.5Distance to the central

Adair et al (1996), Daly et

al (2003), Haddad et al

(2011), Opoku & Muhmin (2010), Ratchatakulpat (2009)

al (2003), Haddad et al

(2011), Mateja (2009), Ratchatakulpat (2009),

Sengul et al (2010), Xiao

Planning to buy a new house X7.1 Ajzen (1991), Han & Kim,

(2010), Kunshan & Yiman,

(2011) Making effort to buy a new

3.4 MAIN SURVEY

The questionnaire survey was conducted at the ISB-Mbus class and four of the economic night classes of UEH in 59C Nguyen Dinh Chieu Street Besides, three of the economic night classes of UEH in Nguyen Tri Phuong Street were also delivered the questionnaires The last surveyed place was a small PNC book coffee

in Nguyen Oanh Street Timeframe to survey was from the middle of September,

2012 to at the end of October, 2012

There were 263 hand-delivered questionnaires, only 239 respondents gave feedback immediately, but quantity of available respondents was 230

After data collection, the first step would be data preparation with editing, coding, and data entry to ensure accuracy of data from raw data and to detect errors or omissions to correct Next, data were classified to arrange them into groups or classes of common demographic

Finally, variables would be tested reliable by Cronbach’s alpha, validity by EFA, and hypothesis and model would be tested by multiple regression of SPSS

3.5.1 Reliability measure

In order to check reliability of each of scales with particular sample, as well as consider the internal consistency of the scales, it was necessary to use Cronbach’s Alpha coefficient which should be above 7 (Devellis, 2003)

Also, the corrected item - total correlation values should be at least 3 to ensure each

of items was measuring the same from the scale as a whole (Pallant, 2011)

An important person affecting house purchase decision

X7.3

3.5.2 Validity measure by EFA (Exploratory Factor Analysis)

In order to evaluate the validity and the correlation among variables to identify underlying factors or define number of extracted factors, EFA was applied with the oblique approach using the Promax method However, some requirements of EFA should be satisfied (Pallant, 2011):

- The minimum of sample size should be at least 100 and rate of observations per items of models should be five cases for each of the items, so that meant the minimum required sample size should be at least 5m = 5x34 = 170 cases (where m: quantity of items from the conceptual model) The actual sample size was 230, bigger than 170 so it met the requirement

- The correlations of r of the correlation matrix should show at least 3

- Kaiser-Meyor-Olkin (KMO) test must be equal or above 6 (Tabachnick & Fidell, 2007)

- Barllett’s test of sphericity should have significant less than 5%

- In order to extract factors, the eigenvalue of factors must be greater than 1 (Kaiser, 1956)

3.5.3 Multiple regression analysis

To explore the relationship between independent variables, consisting of “features”,

“living space”, “finance”, “distance” and “environment”, and dependent variable, namely “decision” as well as to evaluate the importance of those independent variables in the framework model, the multiple regression analysis was conducted The multiple regression analysis required that some following conditions should be satisfied:

- The minimum sample size based on the formula:

n > 50 + 8m = 50 + 8x5 = 90 samples, where m: number of independent variables

in the conceptual model

The actual quantity of cases was 230, so this condition was satisfied

- The multicollinearity did not exist, so r value, the correlated score was less than 9

- The collinearity test on variables was via two values “tolerance” and “VIF”, particularly the VIF should not be less than 1, or above 10

- The Normal probability plot (P-P) was required with most of the scores concentrated in the centre (along the 0 point)

- The presence detection of outliers was considered from the Scatterplot

The multiple regression was used to test hypotheses, to explore the relationship between five INDEPENDENT VARIABLEs and one dependent variable, and to consider whether control variables supported or not to dependent variable The generalized equation (Donald & Pamela, 2006) was:

Y = o + 1X1 + 2X2 + 2X2 + …+nXn + 

Where:

o = a constant, the value of Y when all X values are zero

1 = the slope of the regression surface (the  represents the regression coefficient associated with each Xi)

 = an error term, normally distributed about a mean of 0

CHAPTER 4 DATA ANALYSIS & RESULTS This chapter presented data preparation with editing, coding, and data entry from raw data to correct errors Then data were described through frequency tables about the general information Using Cronbach’s alpha to test the reliability of variables and EFA to test their validity, then the multiple regression was run to explore the relationship between independent variables and dependent variable, and to test hypotheses

4.1.1 Editing

After collection 239 cases from respondents, all cases were checked first There were 03 cases of blank sheets, 02 cases of filling in half of I part only, 01 case of no filling in the general information part, and 03 cases of filling almost choosing number 1 or 3 or 4 The last available numbers of cases was 230, and each of all cases was marked a reference number on it to find easily Others did not have any cases of missing data for contend of INDEPENDENT VARIABLEs and dependent variable

4.1.2 Coding

Answers were assigned numbers of symbols so that the responses were grouped into

a limited number of categories (see Table 4.1)

Table 4.1: Codebook of questionnaire items

No Factors Code Description

ble name

Varia-Coding/

Creating Dummy

1 Feature House size Fea01 Record respondents’ numbers

4 External Design Fea04

8

Private

Living

Space

Kitchen Size Liv01 Record numbers

13 Finance House Price Fin01 Record numbers

20

Distance Width of adjacent

street

Dis01 Record numbers

Env01 Record numbers

33 Decision Plan to buy a new

house

Dec01 Record numbers

a new house

Dec02

to make decision to buy a new house

1 = “less than 35”; 0 = “above 35”

1 = “Single”; 0 = “Married”

4.2 DESCRIPTIVE DATA

According to Table 4.2, there were 230 available respondents, the male was thirds of total of cases and almost respondents were single with percent of 83 percent Also, 61.3 percent respondents graduated university and 31.7 percent postgraduates studying master programs Their ages range from 18 year olds to 35 year olds with 99.1 percent of total cases Almost all of them were officers with their ages at least 18 years old and less than 36 years old Besides, the main career

two-of respondents was two-officers with 87.8 percent per total two-of cases, their income was less than 15 million per month with 89.6 percent rate, while the group of managers

or owners at least 15 million per month with 3.9 percent rate

Also, the single house was chosen most with 73.6 percent rate, the second choice of type of house was apartment with 21.6 percent rate The house price which was less than 20 mil./m2 was appropriate with 87.3 percent of cases and the type of small and medium house size of less than 100 square meters was chosen most with 84.3 percent rate

1 = “less than or equal 14 mil.”

0 = “more than 14mil.”

1 = “not yet graduated university”; 0 = “graduated university”

1 = “staff”; 0 = “management board”

Table 4.2: Characteristics of respondents

Characteristics Frequency % Cumulative %

Characteristics Frequency % Cumulative %

In order to evaluate appropriation of a measurement scale, the scale should be checked its reliability and validity The reliability was tested by Cronbach’s Alpha and the validity was tested by EFA

4.3.1 Cronbach’s Alpha

Refer to the Case processing summary of all variables of five concepts, the number samples of each concepts was valid with 230 available cases

Based on the Reliability Statistics Table 4.3 and Table 4.9, all Cronbach’s Alpha values of all concepts were above 7 after deleting items that the “Corrected item-Total correlation” values of them were less than 3

For “feature” concept, both “Fea06_Construction duration” item and “Fea07_Type

of house”, their “Corrected item-Total correlation” values were 253 and 08 For meaning consideration, the “feature” concept could be measured by remaining items, so both of them should be deleted After they were removed, the Cronbach’s Alpha value increased from 748 to 830

For “private living space” concept, the “Corrected item - Total correlation” value of

“Liv04_Living room size” item was low with 371, of “Liv05_Storey of house” item was 012 All other items ensured the content of “living space”, so those two items should be deleted and the Cronbach’s Alpha value of “private living space” increased from 619 to 739, this value was not high, but it could be acceptable For “finance” concept, the “Corrected item - Total correlation” value of

“Fin07_Registration Fee” item was 013, less than 0.3, so this item was deleted Also, “Corrected item - Total correlation” value of “Fin02_Max Mortgage” item was quite low 36, it should be also deleted The Cronbach’s Alpha value of “private living space” increased from 726 to 865, this value was so quite good

For “distance” concept, the “Corrected item - Total correlation” value of

“Dis06_Business Distance” item and “Dis07_Main Access” which were 139 and 202 Those values were too low compared with 3, so they should be deleted Beside, the “Corrected item - Total correlation” value of “Dis05_Recreation Distance” was 377, it also should be deleted The Cronbach’s Alpha value of

“Distance” concept was increased from 765 to 890

For “environment” concept, the “Corrected item - Total correlation” value of

“Env03_View” item and “Env06_Nearby traffic” which were 284 and 272 Those values were less than 3 and the “Env06_Nearby Traffic” and “View” could be

explained by characteristics of “Noise”, “Pollution”, “Neighbour Condition”, and

“Security” of the environment, so they were removed The Cronbach’s Alpha value

of “environment” concept was increased from 767 to 846 after deleting two items above

Finally, “decision” concept had got all the “Corrected item - Total correlation” value of all items were above 4, and its Cronbach’s Alpha value was 816, those values were quite good

Table 4.3: Cronbach’s Alpha test results

Variables

Scale Mean if Item Deleted

Scale Variance

if Item Deleted

Corrected Item-Total Correlation

Cronbach's Alpha if Item Deleted

4.3.2 Exploratory Factor Analysis (EFA)

Exploratory factor analysis was carried out through three steps consisting of the 1ststep was evaluation the suitability of data for factor analysis, the 2nd step was factor extraction, and the 3rd step was factor rotation and interpretation (Pallant, 2011)

Sample size

Sample size and the strength of the relationship among the variables were required

to test suitability of data (see Table 4.12) The sample size was 230 available cases and shown detail in chapter 3 that this requirement was met the minimum required sample size

Factorability of the correlation matrix

From the pattern matrix, low communality values should be removed to increase the total explained variance Removing some low communality variables and repeating the same analysis, the result of EFA was presented in the Table 4.4 From the partial Correlation Matrix Table 4.5, some of the correlation coefficients between variables each others were above 3

The KMO value was 779, exceeding value of 6 and Bartlett’s test of Sphericity value was 000, that means less than the statistically significant at p < 05 (see Table 4.10)

Therefore, the condition about the factorability of the correlation matrix was appropriated with assumptions of EFA

4.3.2.2 Defining number of extracted factors

From Total variance explained value in Table 4.11, there were first six components with eigenvalues above 1 including following values: 6.335; 3.099; 2.860; 2.259; 1.647 and 1.166 And those components explained total 64.26% of the variance, exceeding than 50% explained total, so this value was appropriate

In addition, the Scree Plot Figure 4.1 showed that there was hard break from components 2 and 3, and both of components 1 and 2 xplained 36.64% than four remaining components However, there was slight break after component 6, so the number of extracted factors was six

From the Factor Matrix Table 4.13, the first component presented most of the items loaded on it while the second three components loaded quite the same, and the final two components loaded the least

In the Pattern Matrix Table 4.4, all items loading on six components were above 4 Besides, there were five items loading on component 1, four items loading on component 2, five items loading on component 3, three items loading on component

4, four items loading on component 5, and four items loading on component 6

Table 4.4: EFA results