VIETNAM NATIONAL UNIVERSITY – HO CHI MINH CITY INTERNATIONAL UNIVERSITY
SCHOOL OF BUSINESS
MOBILE USERS' ATTITUDES TOWARD
MOBILE ADVERTISING IN VIETNAM
A STUDY OF MOBILE WEB DISPLAY ADVERTISING
AND MOBILE APP DISPLAY ADVERTISING
In Partial Fulfillment of the Requirements of the Degree of
BACHELOR OF ARTS in BUSINESS ADMINISTRATION
Student’s name: NGUYỄN HỒ BẢO TRÂN (BAIU09153)
Advisor: LÊ ĐÌNH MINH TRÍ, M.B.A.
Ho Chi Minh City, Vietnam
MOBILE USERS' ATTITUDES TOWARD
MOBILE ADVERTISING IN VIETNAM
A STUDY OF MOBILE WEB DISPLAY ADVERTISING
AND MOBILE APP DISPLAY ADVERTISING
APPROVED BY: Advisor APPROVED BY: Committee,
MBA. Lê Đình Minh Trí Ph.D Phạm Hồng Hoa, Chairman
Ph.D Nguyễn Văn Phương
Ph.D Mai Ngọc Khương
MA. Nguyễn Hữu Đăng Khoa
The completion of this bachelor’s thesis is a difficult but interesting task. It is a chance for me to apply knowledge and skills learned throughout four years in International University. During the time of research, I have received supports and helps from many people. I would like to take this opportunity to express my gratitude to all those persons.
In particular, I am really grateful to the help of my kind advisor, Le Dinh Minh Tri, whose gave me valuable recommendations and encouraged me in every step to complete this thesis. His willingness to share his precious time and knowledge to assist me is honorable. My grateful thanks are also extended to Dr. Nguyen Quynh Mai and Ms. Ta Diu Thuong for giving me worthy advices to improve my thesis.
I would like to thank participants for their enthusiastic assistance with the collection of my data. Many thanks sent to my friends in International University and their relatives for their contribution in pilot study. It is also appropriate to thank for spiritual support, so I’d also include my friends and parents for their encouragements.
Nevertheless above persons have supported me so much, I am the only person responsible for the errors in the thesis. My family is my most precious fortune and this thesis is dedicated to them.
TABLE OF CONTENTS
LIST OF TABLES ... vii
LIST OF FIGURES ... viii
ABSTRACT ... ix CHAPTER I: INTRODUCTION ...1 1.1. Background ...1 1.2. Research Purposes ...1 1.3. Research Objectives ...1 2.1.1. Research Questions ...2 2.1.2. Hypotheses ...2 2.1.2. Objectives ...2
1.4. Rationale of the Research ...2
1.5. Significance of the Research ...3
1.6. Organization of the Thesis ...3
CHAPTER II: LITERATURE REVIEW ...5
2.1. Definition of Terms ...5
2.2. Mobile Advertising ...6
2.2.1. Advertising and Advertising Message ...6
2.2.2. Mobile Advertising ...6
2.3. Consumers’ Attitudes ...7
2.4. Attitudes toward Mobile Advertising ...8
2.4.1. Attitudes toward Advertising ...8
2.4.2. Factors underlying Attitudes toward Mobile Advertising ...8
22.214.171.124 Informativeness ... 10
126.96.36.199 Entertainment ... 10
188.8.131.52 Irritation ... 11
184.108.40.206 Credibility ... 11
CHAPTER III: METHODOLOGY ... 12
3.1 Research Design ...12
3.2 Research Model and Hypotheses ... 13
3.2.1 Research Model ...13
3.2.2 Hypotheses ...13
3.3. Questionnaire Design ...14
3.3.1 Questionnaire Design Process ...14
3.3.2 Information Requirements ...14
3.3.1 Format of Questions ...15
3.3.2 Sequence and Layout ...15
3.4 Sampling ... 16
3.5 Data Collection Method ... 17
3.6 Data Analysis Techniques ...17
3.6.1. Internal Consistency Test ...17
3.6.2. Factor Analysis ...18
3.6.3. Multiple Regression Analysis ...18
3.6.3. Chi-square Test for Independence ...18
3.7 Data Quality ...18 3.7.1. Reliability ...18 3.7.2. Valid ...19 220.127.116.11. External Validity ... 19 18.104.22.168. Internal Validity ... 20 3.8 Pilot Study ...20
CHAPTER IV: DATA ANALYSIS AND FINDINGS ...21
4.1 Data Screening ...21
4.2 Description of the Sample ...23
4.3 Descriptive Statistics of Attitudes toward Mobile Advertising ...26
4.4 Internal Consistency Test ...27
4.4.2 Informativeness ... 28
4.4.3 Entertainment ... 29
4.4.4 Irritation ... 29
4.4.5 Credibility ... 30
4.5 Principal Axis Analysis (Common Factor Analysis)...31
4.5.1 Appropriateness of the data ... 31
4.5.2 Factor Structure ... 35
4.5.3 Correlation among Variables ... 39
4.6 Multiple Regression Analysis ...39
4.6.1 Model Fit ... 39
4.6.2 Statistical Significance of Independent Variables ... 41
4.7 Chi-square Test for Independence ...42
4.7.1 Attitudes and Gender ... 42
4.7.2 Attitudes and Age ... 43
4.7.3 Attitudes and Education ... 44
4.7.4 Attitudes and Income ... 45
CHAPTER V: CONCLUSION AND RECOMMENDATION ... 48
5.1. Summary and Discussion ...48
5.2. Limitation of the Study ... 50
5.3. Recommendation... 50
5.3.1. Recommendation for Practical Applications ...51
22.214.171.124 Credibility ... 51
126.96.36.199 Entertainment ... 51
188.8.131.52 Target Segments ... 51
5.3.2. Recommendation for Further Research... 52
5.3. Conclusion ... 53
LIST OF REFERENCES ... 54
APPENDIX 1: QUESTIONNAIRE 1 (Vietnamese Version) ...56
LIST OF TABLES
Table 1: Mobile advertising revenue by type, worldwide, 2012-2016 (millions of US
Dollars) ... 6
Table 2: Previous Studies of Attitudes toward Advertising ... 9
Table 3: Information requirements of Questionnaire ... 14
Table 4: Layout and Scales of Questionnaire ... 15
Table 5: Characteristics Related to Mobile Devices of the Sample ... 25
Table 6: Reliability Statistics of Attitudes toward Mobile Advertising Items ... 28
Table 7: Reliability Statistics for Informativeness ... 28
Table 8: Reliability Statistics for Entertainment ... 29
Table 9: Reliability Statistics for Irritation ... 29
Table 10: Reliability Statistics for Credibility ... 30
Table 11: Retained Items after Internal Consistent Test ... 31
Table 12: KMO and Bartlett’s Test ... 32
Table 13: Measure of Sampling Adequacy for Individual Variables ... 32
Table 14: Communalities ... 33
Table 15: Total Variances Explained ... 34
Table 16: Rotated Factor Matrix ... 35
Table 17: Modified Items and Variables ... 37
Table 18: Pearson Correlations among Variables ... 38
Table 19: Model Summary ... 39
Table 20: ANOVA for Multiple Regressions ... 40
Table 21: Coefficients for Multiple Regressions ... 41
Table 22: Chi-Square Test for Independence between Gender and Attitude ... 42
Table 23: Chi-Square Test for Independence between Age and Attitude... 42
Table 24: Crosstab between Age and Attitude ... 43
Table 25: Crosstab between Education and Attitude ... 44
Table 26: Chi-Square Tests between Income and Attitude ... 45
LIST OF FIGURES
Figure 1: The mobile advertising ecosystem ... 7
Figure 2: Proposed Model of Attitudes toward Mobile Advertising... 13
Figure 3: Demographic Characteristics of the Sample ... 24
Figure 4: Histogram of A1 – Positive Emotions ... 26
Figure 5: Histogram of A3 – Receive Product Information ... 26
The rapid development of smartphones has resulted in the increasing use of mobile devices to deliver advertisements for products and services. Mobile web and mobile application has become a potential advertising channel. Many firms in Vietnam have invested numerous resources in this field and a wise understanding of mobile advertising is necessary to develop a successful mobile advertising strategy. Based on the existing literature about attitudes toward mobile advertising, a questionnaire is constructed to illustrate the factors affecting consumer attitudes toward advertisements on mobile web and mobile application. Informativeness, Credibility, Entertainment, Irritation and demographic characteristics are factors studied in this research. More than two hundred people participated in this research by using convenience sampling method. The findings of this study showed that although many consumers do not have good feelings toward advertising, they cannot ignore the importance of mobile advertising. If mobile advertisers can present credibility and entertainment in their advertisements, consumers will be willing to see the ads and intent to buy products and services. The results also indicated that demographic characteristics have somewhat relationships with attitudes towards mobile advertising. Thus it is not good idea to use mobile advertising to advertise for target segments that
The development of high-tech devices has led to the changing of consumers’ habits over time. Along with the boom of tablets and smartphones, mobile devices became the top channel for media. Vietnamese people spend 35 percent of average hours of media daily (4.5 hours) on mobile devices, more than on TV (25 percent) and computers (18 percent). Mobile devices play an important role in consumer behavior. A half of mobile web users are most impacted by mobile when making purchase decisions and three-quarters of users feel comfortable with mobile advertising as TV or online ads. (InMobi, 2012).
These statistics showed that mobile advertising is the potential form of advertising in the future. In order to take the advantage of this opportunity, marketers have taken more attention on mobile advertising. There are 2.1 billions of adverts are served to mobile devices. (Kemp, 2012) This thesis is considered as a support for marketers to understand consumers’ attitudes and get more attraction from them.
The development of high-tech portable devices such as tablets and smart phones has led to the rise in mobile advertising. However, these advertisements do not get attention as much as other forms of advertising like Television or the Print Media The purpose of this thesis is to provide insight into mobile marketing industry, analyze important issues related to the mobile users’ attitude toward mobile advertising, do further research on consumer behavior and build effective mobile marketing strategies.
1.3.1 Research Question:
The following questions will serve as a basis for addressing the primary research:
What are consumers’ attitudes toward mobile advertising?
Which underlying factors of mobile advertising contribute to consumers’ attitudes? How do attitudes toward mobile advertising differ among demographic segments?
1.3.2 Research Hypothesis
In general, consumers have a positive attitude toward mobile advertising.
Consumers’ evaluation of factors underlying advertising positively contributed to attitudes
Demographic variables have a significant impact on factors underlying consumers’ attitudes toward mobile advertising.
To evaluate the overall mobile users’ attitudes toward mobile advertising To investigate the factors underlying consumers’ attitudes.
To identify the relationship between demographic segments and attitudes toward mobile advertising
1.4.Rationale of the research
Advertising always plays an important role in marketing industry. Awareness of products and brands that affect purchasing decision of consumers is an example for advertising importance. Especially mobile advertising can cover a wide range of audiences of all ages, locations and occupations. To take fully advantage of mobile advertising, a study on consumers’ attitudes toward mobile advertising is very necessary to attract target audiences and improve the effectiveness of advertisements.
This thesis does not only base on existing theories and previous researches, but it also improves the applicability of Consumers’ Attitudes. Studies of Consumer Attitudes toward Mobile Advertising were conducted in many countries before but there is no similar research in Viet Nam. Besides that, because of the rapid changes of
technology, almost the research focused on SMS Advertising and did not include new types of mobile advertising like Mobile Web Banner or Mobile Application Advertising. This thesis will cover recent popular types of mobile advertising: Web banner and poster, in-App, Pop-up, Pop-under ads, etc. and can be applied for new mobile marketing strategies that catch up the development of high technology.
1.5.Significance of the research
The findings of this thesis can be applied in many cases for many users: from the advertisers, marketers to mobile advertising networks (For example: Admicro, Sosmart, MAD, Fmob), website and mobile application developers. It helps advertisers and marketers to have better mobile marketing strategies and better advertising design, especially for Vietnamese people. Mobile advertising networks can refer this research to consult their clients and offer effective solutions.
Furthermore, the research of consumers’ attitudes is the basic foundation to do further research on Vietnamese consumer behavior toward mobile advertising. Based on results of this thesis, studies which determine other factors affecting attitudes can be conducted to have a more accurate insight into attitudes toward mobile advertising. Another related further research is studies on consumer attitudes towards advertised brands on mobile advertising, consumers’ behavioral intention and actual behavior according to Theory of Reasoned Action (Aizen & Fishbein, 1980)
1.6.Organization of Thesis
Thesis can be divided into five chapters: introduction, literature review, methodology, data analysis and result, discussion and conclusion.
This first chapter gives an introduction about the research study. From the background of study, problems leading to the necessity of this research are recognized. To solve the current problems of mobile advertising, research objectives are developed. Significance and
Rationale is also specifically described in order to express the importance of this study. The second chapter – Literature Review will provide basic theories used in
the thesis. Theories about attitudes, mobile advertising, attitudes toward advertising and its underlying dimension can be found here.
The third chapter is methodology section. All methods from sampling, collecting data, designing questionnaire to analyzing data are explained in this section. Another parts included in this section is data quality and the procedure of pilot study.
The fourth chapter, data analysis and results, will describe the process of analysis and show results. The hypotheses will be tested using the achieved results.
The last chapter, discussion and conclusion will give an explanation of results in the fourth chapter. This chapter will come up to a conclusion for this thesis.
2.1 Definition of Terms
- Advertising: any paid forms of non-personal presentation and promotion of ideas, goods or services by an identified sponsor (Koler & Armstrong, 2010).
- Attitude toward Advertising: a learned predisposition to respond in a consistently favorable or unfavorable manner toward advertising in general (MacKenzie & Lulz, 1989).
- Attitude: A lasting, general evaluation of people (including oneself), objects, advertisements, or issues (Solomon, 2013).
- Landing Page: The web page where customers end up after they click the ad. This page is usually the same as the ad’s destination URL. (Google Adwords, 2013)
- Mobile Advertising: A form of advertising that is communicated to the consumer/target via a handset. This type of advertising is most commonly seen as a Mobile Web Banner (top of page), Mobile Web Poster (bottom of page banner), and full screen interstitial, which appears while a requested mobile web page is “loading.” Other forms of this type of advertising are SMS and MMS ads, mobile gaming ads, and mobile video ads (pre, mid and post roll) (Mobile Marketing Association).
- Mobile Marketing: A set of practices that enables organizations to communicate and engage with their audience in an interactive and relevant manner through any mobile device and network (Mobile Marketing Association).
- Pop-under ad: Opens underneath a user’s active browser window and does not appear until the user closes the active window (Laudon & Traver, 2009).
- Pop-up ad: Banners and buttons that appear on the screen without the user calling for them (Laudon & Traver, 2009).
2.2 Mobile Advertising
2.2.1 Advertising and Advertising Message
Advertising is defined as any paid forms of non-personal presentation and promotion of ideas, goods or services by an identified sponsor (Koler & Armstrong, 2010). No matter what advertising objectives, strategy and media types are, advertisements have to get attention from target audiences among 3000 to 5000 commercial messages every day (Koler & Armstrong, 2010). Therefore, the advertising message needs to be examined carefully before publishing advertisements.
2.2.2 Mobile Advertising
Mobile advertising communicates with the target audiences via a handset. There are many types of mobile advertising. The popularity of these forms has changed over time. According to Gartner, the world’s leading information technology research and advisory company, mobile web display, in-app display and search/maps are three types of mobile advertising that have highest revenue. In addition, Mobithinking, predicts that the revenue of these types will continue to grow up in the future.
Mobile advertising revenue by type, worldwide, 2012-2016 (millions of US Dollars)
(Source: (Mobithinking, 2013))
Type 2012 2013 2016
Mobile Web Display 2,332.4 2,855.0 7061.0
In-app display 2,927.7 3,117.7 6,754.0
Search/maps 3.893.9 4,533.7 7,859.2
Audio/video 380.6 696.6 2.627.9
SMS/MMS/IM 224.5 217.0 257.9
This thesis will just focus on advertising on mobile website and in mobile applications, which are common types of recent mobile advertising. In another way, it can be said that mobile advertising has a close relationship with online advertising.
Mobile advertising ecosystem has many players: brands, advertising agencies, advertising networks, publishers and users. The relationship among these players is showed in figure 1.
The mobile advertising ecosystem an
In order to attract more customers, brands suggest the ad agencies to design advertising campaigns. Mobile Advertising Network has a responsibility as a distribution, an intermediary to share the advertisement to audiences through mobile web sites or applications. Publishers will place advertisements on their web sites or apps and make money. The final destination of this ecosystem is mobile users that are potential customers of brands. Their attitudes toward mobile advertising will be discussed in this thesis.
Attitude is a lasting, general evaluation of people (including oneself), objects, advertisements or issues (Solomon, 2013), a mental state used by individuals to structure the way they perceive their environment and guide the way they respond to it. The evaluation of people is more complex than whether they like or dislike an object simply.
In order to have a comprehensive view of attitudes, the ABC Model of Attitudes was developed. This model divide attitude into three components: Affect, Behavior and Cognition which are referred as the verbs “feel, do and think”. Affect is the feeling of a consumer about an object. Behavior refers the intention of consumer to do something. Notice that the meaning of behavior in this model is the intention, not the actual behavior. Cognition is what a consumer believes about an object. These three components have a close relationship with each other. Depending on the situation, the relative impact of these components which is known as hierarchies of effects are diversified (Solomon, 2013).
Similarly with ABC Model, (Aaker, Kumar, & Day, 2000) form an attitude into three components: cognitive and knowledge, affective or liking and intention or action components. Cognitive or knowledge represents a person’s information about an object. The affective or liking component summarizes a person’s overall feelings toward an object, situation, or person. Intention or action component refer to a person’s expectation of future behavior toward an object. (Aaker, Kumar, & Day, 2000)
2.4Attitudes toward Mobile Advertising
2.4.1 Attitudes toward Advertising
Attitude toward Advertising is defined as a learned predisposition to respond in a consistently favorable or unfavorable manner toward advertising in general. Consumer Attitudes toward Advertising tend to affect their attitudes toward specific advertisements. (MacKenzie & Lulz, 1989).
2.4.2 Factors underlying mobile consumers’ attitudes
Many researchers around the world developed many models that pointed out determinants of attitues toward advertising. Some of them are listed in the following table:
Previous Studies of Attitudes toward Advertising
Title of Study Factors affecting attitude toward advertising
Advertising Value and Advertising on the Web (Ducoffe, 1996)
Informativeness, Entertainment, Irritation
The influence of personalization in affecting consumer attitudes toward mobile advertising in China (Xu, 2006 - 2007)
Entertainment, Credibility, Personalization
Public Attitudes Toward Advertising: More Favorable Than You Might Think (Shavitt, Lowrey, & Haefner, 1998)
Enjoyment and Indignity,
Trustworthiness or Usefulness of Ad Content, Demographic Segments
The influence of consumer socialization variables on attitude toward advertising: A comparison of African-American and Caucasians (J.Bush, Smith, & Martin, 1999)
Parental Communication, peer communication, mass media, gender, race
Hispanic Attitudes toward Advergames: A proposed Model of their Antecedents (Hernandez, Minor, & Maldonado, 2004)
Intrusiveness (Congruence, Extended Exposure) and Irritation (Entertainment)
Are we measuring the same attitude? Understanding media effects attitudes towards advertising (Tan & Chia, 2007)
Materialism, Good for economy
variety of other factors. Because this thesis concern on mobile web displays ads and in-app ads, mobile advertising throughout the research is a part of internet advertising. In an article about advertising value and advertising on the web (Ducoffe, 1996), three perceptual antecedents (Informativeness, Entertainment and Irritation) influence on how consumers assess the value of web advertising. Additionally, the findings of this research also pointed out those consumers’ assessments of value have a significant impact on their overall attitudes. Therefore, informativeness, entertainment and irritation are factors that should to be considered when examine attitudes toward mobile advertising.
Consequently, two hypothesized additional variables: Credibility and Demographic Variables were added to the Ducoffe model in an article: Cyberspace advertising vs. other media (Bracket & Carr, 2001) and tested that they strengthened this model.
This thesis will focus on four hypothesized factors: Informativeness, Entertainment, Irritation and Credibility.
Informativeness is a condition of providing useful or interesting information (Oxford, 2013). Informativeness includes a good source of product information, the ability to supply relevant product information and provide up-to-date information (Bracket & Carr, 2001). Previous research has proved the importance of informativeness in advertising. Including specific information in an ad increased its chances of being included in the consideration and calling sets as well as being the advertiser selected to call and visit first (Fermandez & Rosen, 2000)
Entertainment refers to the enjoyment of the message. In a survey of attitudes toward enjoyment, the majority of respondents agree that they like to look at most of the advertisements they are exposed. Practical situations have showed that entertainment like humor easily attracts consumers. With a variety of entertainment tools like music, games, visuals, mobile advertising is a promising form for entertaining advertising. Therefore,
entertainment may be a factor influencing attitudes toward mobile advertising. A research has found that people’s feelings of enjoyment associated with advertisements played the strongest role in accounting for their overall attitudes toward advertising. (Shavitt, Lowrey, & Haefner, 1998)
The meaning of irritation is the state of feeling annoyed, impatient, or slightly angry (Definition of irritation, 2013). The feeling of intelligence insulted, annoying and irritating is elements in irritation (Bracket & Carr, 2001). Internet Advertising is considered less irritating than general advertising because the interactivity of internet advertising allows consumers to tailor the ad to meet their individual needs (Scholosser, Shavitt, & Kanfer, 1999). This assumption may be true with mobile advertising
Advertising credibility is defined as consumers’ perceptions of the truthfulness and believability of advertising in general. Advertising credibility is one of perceptual dimensions underlying ad credibility which is the extent to which the consumer perceives claims made about the brand in the ad to be truthful and believable. Other dimensions is advertiser credibility and perceived ad claim discrepancy. (MacKenzie & Lulz, 1989)
2.5 Demographic of Mobile Users and Attitudes toward Mobile Advertising Based on the results of a study on Public Attitudes toward Advertising “More favorable than you might think” (Shavitt, Lowrey, & Haefner, 1998), attitudes are considered as a function of demographic segments. Specifically, demographical variables including gender, age, education and income may affect attitudes toward advertising
The purpose of this chapter is to point out the guidelines to conduct this research from collecting data to analyzing data. First of all, the research design will developed completely to guide the procedure of the research step by step. Next, specific methods used in each steps: designing questionnaire, sampling, collecting data and analyzing data will be introduced. Finally, procedure of pilot study will be also mentioned in this chapter.
3.1. Research Design
First, an exploratory research was conducted to seeking insights into the problems by doing secondary data research. Secondary data provided problems and hypothesis, supportive researches and information, additional problems that can be explored in the primary research. Next, an initial questionnaire was developed. Then pilot study was conducted to test the framework developed from secondary data research. This helps to transform the secondary data, theory into the practical situation and design the questionnaire which is most suitable to participants.
The survey research was conducted in March, 2013 by paper-based, web-based and mail survey. The questionnaires were provided on a variety of forums, social networks and different types of web sites: schools and universities’ Forums, Facebook, Zing Me, Forums about cellphones, technologies, women, etc, especially websites that support mobile platforms. Zalo – a mobile application was also a source of respondents. Moreover, in order to reach old respondents, paper-based questionnaires were published in public places like Public Park.
The questions covered a wide range of issues that approach a both descriptive research and causal research. These approaches were combined to narrow the factors underlying attitudes.
3.2. Research Model and Hypotheses
3.2.1. Research Model
Proposed Model of Attitudes toward Mobile Advertising
3.2.2. Research Hypotheses
Based on the relationship between underlying factors and attitudes toward mobile advertising described in theoretical framework, the following hypotheses were developed.
H1: Overall consumers have a positive attitude toward mobile advertising.
H2: Informativeness has a positive impact on attitudes toward mobile advertising. H3: Entertainment has a positive impact on attitudes toward mobile advertising H4: Irritation has a negative impact on attitudes toward mobile advertising. H5: Credibility has a positive impact on attitudes toward mobile advertising.
H6: Attitudes toward mobile advertising is related to age of consumers. H7: Attitudes toward mobile advertising is related to gender of consumers. H8: Attitudes toward mobile advertising is related to education of consumers. H9: Attitudes toward mobile advertising is related to income of consumers. 3.3Questionnaire Design
3.3.1 Questionnaire Design Process
Completed questionnaires were built through five stages: (1) planning what to measure, (2) formatting the questionnaires, (3) question wording, (4) sequencing and layout decisions and (5) pretesting and correcting problems.
3.3.2 Information Requirements
Based on research objectives, the following information requirements were listed to navigate the questionnaires. All of them were revised again after additional secondary data collection and exploratory research.
Information requirements of Questionnaire
Research Questions Information Requirements 1/ What are consumers’ attitudes
toward mobile advertising?
- General attitudes toward mobile advertising and specific attitudes toward the forms of mobile advertising.
2/ Which underlying factors of mobile advertising contribute to consumers’ attitudes?
- Overall evaluation of informativeness, entertainment, irritation and credibility in mobile advertising
- Reaction to the ideas and statements that contribute to attitude toward advertising.
3/ How do attitudes toward mobile advertising differ among
- Classification of demographical characteristics and related variables.
3.3.3 Format of Questions
Both open-response and closed-response are chosen in the pilot study. Open-questions were used in exploratory research and pretesting to modify the question for primary research. The questions in primary survey are closed-response. Some fields of question used probe – an open-response question to follow up a closed-response question to obtain additional question and reduce the limitation of incomplete choices.
3.3.4 Sequence and Layout
The questionnaire followed the basic guidelines for sequencing a questionnaire: from general to specific, from broad to narrow, from simple to complicated. Sensitive questions about private information will be placed in the end of the questionnaire. The first question is a simple question that if they own a tablet, smartphones or feature phones. Next, a question about purpose of using mobile devices was created to give participants an imagination about the times they uses cellphones or tablets. Then respondents were asked about types of advertisements that they have seen on mobile device before and draw an overall view of what survey focusing on. Main questions determining dependent variables and independent variables were coming up next. Finally, personal information of participants which is considered as the most sensitive part in questionnaires was requested. The below table will summarize the outline of questionnaire.
Layout and Scales of Questionnaire
Components Contents Scales
Screening Questions Types of owning mobile devices
Multiple Choice, Multiple-Response Scale (Checklist)
Main purposes of using mobile devices
Multiple Choices, Single-Response Scale Data: nominal
Types of advertisements
participants have seen
Multiple Choice, Multiple-Response Scale (Checklist)
Data: nominal Focused Questions Attitudes toward
Mobile Advertising Items
Likert Scale Data: Interval
Sensitive Question Gender Multiple Choices, Single-Response Scale Data: nominal
Ages Multiple Choices, Single-Response Scale Data: nominal
Education Multiple Choices, Single-Response Scale Data: nominal
Income Multiple Choices, Single-Response Scale Data: nominal
Target Population: Mobile Users in Vietnam who own at least one cellphone that can access internet or who have seen focusing mobile advertisements before.
Sampling frame: There is no sampling frame for this research.
Sampling Procedure: Because of a limited time and budget, self-selected web survey, a type of convenience sampling that is modified for internet-based survey is used in this research. This method use open invitations on portals such as Web sites, or (in some cases) dedicated “survey” sites (Couper, 2000).
Sample Size: Time and budget constraint is a critical point in deciding sample size. A sample of 200 respondents will be surveyed under the limited condition. This
number exceeded the requirement quantity of responds needed for factor analysis. 3.5 Data Collection Method
The target respondents of this research is mobile users’, especially who often interact with advertisement on mobile web sites and applications. That means they are quite familiar with internet resources like mail and web sites. Hence, the chosen data collection method in this research is web-based survey and mail survey. Moreover, these methods are an inexpensive, fast, simple and effective. Online survey sites like Google Form also support the required questions that reduce the probability of missing answers.
On the contrary, this data collection method has several disadvantages. A lack of interaction between respondents and interviewer causes some bad effects as the misunderstanding of questions or lack of knowledge about mobile advertising. In order to overcome these advantages, questionnaires need to be prepared carefully and pilot testing is necessary.
In addition, paper-based is used to reach inactive mobile users. They are the persons who use mobile to read news or collect information and do not interact with the above survey.
3.6 Data Analysis Techniques
The decision of data analysis techniques is depending on the types of question and the objectives of the research. In general, there are several steps in data analysis. First each question is analyzed by tabulating the data. Next, hypothesis is tested by various statistical techniques such as reliability test, factor analysis, chi-square test, correlation test and multiple linear regressions are the main technique.
3.6.1 Internal Consistency Test
Internal Consistency Test is one of reliability test. Cronbach’s alpha which a common measure of internal consistency was applied in this research. This test tries to calculate how much items support for variables. When interpreting the result of Cronbach’s alpha, the cronbach’s alpha equal or over 0.7 is considered as acceptable.
Items that increase the appreciable this value will be considered to remove.
3.6.2 Factor Analysis
Common factor analysis is a common procedure of factor analysis. The purpose of this technique is to investigate the underlying meaning of items. Factor analysis transforms lists of items into new variables that is not correlated and emphasize same meaning. The number of new variables is reduced as much as possible. To test factor analysis again, correlation test will be run after factor analysis. Generated components then are rotated by varimax rotation. To sum up, the number of components and items beyond component will be established.
3.6.3 Multiple Regression Analysis
One of the main objectives of this thesis is to determine the predictor factors of attitudes toward mobile advertising and multiple linear regressions is the chosen technique. Multiple Regressions – an extension of Simple Linear Regression allows predicting dependent variable which is attitudes toward mobile advertising based on independent variables came out from the previous techniques. It also determines the percentages of model fit and total variance explained. The measures of items are 5-likert scales and meet the requirement of scale for this analysis.
3.6.4 Chi-square test for Independence
From the sixth to the ninth hypotheses, relationship between attitudes toward mobile advertising and demographic variables are required to be checked. Chi-square test for independence is a good statistical technique that can recognize those relationships. The advantage of this method is being a nonparametric test and it does not require the normal distribution of variables. Besides that, Chi-square test provides Cross-tab table to observe an overview of associations between variables.
3.7 Data Quality
Reliability tests the overall consistent of measures and ensures that the research provides consistent results. There are three perspectives that need to be discussed on reliability: stability, equivalence and internal consistency.
A research that reaches the stability is the one supplying constant results of a same person toward the same measurements over time. Due to the lack of time and contact information, the questionnaires were impossible to ask for the answer of one respondent more than one times and the time between tests and retest more than the suggestion of two months. To reduce the errors of instability as much as possible, in the pilot study, respondents were requested to answer the questionnaire two times and the time delay is one week. Almost the questions were consistent with the previous one they have submitted.
Another perspective of reliability is equivalence. When being asked for answering two versions of questionnaire made by different investigators, variations between results should be not different far from the other one. To approach the equivalence of the measures, parallel forms were established and test simultaneously with test and retest of pilot study. The first version of question was paper-based questionnaire and the second was web-based questionnaire. As the results in stability testing, the errors introduced by different form are not trivial.
The last feature of reliability is internal consistency. Internal consistency indicates that items beyond the same variables emphasizes and support for that variable. The results of internal consistency test will be pointed out in the data analysis and result section.
The approach of validity is the measures can measure what the research study attempt to achieve. Validity can be examined in two perspectives: internal validity and external validity.
The external validity occurs when the research study can be generalized and assumed to be right for other cases on different people, times and places. This questionnaire is collected by using convenience method. Therefore, it is hard to assess the external validity and the generalization can be skipped.
Internal Validity is the ability of the measurement to exactly imply the relationship between causes and effects. Internal Validity includes three main forms: content validity, criterion-related validity and construct validity.
Content validity is satisfied when the measured items adequately represent the topic of the research. Because the items were built on research objectives, all of items try to reach the conclusion of research questions. Items were developed from the broad to narrow, from general to specific to ensure that all items were relevant and support for the topic of the study.
Criterion-related validity is the extent to which it is used efficiently for prediction and estimation. Correlation test among variables will be analyzed later to reflect the differences under various variables.
Construct Validity considers both conceptual framework and the practice of measurement. Items in measurement were derived from famed international journal articles and tested by professors in many countries. Thus, convergent validity existed. During data analysis, construct validity will be tested many times by using data analysis methods.
3.8 Pilot Study
Pilot study is very necessary step in developing questionnaires. It helps researcher to reduce errors and test whether the questions obtain the expected information supposed to be. Sample of pilot study is quite small, around twenty respondents. They are students of universities, officers and random respondents. The mutual requirements of participants are they pay much attention on questionnaire and have passionate to improve this questionnaire. The survey is responded by direct face to face between investigator
and respondents. Direct feedback of each detail in questionnaire is noted carefully by investigator.
Some problems were recognized after pilot study. These problems can be grouped into two groups: problems of specific questions and problems of questionnaire. In the term of specific questions, participants were suggested to explain the meaning of each questions to prove that their understanding were consistent with the meaning researcher observed. They gave opinions about their thinking about questions such as the question have implied answer, been sensitive, made them feel ambiguous or discomfort able. From these opinions, the questions were adjusted to be more understandable and clear. Other issues recognized through pilot study are problems of overall questionnaire. Fortunately, most participants agreed that the questionnaire was logical and easy to follow up. The length of questionnaire was reasonable and not made respondents discouraged. However, the format of questionnaire led to missing responses in some questions. Based on pilot study, several items were removed because they were not familiar with Vietnamese people. Few items were added according to the recommendation of participants.
DATA ANALYSIS AND FINDINGS
This chapter will involve analysis and findings of the data. All procedures of data analysis and results will be described specifically as well as the meaning of data; themes under the data will be observed carefully. Based on analyzed data, the end of this chapter come to a conclusion about whether hypotheses will be accepted or not.
4.1 Data Screening
There are 237 respondents participated in quantitative research. However, 8 questionnaires were not input into the data set because the respondents do not meet the requirements that owning at least one mobile device or have seen at least one type of ads on mobile websites or mobile applications. 23 of 229 remaining questionnaires were eliminated after filtering out the cases in spite of two reasons: missing values and outliers. In the case of missing values, 12 incomplete questionnaires which have more than one missing data were removed. With respect to outliers, outliers of every independent and dependent variables were pointed out through box plots by using descriptive statistics. By this way, eleven cases are excluded and a total of 206 cases are included.
Before analyzing data, several items are reverse coded because these questions are negative worded to positive attitudes. This method is applied for four questions about Irritation: IR1 – Offended, IR2 – Irritating, IR3 – Annoying, IR4 - Disturbing. Reverse coding is not mandatory in some methods but makes further analysis easier.
4.2 Description of the sample
Four demographic variables: gender, age, education and income were measured in this survey. The summary of these demographic characteristics is described in Figure 3. It can be seen clearly that most respondents are from 18 to 40 years old. Female respondents are far more than male respondents. The majority of respondents is studying college or graduated college and has income less than 10 million dong per month.
Demographic Characteristics of the Sample
Besides that, characteristics relating to the using of mobile devices are identified. Almost respondents own at least one mobile device. They spend most time using mobile devices for a variety of purposes but phone calling and answering which is the basic function of mobile phone seems to be the main reasons to use mobile devices. Other purposes like surfing webs and using mobile application also has a significant proportion of their time on mobile devices. In other words, it is meaningful to investigate attitudes toward advertising on websites and applications.
Characteristics Related to Mobile Devices of the Sample
Variables Frequency Valid Percent
Owning a smartphone No 57 27.7 Yes 148 72.3 Total 206 100.0 Owning a feature phone No 130 63.1 Yes 76 36.9 Total 206 100.0 Owning a Tablet No 130 63.1 Yes 76 36.9 Total 206 100.0
Not owning any mobile devices
No 204 99.0
Yes 2 1.0
Total 206 100.0
Main purpose of using mobile devices
Phone Calling 81 40.9 Text Messaging 38 19.2 Surfing Web 40 20.2 Using Apps 30 15.2 Playing Games 9 4.5 Total 198 100.0
4.3 Descriptive Statistics of Attitudes toward Mobile Advertising
Histogram of A1 – Positive Emotions
The neutral emotion is the score attained by most respondents than the others. There is a light tendency to negative emotions than positive emotion. While the number of respondent totally agree that mobile advertising brings them positive emotion is extremely small, the most negative emotion take an appreciable percentage and higher than the score of four.
Respondents seem to agree that they have received production information from advertisements they saw with the mean of 3.19. This statement does not make participants strongly disagree or strongly agree.
Figure 6: Histogram of A4 – Use ads’ information to make purchasing decisions Most of participants do neither agree nor disagree with the opinion that they use ads’ information to making purchasing decisions. The proof is the mean of 2.81. However, their opinions slightly lean on disagreement.
4.4 Internal Consistency Test (Reliability Test)
Internal Consistency Test is the first step in analyzing data. The aim of this test is to check the correlation among the items and reduce the items beyond variables if they are not reliable. Cronbach’s Alpha analysis is used to test internal consistence of five variables: Attitudes toward Mobile Advertising, Informativeness, Entertainment, Irritation and Credibility.
4.4.1 Attitudes toward Mobile Advertising
Four items underlying attitudes toward mobile advertising have Cronbach’s Alpha of 0.638. It is acceptable. However, Cronbach’s Alpha could be much greater (0.704) if the item “Mobile Advertising generates legitimate income for publishers and
developers”. From this reason, this item is removed. Table 6
Reliability Statistics of Attitudes toward Mobile Advertising Items (Cronbach’s Alpha = 0.638) Squared Multiple Correlation Cronbach's Alpha if Item Deleted A1 - Positive Emotions .341 .512 A2 - Legitimate Income .087 .704
A3 - Receive Product Information .242 .531
A4 - Use ads' information to making purchasing decisions
Although the Cronbach’s Alpha value of five items underlying Informativeness is quite high, it would be better if deleting the fifth item “Easy to Understand”. By this way, there are only five retained items measuring informativeness is (1) mobile advertising has various information, (2) mobile advertising has information related to mobile users’ demands, (3) mobile advertising provides up-to-date information and (4) mobile advertising useful for purchasing decision making.
Reliability Statistics for Informativeness (Cronbach’s Alpha = 0.723)
Squared Multiple Correlation
Cronbach's Alpha if Item Deleted
I1 - Various Information .371 .650
I3 - Up-to-date Information .298 .674
I4 - Useful for Decision Making .423 .643
I5 - Easy To Understand .164 .747
Reliability Statistics for Entertainment (Cronbach’s Alpha = 0.791) Squared Multiple Correlation Cronbach's Alpha if Item Deleted E1 - Entertaining .379 .759 E2 - Enjoyable .576 .695 E3 - Satisfied .523 .737 E4 - Funny .337 .764
In the matter of Entertainment, Cronbach’s Alpha value is greatly good (0.791). Moreover, removing any items will lead to a decrease in this value. Hence, all four items are accepted.
Reliability Statistics for Irritation (Cronbach’s Alpha = 0.856)
Squared Multiple Correlation
Cronbach's Alpha if Item Deleted
IR1 - Offended Reverse coded .381 .856
IR3 - Irritating Reverse Coded .678 .776
IR4 - Disturbing Reverse Coded .393 .851
As the above table showed that if any item is deleted, new Cronbach’s Alpha will equal or less than the current value (0.856). Thus, the scale of irritation has a high level of internal consistency with four items: offended, annoying, irritating, disturbing.
Similar with Irritation and Entertainment, five items in Credibility generates high Cronbach’s Alpha. Anyhow, the fifth item “Clear, not misled” has a stronger effect on Cronbach’s Alpha than others. For a better consistency, “Clear, not misled” will be eliminated from this dimension.
Reliability Statistics for Credibility (Cronbach’s Alpha = 0.844)
Squared Multiple Correlation
Cronbach's Alpha if Item Deleted
C1 - Providing Appropriated Evidences .492 .809
C2 – True .591 .787
C3 - Believable .475 .811
C4 - Products live up to promises of ads .424 .817
C5 - Clear, Not Misled .356 .838
After testing the internal consistency, a total of three items are deleted. The first one is “limitative income” in Attitudes toward mobile advertising. The second one is “easy to understand” in Informativeness. The last one is “Clear, not misled” in Credibility To conclude the reliability test, this following table will summarize items are retained after the reliability test and these items will be used for later analysis.
Retained Items after Internal Consistent Test
Attitudes toward Mobile Advertising
A1 - Positive Emotions
A3 - Receive Product Information
A4 - Use ads' information to making purchasing decisions Informativeness I1 - Various Information
I2 - Relevant Information I3 - Up-to-date Information I4 - Useful for Decision Making Entertainment E1 – Entertaining
E2 – Enjoyable E3 – Satisfied E4 – Funny
Irritation IR1 - Offended Reverse coded IR2 - Annoying Reverse coded IR3 - Irritating Reverse Coded IR4 - Disturbing Reverse Coded
Credibility C1 - Providing Appropriated Evidences C2 – True
C3 – Believable
C4 - Products live up to promises of ads C5 - Clear, Not Misled
4.5 Principle Axis Analysis (Common Factor Analysis)
Going along with reliability test, Factor Analysis is a statistical technique to reduce numerous variables to a smaller number of variables called factors – a managerial
set of variables. These factors can indicate implication of variables and characteristics that variables are overlapped. There are many procedures to estimate how many factors are appropriate and which items belong to each factor. Depending on the purpose of exploring the underlying construct of variables, common factor analysis (also called principal axis factoring) is chose. This method will be tested based on variance and find the smallest possible number of factors.
4.5.1 Appropriateness of Data
As shown in Correlation Matrix Table, there is no variable that does not have at least correlation equal or over the minimum requirement of 0.3. In other words, all variables are correlated with at least one other variable and no variable is considered to eliminate from factor analysis.
KMO and Bartlett’s Test
Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .881 1664.616 Approx. Chi-Square
Furthermore, the high Kaiser-Meyer-Olkin Measure of Sampling Adequacy (0.881) show that there is adequacy of sampling and the Barlett’s Test of Sphericity is statistical significant. The significant result is smaller than 0.05. They draw a conclusion that the use of factor analysis is suitable and the process could be kept going.
Measure of Sampling Adequacy for Individual Variables
Variable KMO Measure
I1 - Various Information 0.894
I3 - Up-to-date Information 0.861
I4 - Useful for Decision Making 0.912
E1 - Entertaining 0.837
E2 - Enjoyable 0.897
E3 - Satisfied 0.894
E4 – Funny 0.805
C1 - Providing Appropriated Evidence 0.920
C2 – True 0.886
C3 – Believable 0.900
C4 - Products live up to promises of ads 0.937
IR1R - Offended Reverse Coded 0.839
IR2R - Annoying Reverse coded 0.774
IR3R - Irritating Reverse coded 0.838
IR4R - Disturbing Reverse Coded 0.904
In KMO measure (measures of sampling adequacy) for individual variables; all variables have value excess the middling value of 0.7, so there is adequacy of sampling and no variable need to be removed until this step.
Table 14 Communalities Initial Extraction E1 – Entertaining .431 .437 E2 – Enjoyable .641 .653 E4 – Funny .403 .464
C1 - Providing Appropriate Evidences .638 .702
C2 – True .608 .668
C4 - Products live up to promises of ads .404 .378
Offended Reverse coded .439 .424
Annoying Reverse coded .728 .828
Irritating Reverse Coded .724 .792
DIsturbing Reverse Coded .457 .447
I1 - Various Information .537 .471
I2 - Relevant Information .481 .452
I3 - Up-to-date Information .309 .238
I4 - Useful for Decision Making .542 .488
E3 – Satisfied .620 .579
Extraction Method: Principal Axis Factoring.
One of noticeable fact displayed in Table 14: Communalities is the variable date information” has the lowest communality. This result claims that “up-to-date” item will meet troubles when considering belonging to any factor. The removal of this items should be thought over after evaluate the matrices.
4.5.2 Factor Structure
Total Variances Explained Table provides some critical insights into the factor problems. Referring to the eigenvalues, three factors are expected because they have eigenvalues more than 1 and 61.392 percent of the variance will be explained.
Total Variances Explained
Initial Eigenvalues Extraction Sums of Squared Loadings
Rotation Sums of Squared Loadings Total Cumulative % Total Cumulative % Total Cumulative % 1 6.580 41.128 6.142 38.390 3.506 21.915
2 2.042 53.893 1.652 48.713 2.709 38.849 3 1.200 61.392 .740 53.339 2.318 53.339 4 .901 67.023 5 .869 72.457 6 .674 76.668 7 .607 80.462 8 .529 83.771 9 .488 86.820 10 .440 89.571 11 .387 91.990 12 .351 94.182 13 .288 95.980 14 .264 97.627 15 .220 99.000 16 .160 100.000
Extraction Method: Principal Axis Factoring.
With regard to the number of factors, based on the initial eigenvalues and cumulative percent of variance explained, three factors are suggested.
The rotated Factor Matrix table will display how variables loaded into rotated factors.
Rotated Factor Matrix
Factor 1 Factor 2 Factor 3
C2 – True .789
C1 - Providing Appropriate Evidences .769
I1 - Various Information .575 .339
I4 - Useful for Decision Making .550 .392
C4 - Products live up to promises of ads .506 .302
I2 - Relevant Information .473 .425
I3 - Up-to-date Information .349 .341
IR2R - Annoying Reverse coded .889
IR3R - Irritating Reverse Coded .841
IR1R - Offended Reverse coded .636
IR4R - Disturbing Reverse Coded .622
E4 – Funny .671
E2 – Enjoyable .436 .628
E1 – Entertaining .619
E3 – Satisfied .453 .536
Extraction Method: Principal Axis Factoring. Rotation Method: Varimax with Kaiser Normalization.
a. Rotation converged in 5 iterations.
As showed in the rotated factor matrix, the factor structure has complex structure. There are seven variables loaded into two factors. Variables that have small differences between two factor loadings are removed. These are I1 – Various Information, I4 – Useful for Decision Making, C4 – Products live up to promises of ads, I2 – Relevant Information, I3 – Up-to-date Information and E3 - Satisfied. Although E2 – Enjoyable loads into both the first and third factor, factor loading in the third factor are quite high (0.628) and much greater than factor loadings in the first factor (0.436). So this variable is kept on.
To sum up, three factor revealed by Factor Analysis. The first factor consists of three variables: C1 – Providing Appropriate Evidences, C2 – True, C3 - Believable and named as Credibility. The second factor comprises four variables: IR1R – Offended,
IR2R – Annoying, IR3R – Irritating and IR4R – Disturbing. Three variables: E1 – Entertaining, E2 – Enjoyable and E4 – Funny are included in the third one.
To compare with the model applied in the questionnaire, three factors are consistent with three independent variables. Factor loadings of Credibility items are strong on the first factor, Irritation on the second factor and Entertainment on the last factor. There is a little change of items beyond each variable. E3 – Satisfied and I4 – Products live up to promises of ads do not satisfy the requirement of factor loadings and removed.
The following table lists items and variables after factor analysis and reliability test, warning that H2: Informativeness has a positive impact on attitudes toward mobile advertising is removed from analysis.
Modified Items and Variables
Variables Original Items Modified Items
Attitudes toward Mobile
- Positive Emotions
- Receive Product Information - Legitimate Income
- Use ads' information to making purchasing decisions
Positive Emotions Receive Information Use ads’ information to making purchasing decision
Informativeness - Various Information - Relevant Information - Up-to-date Information - Useful for Decision Making - Easy To Understand Entertainment Entertaining Enjoyable Satisfied Entertaining Enjoyable Funny
Irritation Offended Reverse coded Annoying Reverse coded Irritating Reverse Coded Disturbing Reverse Coded
Offended Reverse coded Annoying Reverse coded Irritating Reverse Coded Disturbing Reverse Coded Credibility Providing Appropriated
Evidences True Believable
Products live up to promises of ads
Clear, Not Misled
Providing Appropriated Evidences
4.5.3 Correlations among Variables
To test the relationship among variables, Pearson correlation is the most suitable method. This method will compute coefficients between each paired continuous variables. From these coefficients, the efficiency and significance of factor analysis will be checked and the results can be supported for multiple regression analysis
Pearson Correlations among Variables
Attitude Entertainment Credibility
Credibility .531** .488**
Irritation .307** .326** .429**
All the statistical significance of the correlation coefficient is below 0.0005. It means that the correlation coefficient is statistically significantly different from zero.
“Irritation” has the smallest positive correlation to Attitudes. Stronger correlations to Attitudes than Irritation is Entertainment and Credibility in order. In other word, Credibility has a strongest relationship with Attitudes, Entertainment has medium relationship an Irritation has weakest relationship.
Entertainment and Credibility somewhat relate to the others because their correlation coefficients to others are quite high and nearly 0.5. On the contrary, Irritation seems to have a low correlation with others (approximately 0.3).
4.6 Multiple Regression Analysis
4.6.1 Model Fit
First of all, how well the multiple regression models fit the data will be determined. Model fit is concluded by assessing multiple correlation coefficients (R), Total variation explained (R Squared) and Statistical significance.
Table 19 Model Summary Model R Square R Adjusted R
Square Std. Error of the Estimate Durbin-Watson 1 .563a .317 .307 .57878 1.519
a. Predictors: (Constant), TIrritationRC, TEntertainment, TCredibility b. Dependent Variable: TAttitudes
Shown in the Durbin-Watson statistic, the serial correlation of residuals is 1.519 and this number falls within the acceptance range from 1.5 to 2.5. It means that there is no auto correlation problem involved in the data.
Multiple correlation coefficient (R) can be presented as a measure of prediction quality. The higher multiple correlation coefficient is, the better the
independent variables affect dependent variable. Multiple correlation coefficient of 0.563 is not too high but more than the average. Some independent variables may not efficiently predict for dependent variable.
R Square indicates the total variation explained and also called coefficient of determination. When R square is high, percentage of variance of dependent variable explained by independent variables is large. This model just explains 31.7% variance of dependent variable. Adjusted R Square generates to reduce sample bias and the total variation explained decreased to 30.7% due to adjusting. According to classification of effect size developed by For Cohen, this statistic illustrates a small effect on dependent variable.
ANOVA for Multiple Regressions
Model Sum of Squares Df Mean Square F Sig. 1 Regression 31.114 3 10.371 30.961 .000b Residual 66.996 200 .335 Total 98.111 203
a. Dependent Variable: TAttitudes
b. Predictors: (Constant), TIrritationRC, TEntertainment, TCredibility
F-ratio is calculated by dividing the mean sum of squares for regression to the mean sum of square for the residual. The good fit of data is confirmed by F sig. smaller than 0.005. F(3,200) equals to 30.961.
In summary, Credibility, Irritation and Entertainment somewhat statistically significantly affect Attitudes toward Mobile Advertising.
4.6.2 Statistical Significance of Independent Variables
Coefficients for Multiple Regressions
Model Unstandardized Coefficients Standardized Coefficients t Sig. Collinearity Statistics B Std.
Error Beta Tolerance VIF
(Constant) 1.039 .200 5.205 .000
TEntertainment .211 .067 .212 3.169 .002 .762 1.312 TCredibility .397 .070 .402 5.668 .000 .680 1.470 TIrritationRC .056 .065 .056 .869 .386 .810 1.234
a. Dependent Variable: TAttitudes
Unstandardized Coefficients states how independent variables vary dependent variable changes. If Entertainment increases to one, Attitudes will increase 0.211. An increase of Credibility will lead to an increase of 0.397 in Attitudes. Attitudes seem to have least change when Irritation changes.
Irritation has the significant value of 0.386, excess the critical significant value of 0.05. Therefore, Irritation does not affect Attitudes. Attitudes are more affected by Credibility than Entertainment.
The hypotheses tested by multiple regressions analysis are:
H3: Entertainment has a positive impact on attitudes toward mobile advertising is accepted.
H4: Irritation has a negative impact on attitudes toward mobile advertising is not accepted.
H5: Credibility has a positive impact on attitudes toward mobile advertising is accepted. 4.7 Chi-square Test for Independence