Internet Privacy
A look into the construct of Privacy Knowledge
Master’s thesis within Business Administration
Author: Michael Nordström
Acknowledgements
We would like to thank the following people for their help and contribution to the com-pletion of this Thesis.
We thank Erik Hunter for his guidance and critique which he has provided during the development of this paper.
We thank Pensionärernas Riksorganisation Dalvik (PRO Dalvik) for helping us get in-to contact with elderly survey respondents.
We also thank all the participants who took the time to fill out our questionnaire. Finally we would like to thank our families and friends, for their support and under-standing in during a hectic period of our studies.
Thank you.
_____________________ _______________________
Master’s Thesis in Business Administration
Title: Internet Privacy: A look into the construct of Privacy Knowledge
Author: Michael Nordström
Sergej Sevcenko
Tutor: Erik Hunter
Date: 2012-05-14
Subject terms: Privacy, Privacy Concern, Privacy Knowledge, Social Media, Computer Knowledge, Internet Knowledge, Regulation Awareness
Abstract
Background: With the increasing use of personalized marketing and the in-creasing ability to collect information on consumers, the con-sumers’ concern of privacy is increasing. Therefore it is im-portant to understand what effects privacy concern, and how marketers can minimize this concern. Previous research suggest that factors such as computer knowledge, internet knowledge, and regulation awareness all affect privacy concern, however we believe that these are all related to each other in a construct we call Privacy Knowledge.
Purpose: To investigate the construct of Privacy Knowledge and to what degree it influences a consumer’s attitude towards informational privacy.
Method: In order to validate the Privacy Knowledge construct and meas-ure its relationship to Privacy Concern we employed a deductive methodology which was comprised of questionnaires. The ques-tionnaires were composed of summative Likert Scales, three of which had been previous validated by previous research. We uti-lized a quota sampling technique in order to gather enough data from each age group. The results were then analyzed by tools such as Factor Analysis, ANOVA tests, and Multiple Regression Analysis.
Conclusion: Through the Factor Analysis we found that the factors Internet Knowledge, Computer Knowledge, and Regulation Awareness were better organized as Basic IT Knowledge, Advanced IT Knowledge and Regulation Awareness. Privacy Knowledge was found to be positively related to Privacy Concern. However we could only conclude of the three factors which make up Privacy Knowledge, Basic IT Knowledge had an effect on Privacy Con-cern. We believe this is due to the exclusion of other factors af-fecting Privacy Concern such as situational factors and suggest conducting further research on the matter including these varia-bles.
Table of Contents
1
Introduction ... 1
1.1 Background ... 2 1.1.1 Personalization ... 2 1.1.2 Privacy ... 3 1.1.3 Social Networks ... 4 1.2 Problem Discussion ... 5 1.3 Purpose ... 6 1.4 Research Questions ... 62
Frame of Reference ... 7
2.1 Privacy Definition... 7 2.2 Disclosure ... 82.3 Internet Privacy Concern ... 8
2.3.1 Factors affecting Privacy Concern ... 9
2.3.2 Consequences of Privacy Concern ... 11
2.4 Privacy knowledge ... 12
2.4.1 Factors influencing privacy knowledge ... 12
2.4.2 Levels of privacy knowledge and privacy concerns ... 13
2.5 Summary of frame of references ... 14
3
Method ... 16
3.1 Research Approach ... 16 3.2 Research Design ... 16 3.2.1 Research Strategy ... 16 3.2.2 Research Choices ... 17 3.3 Questionnaire Design ... 18 3.4 Sampling ... 183.4.1 Defining target population ... 19
3.4.2 Sampling type, technique and size ... 19
3.5 Data Analysis ... 20 3.5.1 Descriptive statistics ... 20 3.5.2 Inferential statistics ... 21 3.6 Quality of Method ... 21 3.6.1 Reliability ... 21 3.6.2 Validity ... 22 3.6.3 Generalizability ... 22
4
Results ... 23
4.1 Demographics ... 23 4.2 Factor analysis ... 23 4.2.1 Reliability analysis ... 25 4.3 Description of answers ... 264.3.1 Privacy Concern Factor ... 26
4.3.2 Original Three Factors ... 27
4.3.3 Restructured Factors ... 28
4.4 Hypothesis Testing ... 30
4.5 Hypothesis Tests Summary ... 33
5.1 Privacy concern ... 34 5.2 Privacy knowledge ... 35 5.2.1 Research Question 1 ... 35 5.2.2 Research Question 2 ... 37 5.2.3 Research Question 3 ... 38
6
Conclusion ... 40
6.1 Critiques of the Study ... 41
6.2 Suggestions for Future Research ... 41
Appendix
Appendix A – Questionnaire ... 46
Appendix B – Description of answers ... 51
Appendix C – Additional Data Hypothesis 2 ... 53
Figures
Figure 1 – Privacy knowledge and privacy concerns interrelation ... 15Figure 2 – Sampling Design ... 18
Figure 3 – How many people use the internet among various ages (Findahl, 2011) ... 19
Figure 4 – Factor Scores ... 25
Figure 5 – Questionnaire Page 1 ... 46
Figure 6 – Questionnaire Page 2 ... 47
Figure 7 – Questionnaire Page 3 ... 48
Figure 8 – Questionnaire Page 4 ... 49
Figure 9 – Questionnaire Page 5 ... 50
Tables
Table 1 – Respondents’ demographics... 23Table 2 – Privacy Knowledge Rotated Component Matrix ... 24
Table 3 – Cronbach alpha values for all the scales ... 26
Table 4 – Privacy Concern Welsh Test & Brown-Forsythe Test ... 26
Table 5 – Privacy Concern Independent Sample t-test ... 26
Table 6 – Computer Knowledge Post Hoc Test ... 27
Table 7 – Computer Knowledge Independent Sample t-test ... 27
Table 8 – Internet Knowledge Post Hoc Test ... 28
Table 9 – Internet Knowledge Independent Sample t-test ... 28
Table 10 – Basic IT Knowledge Post Hoc Test ... 28
Table 11 – Basic IT Knowledge Independent Sample t-test ... 29
Table 12 – Advanced IT Knowledge Post Hoc Test ... 29
Table 13 – Advanced IT Knowledge Independent Sample t-test ... 29
Table 14 – Regulation Awareness Post Hoc Test ... 30
Table 15 – Privacy knowledge descriptive statistics ... 30
Table 16 – t-test of privacy knowledge factors ... 30
Table 17 – Correlation of age and education level variables ... 31
Table 18 – Cross tabulation of educational level and age group variables ... 31
Table 19 – Liner regression of privacy knowledge towards privacy concern ... 32
Table 20 – Hierarchical Regression Model ... 33
Table 21 – Hypothesis tests summary ... 33
Table 22 – Computer knowledge factor: test of homogeneity of variances ... 51
Table 23 – Computer knowledge factor: Robust Tests of Equality of Means ... 51
Table 24 – Internet knowledge factor: test of homogeneity of variances ... 51
Table 25 – Internet knowledge factor: One-way ANOVA... 51
Table 26 – Basic IT knowledge: test of homogeneity of variances ... 51
Table 27 – Basic IT knowledge: One-way ANOVA ... 51
Table 28 – Advanced IT knowledge: test of homogeneity of variances ... 51
Table 29 – Advanced IT knowledge: One-way ANOVA ... 52
Table 30 – Regulations’ knowledge: test of homogeneity of variances ... 52
Table 31 – Regulations’ knowledge: One-way ANOVA ... 52
Table 32 – Regulations’ knowledge: independent samples t-test ... 52
Table 33 – ANOVA Test Privacy Knowledge and Education ... 53
Table 34 – Independent Sample T-test (Education) ... 53
Table 35 – Independent Sample T-test (Education 2) ... 53
1
Introduction
In February of 2012 a man barged into Target, a discount retailer focusing on a large va-riety of products, and urged that the manager of the store explain why they had sent his daughter printed advertisement and coupons for baby related items.
“She's still in high school, and you're sending her coupons for baby clothes and cribs? Are you trying to encourage her to get pregnant” (Golgowski, 2012).
The manager confused by the situation apologized profusely and said that he would get to the bottom of the problem. The manager proceeded to investigate what had happened. What he found was that the analytics department of the Target store chain had been ana-lyzing their consumers purchasing habits. The reason for this was in order to create a more personalized advertisement for its customers (Golgowski, 2012). In the case of this teenage daughter, the department had come to the conclusion that she was pregnant and therefore had sent her printed advertisements and coupons which would benefit a consumer in this situation (Golgowski, 2012). The manager called back to the upset fa-ther, and explained the situation, however what he heard next was very surprising. It seems that the father had been informed by his daughter that she in fact was pregnant. Both the store manager and father were baffled by the fact that a group of analysts could have figured this out without ever meeting the daughter (Golgowski, 2012).
This is a very recent example of marketers using consumer information to create very specific and efficient marketing communication to its consumers. However it also brings up another important area, and that is Privacy. This news article was very inter-esting and evoked a sense of technological marvels, however at the same time a sense of eeriness was evoked. How much do these marketers know about me? What do they do with this information? How are they getting this information? These are just some of the questions consumers might be asking themselves after hearing of such a story.
In times where the amount of consumer information being collected is only growing, privacy is at the forefront of the consumers mind, and therefore is a very important topic to understand. With this research paper we will be looking at consumer’s attitude to-wards privacy. More specifically we will be looking at internet privacy, and what fac-tors may influence a consumer’s privacy concern. This research paper will be looking at privacy knowledge factors, and how a consumers understanding over privacy policies, regulations, and mechanisms affect said consumer’s concern with his or her privacy.
This section will introduce the context in which the study is being performed. Background information on the area of privacy and personalization will be dis-cussed, as well as the purpose of the research paper. Furthermore we will discuss
1.1
Background
In order to better understand this phenomenon, we will discuss three related topics. Per-sonalization, which refers to the specification of marketing content in order to be more appealing to consumers. Privacy, the research which has been performed in this area and its various segments. Social Networks, here we will be introducing the recent area of privacy within social networks, and why this is such an important area.
1.1.1 Personalization
Consumer buying behavior and information search patterns have always been in the center of marketers interests (Bhatnagar & Ghose, 2003). In order to effectively market products and services, companies need to know the target consumer’s demographical and socio-economic personal information (Christiansen, 2011; Keh, 2007, Alatalo & Siponen, 2001). In terms of the growing advertising industry and as a result the increas-ing informational noise (Solomon, Bamossy, Askegaard & Hogg, 2010), the necessity of more relevant information about consumer interests is becoming apparent. The rea-son for this is that a consumer is more likely to pay attention to a piece of marketing communication which appeals directly at them and their interests (Solomon et al., 2010; Culan & Armstrong, 1999). Furthermore, in the eyes of consumers, usefulness of the website increases, if there is a high level of personalization (Lee & Cranage, 2010). Personalized marketing began from the address and telephone books for newsletters and personal selling via telephone. In the 80s – 90s it was developed to computerized data-bases and moved towards online advertising and social media in the recent 15 years (Nowak & Phelps, 1995). Modern technologies give companies the possibility to track, store and analyze consumers’ information at a very low cost (Norberg, Horne & Horne, 2007). Using consumers’ personal information and browsing history, companies are able to deliver them offers appealing directly to their personal interests, mood, plans or desires at the current moment. As a result, technologies allow companies not only to lower the cost of B2C and B2B communication, but increase its’ efficiency and viabil-ity. Highly personalized advertisements and other online services increase the possibili-ties of persuading the consumer to behave in a desired way (Moon, 2000).
In the most common way, advertisers create a number of different offers, based on con-sumers groups. Advertisers identify these targeted groups by defining the most im-portant characteristics. Examples of this could be demographic information, search keywords, and websites they visit. All that information is collected by cookies that are stored on the consumers’ computers (Schedwin, 2008). However, some of the technolo-gies are able to deliver personalized advertising and without using cookies. What is more important is the fact that information can be collected on consumers even without their knowledge of this activity (Christiansen, 2011).
The most well-known examples of consumer personalized information tracking tech-nology developers and service providers are Microsoft, Google and Facebook (Sched-win, 2008; Tucker, 2011). Information is used to deliver personalized advertisements based on online behavior history (Google and Yahoo services), interests and social ac-tivity (Facebook and MySpace). Personalized advertising services became a significant source of income for these service providers, for example, Google Adwords and Ad-sense personalized advertisement services are one of the main sources of income
(Google Annual report, 2011). Social networks have been also increasingly adopting this form of advertisement as a form of income (Kannan, Chang & Whinston, 1998). However, utilization of such technologies has become a source of public debate (Schedwin, 2008; Culan, 1993). While society understands the benefits which can be brought from personalization, such as increased convenience, filtering advertisement noise, and an increased willingness by the consumer to purchase the advertised product, it also fears it. This is because of the concerns that are brought up regarding privacy on the internet, and possible threats of such information collection and possession. (An-drade et al., 2002)
1.1.2 Privacy
The area of privacy, in terms of consumer information collection, became known to the world in the early 1990’s. The reason for this was that it had become more and more apparent to the consumers that companies were selling information which they had col-lected on them (Culnan, 1993). This created an outcry from the public, where most con-sumers feared for their personal information, and furthermore dented the reputation of these companies (Culnan, 1993).
This shows the importance of taking the consumers privacy into perspective when deal-ing with business decisions. Durdeal-ing the previous decade the internet has become a very powerful tool which businesses use, and as was discussed above, the use of consumer information has only increased. Due to this, privacy has yet again become a big issue and has had a wave of academics performing new research into this area.
Much research has been done when it comes to the consumer’s attitude to privacy, in other terms, their concern over privacy. Fogel and Nehmad (2009) found that three fac-tors which affect a person’s attitude towards privacy are their risk orientation, their level of trust, and the person’s belief in their ability to control their own information. At the same time Bellman, Johnson, Kobrin, and Lohse (2010) found that the differences in various individuals’ privacy attitudes could have to do with difference in their cultural value, the varying internet experience between users, and the national regulations where the user lives. This study was focusing on comparing the differences between interna-tional consumers, and did not focus on a homogenous group of people.
Furthermore it has been found that there are different classes of users when it comes to privacy attitudes. Sheehan (2000) argues that there are three different levels of internet users in terms of privacy. There are fundamentalists, who are very averse to the divulg-ing of private information. There are the unconcerned, who do not think much of their own privacy. And finally you have the pragmatists, those who are willing to divulge private information, but only if there is a benefit for doing so. It was found from Sheehan (2000) that roughly 25% of the United States’ population were fundamental-ists. Furthermore these individuals tended to be of higher education. 50% of the United States’ population were pragmatists, and the last 25% of the United States were uncon-cerned about their privacy.
More recently there has also been a lot of investigation into how one might minimize the concern a consumer has towards their privacy. It has been seen that if informational practices are put in place, then the overall concern over one’s own privacy diminishes (Culnan & Armstrong, 1999). In fact, it was shown that if a consumer is made aware that fairness procedures have been put in place, then only previous experience from that
consumer would affect their decision to join the proposed program or not. This differed from the case when this information was not divulged, at which time both privacy con-cerns and previous experiences came into play (Culnan & Armstrong, 1999).
Moreover it was shown that as privacy assurance was put in place, the consumers per-ceived benefits, reper-ceived from divulging their private information, increased (Lee & Cranage, 2010). This led to a study which showed that companies who are introducing privacy policies and privacy assurance can take a competitive advantage from this area (Lee, 2011) and that as consumers become more concerned with their privacy, the reali-zation will most likely be that all firms will have to adopt some form of privacy policy (Lee, 2011).
1.1.3 Social Networks
At the end of 2011, there were 12 Internet social networks with more than 100 million active users each. The most popular, Facebook, has reached 845 million, Qzone has 536 million and Twitter with 380 million users. The extent to which people are using social networks is the cause for many ethical issues that face both marketers and technology developers. With the development of social networks, a number of studies have been made in that area (Tucker, 2011; Tufekci, 2008; Pitta & Fowler, 2005, Norberg, 2007). Researchers have noticed a few important aspects of social network users’ privacy relat-ed behavior. It was found that users tend to disclose their sensitive personal information (telephone number, address), even without asking them to (Fogel & Nehmad, 2008; Tufekci, 2008). When researchers were trying to understand the roots of such irrespon-sible behavior, they founded that a “privacy paradox” is taking place. People claim that they are concerned about their privacy and have negative attitudes toward obtaining their private information via the Internet, but behave in the opposite way (Norberg, Horne & Horne, 2007).
It was also found that social network users do not behave in the same way. Men are more likely to disclose contact data and women are more likely to disclose their inter-ests (Zheleva & Getoor, 2009). In terms of race, Anglo people more often use their real names, indicate relationship status and interests, while Afro-Americans more often indi-cate their religion. It is also a tendency that the older a user is, the more concerned about privacy they are. Another factor is education. People with IT and engineering back-grounds are more concerned about privacy and more often use privacy settings (Tufekci, 2008). Users subconsciously evaluate trust and risk in regards of various ac-tions online (Suh & Han, 2003).
In terms of social networks, personalized advertising has become a key feature in re-gards to advertisement. Tucker (2011) demonstrated that personalized advertisement, where consumers’ institution and favorite celebrity was mentioned on Facebook, was twice more effective than not personalized.
In return to consumers’ concerns towards their online privacy, Facebook and MySpace introduced more convenient and easy-to-use privacy settings. Consumers try to manage their privacy settings and limit sharing of their private information, but they don’t take into account all the possible threats (Tufekci, 2008). For example, researchers have demonstrated that even with improved security settings it is possible to access users’ private information. Through the social groups, users can lose significant amount of
in-formation because of the groups’ settings. This makes it possible that inin-formation is shared with unknown people (Zheleva, Getoor, 2009).
The source of threat for privacy in social media is not only in conscious publishing of private information. All the information that consumers leave in social media remains saved on the service providers’ servers and can be accessed. Communication online is not as the live communication. Potentially, everything can be founded using search en-gines (Tufekci, 2008). As a result, Christiansen (2011) found cases when such infor-mation was used against the users. Examples are thefts and employers shadowing em-ployees.
With the increasing number of people going online and joining social networks and the willingness of businesses to obtain their private information for personalized advertis-ing, the problem of privacy in the online social world becomes more and more im-portant.
1.2
Problem Discussion
Consumers’ privacy attitude and privacy behavior influence business in a many ways. Growing consumer concerns have already resulted in more strict regulations of obtain-ing, storage and using of personal information on the Internet (Schedwin, 2008). If con-sumers will continue to be persistent in the protection of their rights for privacy, it can increase advertising costs for companies.
Researchers found that different users of social networks have different attitude towards privacy online and behave in a different way. Several different factors, influencing both attitudes and behavior, were investigated by authors – gender, race, education, regula-tions, level of trust and risk perception (Fogel & Nehmad, 2009; Bellman, Johnson, Ko-brin & Lohse, 2010 ; Zheleva & Getoor, 2009; Tufekci, 2008; Suh & Han, 2003). Background research shows that computer and the Internet proficiency, educational ar-ea, and regulation awareness are reoccurring phenomenon. This makes sense as all these factors have something to do with the potential knowledge the consumers may have had in regards to internet privacy. It is important to note that while figures about these fac-tors effects have been obtained, no research takes into account all three of these facfac-tors when looking at privacy concern. As we believe that these three factors are related, that is to say they all increase or decrease an individual’s understanding of online privacy, we feel it is important to understand how these factors together effect privacy concern. We feel that this is an important area of research since there is clear academic back-ground showing that a consumer’s understanding of information collection and privacy regulations affect their privacy concern, however no one has summarized these factors into a single concept such as privacy knowledge.
This research can also be beneficial for businesses. When debating whether or not there is a need to implement tools in order to safeguard the consumers’ privacy, it is benefi-cial to understand how the consumers will react to certain privacy changes. For exam-ple, a company whom is collecting information on its consumers that are generally more knowledgeable in terms of computers might need to take more precautions when col-lecting data on their consumers since they might have a stronger attitude towards priva-cy and be more concerned.
1.3
Purpose
To investigate the construct of privacy knowledge and to what degree it influences a consumer’s attitude towards informational privacy.
1.4
Research Questions
In order to fulfill our purpose we have derived the following research questions which will guide our study.
As mentioned in the introduction we feel that Privacy Knowledge is a construct which is comprised of Internet Knowledge, Computer Knowledge, and Regulation Awareness. However it is important to know if all these factors are important, and which factors matter the most in the construct. To fill this knowledge gap we propose the first re-search question:
Q1. Which factors compile the construct of privacy knowledge?
We also have seen that each factor in itself has had an effect on privacy concern, how-ever it is important to see how the construct as a whole affects privacy concern. There-fore we have proposed the second research question:
Q2. How does privacy knowledge affect the privacy concern of an individual?
Lastly, we realize that privacy knowledge may act differently in different demographic groupings, and furthermore it is important to see what the overall level of privacy knowledge is at this point in time, therefore we have proposed the third research ques-tion:
Q3. What is the overall level of privacy knowledge within the consumer’s demographic groups?
In order to answer these research questions we shall be utilizing a quantitative question-naire for our empirical data collection. This questionquestion-naire will include a predefined scale which measures an individual’s privacy concern, developed by Buchanan, Paine, Joinson and Reips (2006). The questionnaire will also include our own scale which measure privacy knowledge, and lastly questions relating to demographical data. This data will then be used in order to map how privacy knowledge affects the privacy con-cern of individuals. We will also be taking a look at if these results are consistent through various demographic groups.
In order to map our results, we will be utilizing regression analysis, with the individuals score on the privacy concern scale as the dependent variable, and the privacy knowledge score and demographical information as independent variables. This will statistically show how privacy knowledge affects privacy concern.
2
Frame of Reference
An overview of previous research in the areas of Privacy, Privacy Concern, and theories related to Privacy Knowledge, helps us to understand the roots of privacy concerns and interconnections between privacy knowledge and privacy concern. After defining a concept of privacy knowledge and its influence on privacy concerns of the Internet us-ers, research is followed by deeper investigations of factors, influencing privacy knowledge. Theoretical understanding of a concept led us to building hypotheses, which will help to answer the research questions and empirically test theoretical model:
Q1. Which factors compile the construct of privacy knowledge?
Q2. How does privacy knowledge affect the privacy concern of an individual?
Q3. What is the overall level of privacy knowledge within the consumer’s demographic groups?
2.1
Privacy Definition
As a concept, privacy was first seen in the late 19th century when Warren and Brandeis (1890) wrote an article in the Harvard Law Review in order to introduce a definition of privacy and argued the need to protect this privacy. This effort was sparked by the in-troduction of the photographs and the ability to take snapshots of people and include them in newspapers or other public materials. Warren and Brandeis argued that every individual has the right to be left alone, and if he or she does not want to be included within a public forum, then that person should have the right to remove themselves from the public setting (Warren & Brandeis, 1890).
However, this is only one kind of privacy which the public feels necessary. The more critical and important perspective of privacy which this paper will be dealing with is that of seclusion and information ownership. Westin introduced a key definition in this sense, with saying:
“Privacy is the claim of individuals, groups, or institutions to determine for themselves, when, how, and to what extent information about them is communicated to others” – (Westin, 1967, Cited by Nowak & Phelps, 1995, p 49)
This definition of privacy constructed an outline for the Privacy Act of 1974, a law within the United States of America, and furthermore introduced a framework of under-standing which would form future research within privacy (Nowak & Phelps, 1995). The next evolution of privacy was with Stone and Stone’s definition of privacy, which stated that privacy was the ability to control the release or subsequent dissemination of information about him or herself, the ability to regulate the amount and nature of social interaction, and the ability to exclude or isolate himself or herself from unwanted audi-tory or visual stimuli (Nowak & Phelps, 1995). This definition introduced various situa-tions where privacy would matter, and furthermore introduced the fact privacy not only had to do with information.
This section of the research paper will introduce the theoretical framework that will be used within the analysis. Here we go into more depth on the topics of
Prosser suggested that privacy should take into account four different legal torts which it affects. These are intrusion, disclosure, false light, and appropriation (Prosser Cited by Nowak & Phelps, 1995). Intrusion deals with someone invading an individual’s per-sonal space without permission. Disclosure deals with publicly releasing information about an individual without that individual’s approval. False light deals with creating a false depiction of someone in a public setting, and appropriation deals with using some-one’s identity without their approval (Nowak & Phelps, 1995).
From this last definition we believe that a comprehensive picture of privacy is formed. With ever increasing exposure of personal information online, today individuals have to deal with persons and companies invading their personal space, publishing and selling of private information and misinformation in virtual space. All these issues hinder the ability of people to control their private information. Therefore we shall adopt the Defi-nition composed by Prosser (Cited by Nowak & Phelps, 1995).
2.2
Disclosure
Self-Disclosure is the act of communicating personal information to another entity (Moon, 2000). The reason that this area of research is important to look at is that to some extent the information which has been collected from the Internet, has in one form or another been self-disclosed. In social media, the essential part of the modern Internet, all the published information relates to self-disclosure phenomena.
Self-Disclosure has several interesting aspects, for example intimate disclosures are said to convey the emotions, attitudes and feelings of an individual. These disclosures are said to cause a sense of extreme personal vulnerability (Moon, 2000). This could be a large source of privacy concern, since what an individual believes is being disclosed to a close friend or acquaintance is then being collected by a marketing agency. However as we shall discuss later on, this feeling of vulnerability could be diminished by a higher knowledge within privacy aspects of the websites and social networks.
Another aspect of self-disclosure is the principle of reciprocity. This principle says that an individual is much more likely to disclose personal information if the party which is receiving said information will be returning the favor, i.e. giving some personal infor-mation back to the sender (Moon, 2000). This principle explains, why people tend to share their private information to unknown websites and within social networks.
2.3
Internet Privacy Concern
With the topics of privacy and disclosure now discussed, we can form an understanding as to what privacy concern is. As seen previously when defining privacy, certain aspects pertained to that of information privacy, and furthermore what eventual repercussions could be possible if ones privacy was not respected, i.e. public humiliation, and or harm to one’s self-identity. From this we can gain an understanding that most individuals are afraid of their privacy, and any information which can be used to negatively affect them. This is the area of privacy concern, an individual’s concern over their privacy and thereby their interests in how to best protect their privacy.
Much research has been done on privacy concern, and one of the most frequently reoc-curring definitions of internet privacy concern is an individual’s concern for being able to control the acquisition and thereafter use of information which he or she has put on the internet (Culnan & Armstrong, 1999; Culnan & Milberg, 1999; Goodwin, 1991;
Hoffman, Novak & Peralta, 2006; Phelps et al., 2000). That is to say an individual is concerned with the ability to control who can obtain their information, and how it gets used.
Internet privacy concern has many various contexts in which it can be found. The re-search area which this paper focuses on is that of information collection in terms of di-rect marketing. With this context in mind there are certain aspects of privacy concern which must be taken into account. The biggest of these aspects is that direct marketing is offering the consumer a benefit in exchange for their privacy information (Wirtz, Lwin & Williams, 2007). That is to say, while the consumer is giving up information about him or herself, in return for that the consumer is receiving marketing communica-tion which is personalized towards him or her.
However, since we are also looking at a context of the internet and online privacy, other aspects of privacy concern become more important. One such aspect is the fact that cer-tain technological tools, such as cookies, can track a user’s usage of the internet, what websites he or she has visited, what he or she has talked about and so on (Wirtz et al., 2007). Furthermore there is the aspect that information becomes more freely available, companies can now gather information not only those who shop or register on their websites, but also individuals who are just browsing (Caudill & Murphy, 2000). This af-fects level of privacy concern, since companies which an individual has no affiliation with are gaining access to information about them.
Privacy concern can therefore be seen as an important aspect for consumers which are using the internet, and further on within this research paper which shall discuss the con-sequences of privacy concern within consumers, and how that can affect companies. However, before we move on to factors which affect an individual’s privacy concern, we shall discuss the methods for measuring privacy concern.
There are numerous scales which have been developed for the measurement of privacy concern, i.e. privacy attitude (Buchanan et al., 2006; Malhotra, Kim & Agarwal, 2004). The researchers in this area have looked into the measurement of privacy concern, and from here determined that this measurement is a one-dimensional measurement (Bu-chanan et al., 2006). Bu(Bu-chanan et al (2006) tried to identify more dimensions within this scale, such as the various types of privacy which were earlier discussed, however their research concluded that privacy concern was a one-dimensional measurement. However this does not mean that there are no multiple factors that may affect the dimension of privacy concern, some of which may have profound effects.
2.3.1 Factors affecting Privacy Concern
In regards to factors which affect a consumer’s level of privacy concern much research has been done and several factors have been found. Castañeda and Montoro (2007) or-ganized these various factors into four groups. Customer-intrinsic characteristics; Per-ceptions, beliefs and attitudes of customer towards the control mechanism; Website re-lated variables; and situational variables.
Customer-intrinsic Characteristics
These variables as the title suggests have to do with internal characteristics of the con-sumer. Examples of such characteristics are socio-demographic factors, or internet ex-perience (Castañeda & Montoro, 2007). In terms of socio-demographic factors the
fol-lowing has been found. Phelps et al (2000) found that the only socio-demographic fac-tor which had any effect on the level of privacy concern was that of education. From their research they found that those with the highest level of privacy concern were those which had attended a vocational school. Moreover they found that that the lowest priva-cy concerns in terms of educational segments was that of college graduates. This indi-cates that education has a negative relationship to privacy concern. Furthermore Milne (1999) found that women were more likely to be concerned about privacy than men, and moreover that the older generation was more concerned than the younger generation. Internet experience was also included as a factor under customer-intrinsic characteris-tics. In these regards it was found that the more internet experience a consumer had the less likely that consumer was to shop online due to a lack of trust in security and the will to have his or her information remain private (Hoffman, Novak & Peralta, 1999).
Perceptions, Believes, and Attitudes of Customers Towards the Control Mecha-nisms
This category of factors deals with the consumers various beliefs about how their in-formation is going to be handled. Castañeda and Montoro (2007) identified one factor within this category to be the attitudes and perceptions towards direct marketing. Cul-nan (1993) found that those which had a positive attitude towards direct marketing, more specifically those who were positive to receiving directing marketing within the mail, were less concerned with their privacy. This result was reinforced by Milne and Boza (1999), Phelps et al (2000), Phelps, D’souza & Nowak (2001).
Another factor which has been identified is that of trust towards marketing. Culnan & Armstrong (1999) found that if a consumer was made aware that the marketing team in question would use their information in a fair manner which would maintain their sense of privacy, then the privacy concern of that individual would become significantly low-er when thinking about disclosing information. Furthlow-ermore Miyazaki and Flow-ernandez (2000) found that if more information was given on the marketing procedure, and how the marketing team would protect privacy, privacy concern decreased.
This factor greatly overlaps with that of control mechanisms. While the above research focused on the consumers trust for a marketing team, the following factor takes a look at the control mechanisms which are in place during data collection and use. In regards to control mechanisms it was found that privacy concern and the desire for more control mechanisms had a positive correlation (Phelps et al., 2000; Phelps et al., 2001; Milne & Boza, 1999). Furthermore it was found that consumers with high privacy concern de-sired more information about how the information that was being collected was going to be used (Phelps et al., 2000). This helps explain Milne and Boza’s (1999) finding that the consumers who had been informed about privacy policies had a lower privacy con-cern.
Website Related Variables
This category of factors deals with variables which are controlled by websites. One of the most important factors in this category, which was briefly discussed in the previous category, is that of privacy policies. It has been seen that as a website increases the availability and understandability of privacy policies on their webpages, the privacy concern of the user’s decreases (Culnan & Armstrong, 1999; Miyazaki & Fernandez,
2000). Furthermore it has been seen that as the privacy policies go into deeper detail, adding more information, the privacy concern also decreases (Andrade et al., 2002).
Situational Variables
The final category of factors which affect privacy concern is that of situational varia-bles. These factors have to do with factors that are relevant only to the exact exchange of information at one period of time. An example of this is the type of information which is being asked for. It has been seen that as more sensitive information is being asked for, or collected, the concerns of the consumer will increase (Andrade et al., 2002; Phelps et al., 2000). This is understandable since when more sensitive information is be-ing exchanged, the risks for the consumer, identified in section 2.1, increase.
Another situational variable which has been found is that of rewards. If a consumer sees a higher reward when deliberating over informational exchange, then the consumer will be more likely to have a lower concern over privacy (Andrade et al., 2002; Culnan & Armstrong, 1999; Culnan & Milberg, 1999; Phelps et al., 2000).
While these four categories identified by Castañeda and Montoro (2007) include a large amount of research within factors that affect privacy concern, there are still some factors which have not been included, however we still feel them to be of importance. Two of these factors which we believe to be very important are culture and governmental regu-lation. Bellman et al., 2010 found that differences in consumers’ privacy concerns could be explained by their cultural differences, and the level of governmental intervention which was present within their country. More specifically, Bellman et al. (2010) found that consumers which from countries which have a consolidated regulation system in place for informational privacy had a higher level of privacy concern.
2.3.2 Consequences of Privacy Concern
While we have seen what can affect privacy concern, what is also important to under-stand is what effects privacy concern can have on information collection and further-more the organizations and companies which are collecting information. Castañeda and Montoro (2007) identified two areas of consequences for not taking privacy concern in-to consideration. These are direct response behaviors and indirect response behaviors. Direct response behaviors deal with the consumer’s reaction to having a heightened pri-vacy concern and how to deal with this (Castañeda & Montoro, 2007). It has been seen that this can take several different forms. One of which is that the consumer will try to erase their information from websites or databases in which they don’t believe their pri-vacy is being protected (Milne & Rohm, 2000). Another response behavior is that con-sumers will actively try to avoid the direct marketing which has become a result of the information collection. This can be done using various programs, or by the consumer contacting their internet service provider (Sheekan & Hoy, 1999).
Indirect response behaviors on the other hand deal with the consumer’s relationship to-wards the companies or organizations which are poorly keeping up with privacy con-cerns. The most reoccurring observations in this area are that consumers will be less likely to register for websites without privacy policies, and furthermore websites with looser privacy controls will start to lose customers (Castañeda & Montoro, 2007).
While all response behaviors are important for a marketer to take into account, it is the indirect response behaviors which have had most focus from academics (Castañeda & Montoro, 2007; Culnan & Armstrong, 1999; Lohse, Bellman & Johnson, 2000; Sheekan & Hoy, 1999).
2.4
Privacy knowledge
Privacy concern, recognized by different scientists as a widespread phenomenon, has its social and psychological roots. Balabanis & Reynolds (2001) found that information consumers’ possess at certain moments, influence how they use different websites. In-formation in that context is the initial knowledge of the consumers about the website or company which created it. Initial knowledge influences consumers’ willingness to pro-vide personal information for the website’s owner. Eastlick et al (2006) have created privacy concerns model, which describes the sources and consequences of privacy con-cern on the Internet.
Dommeyer & Gross (2003) claim that consumers are lacking information on business practices related to privacy, private information collections and use. This leads to incor-rect evaluation of risks. Dommeyer & Gross (2003), Dinev & Hurt (2003) demonstrated that customers knowing how private information is collected and can be protected, per-ceive more control over personal information and have less privacy concerns. Summa-rizing researchers’ findings, it is suggested that initial privacy related knowledge, is the main source of privacy concerns and motive of consumers’ actions on the website.
2.4.1 Factors influencing privacy knowledge
In order to construct the Privacy Knowledge model we needed to find items which could be included in this model. In our investigation we found evidence that the follow-ing three factors belong to Privacy Knowledge.
Computer knowledge
Consumers have different sources of information related to the privacy on the internet. Potosky in a number of studies (1998, 2003, 2007) have found that initial consumers’ computer knowledge have strong relation with their future development in that area. The more consumers know about, the more information on those issues they will try to obtain in the future, and the more efficient will be their learning (Potosky, 2003). By “computer knowledge” Potorsky understand a range of attributes developed in the earli-er research by Potosky & Bebko (1998). In the research authors have created a scale for a measurement of the computer understanding and experience. It measure user’s knowledge about using computer, the extent to which they use computers, and the level of perception of how they good in using computer.
Internet knowledge
The internet experience and knowledge was not measured in the initial scale of comput-er knowledge. In furthcomput-er research Potosky (2007) introduce a tcomput-erm of “iKnow” which define the level of the internet knowledge. According to the author, the internet knowledge has to be measured separately from computer knowledge, and is defined by computer knowledge, frequency of the internet use, extent to which people use the
in-ternet for information search and email, the inin-ternet self-efficacy beliefs and individual difference variables (age and gender). Wei & Zhang (2008) in general call these factors “Internet experience”.
Work and education
The sources of mentioned computer and internet knowledge are not only internet expe-rience. Work and education have strong relation privacy concerns (Wang & Petrison, 1993; Sheehan, 2011). The source of such connection is the knowledge which consumer obtains during education period and work. If consumer’s current or future work is con-nected with the internet, it will be more interested in all the policies regulating it. The more consumers’ work will or is involved in the internet technologies and issues related to obtaining private information, the more consumer will be interested in obtained such information (Stoel & Lee 2003)..
Socio-demographic factors
According to Potosky (2007), the internet knowledge is strongly related to the consum-ers’ age and gender. In general, men’s internet knowledge is significantly greater then women’s. Findings of Potosky (2007) can be connected with the Fogel & Nehmad (2008), Tufekci (2008) findings that men are less concerned about privacy, than women and explain Dommeyer & Gross (2003) statement that “the more consumers know about the privacy, the less concerned they are”. Potosky (2007) found that consumers’ age has influence on the internet knowledge when young and old people are compared, but differences between other age groups are not significant. Internet knowledge has strong correlation with current occupation, and all the age groups having higher IT in-volvement on their work, have greater internet knowledge, compared to the same age group consumers.
Governmental regulations
People obtain information from various sources. Their knowledge increase, when some issues become an object of public discussion. As a result, government privacy regula-tions and society sensitivity to privacy issues are additional factors influencing privacy knowledge (Bellman et al., 2004). As it was mentioned earlier, internet users in general are more sensitive to everything that is related to political and social issues, and privacy on the Internet is one of such issues.
It can be noticed, that by their nature factors of privacy knowledge and factors of priva-cy concerns are similar. Factors forming certain level of knowledge lead to a certain level of concerns. Concerns lead to a further development of the consumer in the area, and changing over time factors lead to a new level of privacy knowledge.
2.4.2 Levels of privacy knowledge and privacy concerns
Considering all the mentioned theories about privacy knowledge and privacy concerns and also taking into account that people live in a changing environment and continuous-ly developing, it can be supposed, that privacy knowledge of consumers is changing with his or her life stages. Based on the privacy concerns research and sources of
priva-cy knowledge, certain interconnections of the level of privapriva-cy knowledge and level of privacy concerns are found:
1. Absence or little privacy knowledge leads to a very low level of privacy con-cerns (Sheehan, 2011; Phelps et al, 2000);
2. Increased privacy knowledge leads to a high level of privacy concerns (Sheehan, 2011; Phelps et al, 2000);
3. Extensive privacy knowledge leads to a decrease in privacy concerns (Phelps et al, 2000).
At the first stage consumers, having no computer or the internet background, not strong-ly involved in IT technologies on their job or during studies, not interested in IT tech-nologies, have no knowledge about collection of the private information on the Internet methods and its purpose. As a result, they have no privacy concerns. Once they realize that private information is an object of the commercial interest, and in some situations they are not able to notice private information leak, privacy concerns start to rise. After became concerned, consumers start to search for privacy related information and learn how to protect and control his or her private information. As a result of additional edu-cation and rise of privacy knowledge, their privacy concerns decrease.
Those steps of privacy concerns’ and privacy knowledge development are different by their length in different consumer groups. People obtaining education or/and working in a sphere of IT are going fast through all the steps and finally are less concerned about the privacy.
2.5
Summary of frame of references
In the frame of reference we have demonstrated that privacy knowledge defines level of privacy concerns of the internet user and there are a number of factors continuously in-fluencing privacy knowledge of the user. This approach as a model is demonstrated in Figure 1.
In order to answer the first research question, “What is the overall level of privacy knowledge within the consumer’s demographic groups?”, we suggest that:
Hypothesis 1 – The overall privacy knowledge of the respondents is high
Hypothesis 2 – Individuals in different areas of study will have differing levels of
pri-vacy knowledge
Hypothesis 3 – Individuals in different age groups will have different levels of Privacy
Knowledge.
Privacy Knowledge is formed by a number of factors. All of them not only form current level of privacy knowledge, but influence privacy concerns of the consumer. In order to understand which factors are most prevalent within privacy knowledge and have further influence on consumer, we suggest that:
Hypothesis 4 – Privacy Knowledge can be measured by three separate factors. Internet
The Theoretical mechanism of how privacy knowledge affects the privacy concern of an individual was discussed in the section 2.4.2. We saw that privacy concerns have been noted to increase as privacy knowledge components increase, however after a certain point the increased privacy knowledge will lead to a decrease in privacy concern, there-fore we propose the following hypothesis.
Hypothesis 5 – Privacy Knowledge has a positive relationship with privacy concern
un-til a point where the relationship becomes negative.
The tools for Privacy knowledge measurement and the collection of empirical data in order to test research model and hypotheses are demonstrated in methodology section.
Privacy Concern
Situational
Factors Attitude Factors
Demographic Factors Website Factors
Privacy Knowledge
Socio demographics Internet Knowledge Education ComputerKnowledge Knowledge Regulation
Privacy Concern:
Privacy concern is the over-all subject of which we are investigating.
Factors affecting Privacy Concern:
These are the factors which have been found to affect privacy concern.
Privacy Knowledge:
We believe that privacy knowledge is yet another factor which needs to be in-vestigated
Factors Affecting Privacy Knowledge:
From our research, these are factors we believe should affect the amount of privacy knowledge a person has.
3
Method
3.1
Research Approach
The first question which must be tackled in terms of this papers method is that of the re-search approach. While there are many variations on different rere-search approaches, in-duction and dein-duction are the two most established approaches (Saunders, Lewis & Thornhill, 2009). Deduction is most effectively described as when theory precedes the data, and conversely induction, is when data precedes the theory. In other words, the deduction approach to research is when one finds theories, formulates hypotheses, and then proceeds to test these hypotheses with empirical data which has been collected (Saunders et al., 2009). Induction, on the other hand, is when one collects his or her em-pirical data and from this develops new theory or builds upon existing theory (Saunders et al., 2009). Regarding the purpose of this research paper, the deductive approached seemed most suitable. The reason for this is that the theory around privacy concern has already been established, however the authors are taking existing theories and testing a new variation of dependent variables upon this.
The reason that we did not opt for an inductive research approach is because for the purpose of this research paper such an approach would not have been suitable. We are aiming to measure the effect of one variable upon another. Had we instead wanted to re-search the reasons for why privacy knowledge may affect privacy concern, and gained deeper insight into the minds of the consumer, an inductive approach may have been more suitable (Saunders et al. 2009).
The next area of research approach which must be addressed is that of the purpose. The three most prominent classifications of research purposes are exploratory, descriptive, and explanatory purposes (Saunders et al., 2009). In terms of the purpose identified in the introduction of this research paper, we believe that the most fitting classification would be that of explanatory. The reasoning behind this choice is because our research purpose is looking to explain a causal relationship, i.e. the effect privacy knowledge has on privacy concern, which is one of the cornerstones of explanatory purposes (Saunders et al., 2009). This classification of a research purpose can also be known as casual re-search (Malhotra & Birks, 2007).
3.2
Research Design
This section will describe the research design which has been utilized during the inves-tigation. The section is structured in the following manner, firstly, we will introduce the research strategy which we employed, the research choices which accompanied this choice, and finally discuss the area of time horizons.
3.2.1 Research Strategy
When considering the research strategy, we decided to collect quantitative data instead of qualitative data. Quantitative data is defined as data which “seeks to quantify data and typically, apply some form of statistical analysis” (Malhotra & Birks, 2007). We
This section discusses the methodology which was utilized in the research. The sec-tion is split up into the research approach, the research design, The Quessec-tionnaire design, Sampling technique, Data analysis, and finally the quality of the method.
feel that this is the most suitable data collection technique for the needs of this research paper’s purpose. Since we are going to test hypotheses which were created out of col-lected theory, collecting quantitative data will help us to statistically test these hypothe-ses (Saunders et al. 2009).
Furthermore, considering that the aim of this research paper is to identify the relation-ship between privacy knowledge and privacy concern, and not investigate the consum-ers’ reasoning behind these concepts, qualitative data seemed unsuitable since its prima-ry use is for gaining greater insight and understanding (Malhotra & Birks, 2007). The research strategy which we chose was that of surveys and moreover questionnaires. The reasoning behind this decision is that since we are working from a deductive per-spective we wanted to remain as separate from our respondents as possible. It has been noted that surveys are the research strategy which is most associated to the deductive approach (Saunders et al., 2009). Furthermore with a questionnaire we are able to opera-tionalize the research with structured questions, in such a way that we can systematical-ly measure the concerns and privacy knowledge of our respondents on Likert scales, thereby introducing a form of analysis which we can use in order to measure the casual relationship between the two (Malhotra & Birks, 2007; Saunders et al., 2009).
3.2.2 Research Choices
Now that we have defined that we will be using surveys in order to collect our primary data, there are a few research choices which must be taken into consideration. First of all is if we will be utilizing a multi-method or mono-method data collection technique. Multi-method techniques can either be used to gather data from multiple sources, or to mix quantitative and qualitative data. Mono-method on the other hand is when data is collected from a single source (Saunders et al., 2009). Due to the time and budget con-straints which the authors have during this investigation, we feel it is most suitable to perform a mono-method data collection technique. While this may constrict the research in terms of capturing all information, we believe that if we utilized a multi-method data collection technique, our data would suffer as the collection required would manifest poor quality due to time. It is for this reason we chose the mono-method approach. Furthermore there is the aspect of the time horizon. Here one can choose between per-forming a cross-sectional study, or a longitudinal study. A cross-sectional study will give the researchers a snapshot of the current situation, describing what is currently go-ing on. This time horizon entails that the empirical data is collected at a certain point in time and the research thereby describes the phenomenon at that specific point in time. Longitudinal studies on the other hand perform empirical data collection at several points in time. This is done in order to study the phenomenon over a period of time. This can be useful if one is trying to identify trends (Saunders et al., 2009).
For the purpose of this research paper, a cross-sectional study is most suitable. Again one of the larger factors which play into this is that of time constraints. We do not be-lieve it is realistic for us to perform a longitudinal study with the time limits which have been set upon us. However, at the same time, we believe that the study would not exact-ly benefit from a longitudinal approach, as the purpose of this research is to identify a relationship between two concepts. A cross-sectional approach is more than capable of letting us do this.
3.3
Questionnaire Design
The Questionnaire can be found in Appendix B. The design of this questionnaire is based off of three scales which we have found in academic literature. The first meas-urement scale which we use is that developed by Buchanan et al. (2006). This scale was developed to measure privacy concern of respondents. It consists of 14 statements which the respondents must rate on a Likert scale. For this scale a 5 point Likert scale was used, and ranged from 1, noted as not at all, to 5, noted as very much. As is dis-cussed in depth in Appendix B we decided to change some terminology of this scale due to its outdated nature. The specific terms which we changed were that of email. Ques-tions 10 through 12 of the scale used the term e-mail where we believe the authors were suggesting communication over the internet. While the main form of communication may have been email at the time of the original study which developed the scale, we be-lieve that in today’s society email is a much more formal less used form of online com-munication, and furthermore it excludes all other forms of online communication. For this reason we decided to change the term email in question 10 through 12 to “online message”.
The second two scales, namely the iKnow scale developed by Potosky (2007) and the computer knowledge scale developed by Potosky and Bobko (1998) were used in order to derive a privacy knowledge measurement within the questionnaire. As we saw in the theoretical framework, the three variables which we identified as privacy knowledge variables were that of internet knowledge, computer knowledge, and regulation knowledge. We could not find a previously developed scale for the regulation knowledge and therefore created five Likert scale questions for this variable.
3.4
Sampling
In the survey we are using the sampling design offered by Cooper & Schindler (2006), for business research process (Figure 2).
3.4.1 Defining target population
Since the purpose of our research is to investigate to what degree privacy knowledge in-fluences a consumer’s attitude towards informational privacy on the internet, surveys have to concentrate on Swedish users of the internet.
According to European Travel commission report (2011), in Sweden 92,9% of the pop-ulation are using the internet and users from the country spend in average 24 h. per month online. There is minimal difference between genders, with 93% of women and 95% of men using the internet. Percentage of users using the internet differs depending on the age (Figure 3).
Figure 3 – How many people use the internet among various ages (Findahl, 2011)
Malhotra & Bricks (2006) define target population as objects that possess necessary for researcher information. As it is seen from statistics, all the main age groups in Sweden are using the internet. Therefore, the target groups of current research are people living in Sweden and belonging to the following age groups:
18 – 24 age people, studying or still depend on parents.
25 – 64 age people, majority are working and living independently.
65 and higher - the majority of these people are retired.
Statistics Sweden (2012) provides information that on 2011 December 31st in Sweden were registered 7 795 572 people in mentioned age groups, 50.48% of them were wom-en. However, due to limited time and financial resources we have to limit our target population by Jönköping inhabitants. On 2011 December 31st in Jönköping city were registered 101 963 people in mentioned age groups, 50.52% of them were women This information is necessary to test representativeness of the sample exploring different de-mographic groups.
3.4.2 Sampling type, technique and size
In scientific research two fundamental sampling techniques are used – probability and non-probability sampling (Malhotra, Bricks, 2006; Fink, 2003). A non-probability sam-pling technique is chosen for the survey, because for probability samsam-pling access to tar-get population personal details would be necessary, which is impossible because of pri-vacy policies and budget of the research (Saunders et al., 2009).
Quota sampling will be used in order to collect necessary data from all the population subgroups at low cost and time (Saunders et al., 2009). Population is divided on men-tioned earlier specific age groups and information is collected from them separately.
There are 14 569 representatives of 18-24 age group people, 63 870 in 25-64 age group, and 23 524 in 65 and higher age group.
Sample size depends on size of target population and quota variables (Cooper & Schindler, 2006). However, a number of researchers have developed tables for deter-mining minimal sample size (Bartlett et al, 2001; Krejcie, Morgan, 1970). According Bartlett et al (2001) and Krejcie, Morgan (1970) tables, for 14 569 people group we need a sample of 375 respondents, for 63 870 people group 381 respondents‘ sample is necessary, and in the research of 23 524 people, 377 respondents have to participate. According to Saunders et al. (2009) 30 respondents is a minimum number of objects for every single demographic group in the research for conducting statistical analysis. In that case sampling distribution for the mean of every sample will be very close to a normal distribution. As a result, optimal quotes for the entire groups are 375, 381 and 377 respondents. However, research will still be statistically relevant with 30 respond-ents in every group. It is a real number of respondrespond-ents we are able to reach in given short period of time and budget.
18-24 age group will be reached in Jönköping University, 25-64 age people will be reached on their workplace, randomly visiting various companies in the city, 65 and el-der will be reached in retired people organization PRO (Pensionärernas riksorganisa-tion). In order to handle response rate problem and fasten data collection, questionnaires were distributed personally and respondents had possibility to ask researcher questions. Since respondents were selected based on convenience, selection bias can take place. However, even in that situation quota sampling obtains similar to probability sampling results (Malhotra, Birks, 2006).
3.5
Data Analysis
In order to analyze the results of the questionnaire we will use descriptive and inferen-tial statistical methods. Descriptive statistics are used to present general sample results, tendencies and measures. Inferential statistics’ methods will allow going beyond availa-ble data and making predictions about population behavior (Fink, 2003).
3.5.1 Descriptive statistics
Analysis of ordinal and nominal scales using descriptive statistics is necessary in order to understand the central tendency and dispersion of the results. Mean will be used to define a central tendency of nominal scales’ questions and standard deviation is used to measure the extent to which data values differ from the mean. The reason of using the mean is necessity of further data analysis with parametric statistics tools. Such approach is widely used in studies of privacy concerns and knowledge (Potosky, 2006; Potosky, Bobko, 1998; Phelps et al. 2000, Dinev, Hart, 2003). However, some scales used in the survey differ in their magnitudes and in order to compare different variables coefficient of variation has to be used (Saunders et al., 2009).
We will also be using contingency tables and cross-tabulation descriptive statistics techniques in order to demonstrate specific values and compare the results. In the graphs differences in answers of respondents can be shown in most convenient and easy-understandable way.
