Master Thesis

(1)

Master Thesis

A perspective on the integration of contact

tracing apps during the COVID-19

pandemic

Master of Supply Chain Management

University of Groningen, Faculty of Economics and Business

Student: Jakob Lustig Student number: S3825612

25.01.2021

Supervisor: dr. E.I. (Esther) Metting

(2)

List of figures

Figure 1. App use and benefit for society ... 14

List of tables

Table 1. Summary of theories and concepts (Venkatesh et al., 2003) ... 18

Table 2. UTAUT (Venkatesh et al., 2003) ... 19

Table 3. HBM (Hochbaum, 1958) ... 20

Table 4. Interview cases of residents ... 23

Table 5. Interview cases of professionals ... 23

Table 6. Example of coding ... 25

Table 7. Summary of the interviews with residents ... 40

(5)

Abstract

Purpose: During the times of the COVID-19 pandemic the use of technologies play a vital role. Especially, contact tracing app could have a big impact on containing the disease. In prevalent literature researcher used both the unified theory of use and acceptance of technology or the health belief model to investigate the use of technology during the pandemic. However, researcher suggest that one of these models alone is not enough to examine such a technology. Therefore, this research aims to combine the two concepts to explore the society’s perception of the contact tracing app.

Methodology: The setting of the research is the Dutch and the German network and consists of an exploratory case study. Both deductive and inductive approaches aim to explore how people perceive the app and the illness.

Findings: The findings propose that, especially the privacy is a major concern for people and can act as a barrier to use the app. Additionally, some Dutch residents mostly do not know how the technology of the app actually works, and therefore perceive these aspects as a barrier as well. What is more, some interviewee did not see the positive correlation between the use of the app and the containment of the disease.

(6)

1 Introduction

The outbreak of COVID – 19 has affected lives in most countries around the globe. Many people suffer from the virus that causes pulmonary problems. Lockdowns, restrictions and regulations impede the life of millions. Restrictions and regulations aim to reduce social interaction and thus, new infections. While keeping the infection rate down, due to those restrictions other problems arise. On the one hand, psychological issues from feeling trapped at home affect the health of the people. On the other hand, the economy suffers because of the restrictions as well. As a result of lockdowns, restaurants may not be able to stay open and therefore they could go bankrupt. In order to further keep new infections in control, beside many measures, there is a push to use the modern technology available (WHO, 2019).

All sorts of technology are used to help cope with COVID-19 in different sector (Whitelaw et al., 2020a). Those can be simpler technologies that have been around for years, like video chatting or shared drives. The technologies can be more advanced as well, like surveillance or contact tracing (Ting et al., 2020; Zeng et al., 2020).

To help fight the spread of COVID – 19 many countries are making use of contact tracing. This can be done special agencies or by using ICT systems, like a contact-tracing app. This app will send you a message if you had contact with a person that is confirmed to be infected with the corona virus. The technology of those apps is often similar across different apps. The system is either centralized or decentralized. This determines whether the data is stored in a central storage system or on the individual users’ device. To check the contacts with other users, in Europe, in favor of privacy, Bluetooth is the prevalent technology (Kretzschmar et al., 2020).

While the contact tracing apps bear many opportunities for controlling the spread of the pandemic, there are also issues with the technology like for example:

• The privacy of the users

• The accuracy of the technology

• People are not using the app (properly)

(7)

(2003) formulated the unified theory of use and acceptance of technology (UTAUT). As this technology aims to assist the fight against a deadly disease, people’s perception of the illness is also important. If people do not take the illness seriously, they might not take steps against it. Therefore, the health belief model (HBM) determines people’s views on diseases and actions against it (Becker, 1974).

Those two concepts were used to investigate the outlook of the public on the integration of the app. Based on this stems the following research question:

How do residents perceive the integration of contact tracing apps?

In order to answer this research questions an exploratory case study research design was chosen (Barratt et al., 2011). The setting of the research is the Dutch and German contact tracing app. The German Corona Warn App is a bit ahead of the Dutch Corona Melder as it was launched earlier. As a result, the Dutch authorities might learn something from the German integration. Semi-structured interviews were conducted both with residents of each country and professionals in this field.

As an outcome of this research, the practical implications may contribute to the understanding of what can be changed in order to successfully implement the app. Theoretical implications are formed around the two existing models and how they could be extended or combined.

(8)

2 Theoretical background

The theoretical framework forms the basis of the research. In the following sections the outbreak of the coronavirus builds up to the conceptual frameworks.

2.1 Pandemic and COVID-19

The term pandemic describes a wide spread of a human disease. To be classified as a pandemic it must spread over continents or even worldwide. Compared to conventional endemic disease the number of infections is unstable, which makes it more dangerous. Humans were affected by pandemics across history. The. most famous and probably deadliest pandemic was the plague in the 14th_{century, that killed over 75 million people (Echenberg, 2011). Currently two} diseases can be classified as pandemics, HIV/Aids and COVID – 19. Especially COVID – 19 has made a big impact on the day to day lives of people.

The corona virus first started in Wuhan, which is a city in China in the province of Hubei. First cases occurred at the end of the year of 2019. Since then, the world has changed significantly. The disease can cause respiratory problems, which can also lead to death. The virus fatally affects mostly the elder generation or people with pre-existing illness. What is more, the virus is highly contagious. Thus, it is important to minimize the risk of infection. There are various ways of getting infected (Robert Koch Institut, 2021). For example:

• Private households • School/university • Care homes • Work

• Public transportation

• Indoor sports and/or team sports • Restaurants

• Shops

• Supermarkets

(9)

essential stay open most of the time. This includes supermarkets for example. However, in many countries people have to wear masks, wherever they might get in contact with other people. Additionally, the government instructs residents to wash and disinfect their hands regularly.

Once people were infected, they have to get into quarantine to prevent infecting other people. Yet, especially dangerous about the disease is that a high number of people that get infected do not show any symptoms (Lango, 2020). If they do not get tested, they are not aware of their infection which means they cannot self-isolate and are therefore able to infect more people. In order to work against the uncontrollable spread of the virus modern technology is envisaged to be used at a bigger scale. The easiest manner to use modern IT is the share useful information.

2.2 The use of Information and Communication

Technology

The following sections are directed towards the technologies. It examines ICT systems and its integration and leads to the different uses of ICT systems during the pandemic.

2.2.1 Information and Communication Technology

Sharing information and communication with each other has been a vital part of society ever since. The term ICT emerged around 1980 for the first time. As the name suggests information and communication technology describes the assistance of technology when sharing information. The technology can be simply applications such as telephone use. However, as technology advances ICT systems change as well. With the commercialization of the internet new ways of ICT usage emerged (Panel, 2002). The world wide web enabled people to communicate with one another and share and store information. Interaction, especially via video, further rose with the beginning of the web 2.0 (Zakaria et al., 2010). The implementation of cloud computing and artificial intelligence (AI) are two additional factors why ICT systems advanced over time.

(10)

Whether an ICT system succeeds depends on different factors. Trust is an important one (Cai et al., 2010). If the consumer fails to trust in the circumstances of the system, it might not be used as much as the initiators want it to be. In this case trust in the technology and trust in the government is of high importance. Trust in the government ascribes the concerns people have whether their data is used for beneficial purposes only. Especially during this time there are many corona deniers, that say this disease is not that harmful and therefore they do not trust the government with their restrictions. There are also other influences that determine if the contract tracing app gets used, like the accessibility and usability of the app (Eligi & Mwantimwa, 2017).

2.2.2 Implementation and integration

The integration and the implementation of new technology is part of the change of an organization. In this case the integrating organization is the country and more specifically the ministry of health. First of all there needs to be a reason and a need for the change (Cawsey et al., 2011). Using regular contact tracing without the app the time between an official infection and the notification of all of his or her contacts is relatively long. In the meantime, there is a possibility that other people get infected as well. What is more, using conventional contact tracing not all people can get contacted as the infected people might not have any information about the others. For example, if you are on the train you will most likely have no information about the person sitting next to you. Additionally, with rising numbers of new infections conventional contact tracing agencies come close to their capacities and the time between infection and notification of contacts gets even longer (Malmberg & Britton, 2020).

When opting for the integration of a change multiple aspects have to be taken into consideration (Damschoder et al., 2011). One of the aspects deals with the complexity of the problem. In this case the complexity is to get people to accept the app.

(11)

2.2.3 Usage of technology during the pandemic

With all the lockdowns and people working or studying from home, video chatting has become a state-of-the-art tool when it comes to communication. Instead of physical meetings many companies hold their meetings online. Students from elementary school up to university were obliged to stay at home during the lockdowns. Therefore, classes were held online as well (Wiederhold, 2020). Contactless payment also aims to reduce physical contact. Additionally, in China for example robots are used for delivery purposes (Zeng et al., 2020).

It is a long-term trend that the use of technology in the health care sector becomes more and more important. The use of technology can help with processes like facilitating care, organizing care, surveilling, planning or testing (Ting et al., 2020).

Whitelaw et al. (2020) distinguish between five different types that include the direct health care process supported by technology during the corona crisis. Those types are:

• Planning • Tracking

• Screening for infections • Quarantine and self-isolation • Clinical Management

• Contact tracing

Making use of data helps keeping track of the spread of the virus and plan ahead. Acquiring and visualizing the data is an effective tool to manage quarantine measures for example. The data provides clear insight of rising or falling infection rates. Additionally, data determines geographical differences, so different measures can be applied on regions with more infections (Whitelaw et al., 2020b).

(12)

Screening for infections use artificial intelligence technology for example. Tools like digital thermometers or thermal cameras are utilized to check whether people are sick and might be infected. Based on this data people can get advised to test themselves or to go into quarantine.

This leads to the type which is directed at quarantine and self-isolation solutions. In some countries people are monitored, if they really stick to the quarantine measures. In some cases, regulatory offices physically check whether a person is actually following the rules. However, there are also ways of using modern technologies. Again, artificial intelligence assists cameras, GPS, apps or response codes. In China for example drones control if the residents stayed at home.

For clinical management technologies like AI help diagnosing the patients (Ting et al., 2020). What is more, it enables remote care so again physical contact can be avoided. As this disease might have a fatal outcome, the risk can be predicted of how severe the infection is and therefore, help with planning of hospital beds. The amount of hospital beds for patients with a severe course of the disease is an important factor, when it comes to dealing with the virus (Whitelaw et al., 2020b).

The last aspect is contact tracing. Depending on the country different technologies aim to track contacts with positive cases. The next chapter further discusses the contact tracing app.

2.3 Contact tracing

This section aims to investigate the likes of conventional contact tracing and the contact tracing apps.

2.3.1 Conventional contact tracing

The aim of contact tracing is to determine possible further infections, once a person got infected by the virus. The essence is to track down a person’s contacts and inform them that they were in contact with a positive person and might got infected as well.

There are two ways of how contact tracing assists fighting the disease. The first one is forward tracing, which detects people that might were infected as well. The second way is tracing backwards. This aims to identify the source of the infection (Endo et al., 2020).

(13)

cluster refers to the amount of infected people coming from one source of infection (Hellewell et al., 2020).

Several countries make use of technology as well by implementing a contact tracing app. In most cases the app assists the conventional contact tracing agencies. With conventional contact tracing the time between the positive test and alerting all the possible contacts is rather long. The app aims to reduce this time to keep further infection low (Ahmed et al., 2020).

2.3.2 Contact tracing app

The technology of the apps is different for many countries. The tracing part mostly works either with GPS or Bluetooth technology. Due to privacy regulations, European countries use Bluetooth. Other countries outside of Europe like China for example use GPS. The Bluetooth approach connects the phones once they get in close proximity to each other. The phones send codes which they exchange. Whenever a person gets tested positive, he or she should indicate it via the app and the app sends a message to the phones it exchanged codes with. Hereby, there are differences between the apps and the risk of an infection (Bay et al., 2020; Hatke et al., 2020). Generally speaking, only direct contact within 1.5 meters of about fifteen minutes is considered to be of higher risk. Some apps also indicate a mild risk of infection to the user, but not all do. It sometimes coheres with the government’s approach of each country as to what the health regulations are (Martinez-Martin et al., 2020).

There is a distinction between centralized and decentralized operating models. Regarding the centralized approach each phone has an ID on the server and the server provides pseudonyms for the phone. The pseudonyms are the codes that are exchanged with the other phones. Using the decentralized approach, the phones generate the pseudonyms themselves. Most countries in Europe use a decentralized approach as it fits more to the privacy regulations (Michael & Abbas, 2020).

(14)

Figure 1. App use and benefit for society

Figure 1 indicates that the user, who is not using the app, does not contribute to fighting the disease in that regard. On the contrary people that are using the app contribute to national health.

The impact of the app can be ascribed by the aspects of efficacy, effectiveness and efficiency. Efficacy describes whether the app works. The efficacy is studied by many researchers. These studies are mostly based on estimations and not on actual data. This is due to the recency of this topic and the privacy regulations that hinder extracting data from the app. Additionally, the apps are not mandatory to use and some are still in a testing phase. Notable about the studies is that most researchers agree that a high number of users is required in order for the app to significantly reduce the number of infections (above 60%) if contact tracing is the only measure taken (Braithwaite et al., 2020). Using contact tracing alongside measures like social distancing, lockdown, et cetera an usage between 10 and 15 % makes already an impact (Kretzschmar et al., 2020).

The effectiveness of contact tracing apps explains the fact that the apps also work in the real world. For example, the app classifies as effective, if there is an impact on national health. As previously mentioned, this can be achieved if enough people actually use the app.

(15)

The efficiency is not included in this research. There is not enough data on efficiency and it would make the research too broad (Marley, 2000).

2.3.3 Different approaches of the app

Many countries in Europe use the decentralized approach, but there are exemptions as well. For example, the UK used a more centralized and also more private approach. Compared to Germany for example the effectiveness seems to be lower as Germany had fewer infected cases per 100.000 residents compared to the UK, although the lockdown in Germany was not as though. Of course this cannot be attributed to the contact tracing app alone, but it definitely plays a role in it (Reintjes, 2020). The result was that the UK dropped their own app and switched approaches (Wise, 2020).

With privacy regulations that are not that strict outside of Europe, different approaches can be used. One example is South Korea. Based on the number of infections, the implementation of their app-based contact tracing is generally seen as very successful. An outbreak of a disease in 2015 led to gathering data from private companies about their customers. Therefore, when the corona virus broke out in South Korea a lot of data was already gathered. What made the contact tracing app even more effective is the fact that South Korea is a highly technological country. Nine out of ten people have a smartphone. The South Korean app makes use of a centralized approach as well. However, what really distinguishes the South Korean app from all the European ones is that they are using GPS instead of Bluetooth. Hereby, the users routes are tracked and once he is infected the people he crossed paths with receive a message that they were in contact with an infected person (Ryan, 2020).

2.3.4 Threats concerning the app

The benefit of the contact tracing app is undeniable. However, also problems arise when it comes to the use of the app.

(16)

use both apps at the same time, so they might not get the alert if they were in contact with an infected person (Salathé et al., 2020).

The next problem the app might have is, whether people are actually properly using the app. As there is no obligation to use the app it is also no must to indicate it once you got infected. It could be possible that some people only use the app to see if they were in contact with an infected person. Only when people use it properly, the app can benefit the national health. The third aspect was already mentioned before, which is the privacy concern. This point also somewhat includes the previous two aspects. Due to privacy regulations not all technology can be used to ensure a properly working app (Cho et al., 2020; Keesara et al., 2020). For example, also including GPS alongside Bluetooth could help reduce those technical issues. What is more, people also might choose not to use the app because of privacy concerns. Hereby, the trust in the government and technology is an important part (Glanz et al., 2008).

All the threats can be seen as a barrier to use the app. However, in order to integrate the app properly the potential users have to accept it. The following chapter deals with the acceptance of the society.

2.4 Acceptance of the app

Two models are used to investigate the behavior towards the acceptance of technology and the perception of disease.

2.4.1 Unified theory of acceptance and use of technology

When implementing a new technology, it is important that the potential users accept it. The origins of technology acceptance underly conventional social psychology. The following paragraph provides the theories that build the basis for a unified theory.

The theory of reasoned action (TRA) explains human behaviors (Sheppard et al., 1988). TRA includes two concepts: the personal feeling about a behavior and the influence of important people. Davis et al. (1989) first adopted TRA for the use of technology and found very similar results than in other sectors. In the same year the theory acceptance model (TAM) was formed. It included perception about usefulness and how easy it is to use (Davis, 1989).

A vital part of psychology is also motivation. Intrinsic and extrinsic motivation form behaviors. Davis et al. (1992) put the motivation model (MM) into a technology context.

(17)

Later Taylor and Todd (1995) combined both the TAM and TPB (C-TAM-TPB). The model includes the concepts about the perception on influence of others, usefulness, feelings about behavior and control.

Building on older theory about behavior Thompson et al. (1991) formed the model of PC utilization (MPCU). The model’s main purpose is to forecast usage and acceptance of personal computers. Nonetheless, it also allows to examine information systems. The model incorporates concepts such as complexity, consequences, affection, fit, facilitating conditions and social aspects.

When implementing new technology innovation is a big part of it. In sociology the innovation diffusion theory (IDT) investigates innovation in all kind of sectors (Rogers Everett, 1995). Moore and Benbasat (1991) fitted the constructs of the model into a technology perspective. In total seven constructs build the IDT model. The constructs are image, visibility, voluntariness, demonstrability, compatibility, ease and advantage.

(18)

Table 1. Summary of theories and concepts (Venkatesh et al., 2003)

Models

Theory of reasoned action (TRA) Attitude towards behavior Subjective norm

Technology acceptance model (TAM)

Perceived Usefulness Perceived ease of use Subjective Norm

Motivational model (MM) Extrinsic motivation

Intrinsic motivation

Theory of planned behavior (TPB)

Attitude towards behavior Subjective Norm

Perceived behavioral control

Combined TAM and TPB (C-TAM-TPB)

Attitude towards behavior Subjective norm

Perceived behavioral control Perceived usefulness

Model of PC utilization (MPCU)

Job-fit Complexity

Long-term consequences Affect towards use Social factors

Facilitating conditions

Innovation diffusion theory (IDT)

Relative advantage Ease of use Image Visibility Compatibility Results demonstrability Voluntariness of use

Social cognitive theory (SCT)

Outcome expectations - performance

Outcome expectations - personal Self-efficacy

Affect Anxiety

(19)

user believes he has to put in the technology to use it. Performance expectancy refers to the degree the user believes in positive correlations with the technology. Facilitating conditions contain the infrastructure that is necessary for the technology. Social influence is the perception whether others think one should use the technology. Those four elements can be moderated by things like gender, age or experience. There is also an extended model called UTAUT2, which adds price value, hedonic motivation and habit to the existing elements (Venkatesh et al., 2012). However, in the scope of this research the added concepts of the UTAUT2 are not included. Table 2 shows the four concepts of the UTAUT model.

Table 2. UTAUT (Venkatesh et al., 2003)

Unified theory of acceptance and use of technology (UTAUT)

Performance expectancy Effort expectancy

Facilitating conditions Social influence

A survey in Belgium used the UTAUT model to investigate the acceptance of contact tracing apps during the corona pandemic. Apart from effort expectancy all basic elements of the UTAUT model influence the intention to use the contact tracing app. Additionally, the potential user’s innovativeness had a positive influence on the intention. Privacy-regulations had a negative relation (Walrave et al., 2020b).

2.4.2 Health belief model

In the 1950s US psychologists formulated the health belief model (Hochbaum, 1958). The health belief model investigates people’s behavior related to prevent, avoid or control a disease (Green et al., 2020). It originates from two major psychology theories. The first theory is the stimulus response theory, which basically explains that a stimulus is necessary to evoke behaviors. For example, a person tends to repeat his behavior when the action will be rewarded (Skinner, 1938). The second theory is the cognitive theory. It states that someone’s behavior is determined by the personal value and the expectation of an outcome. So in a health context this would be for example the value of staying healthy or the expectancy that measures against the illness succeed (Lewin, 1951).

(20)

someone has to be aware that his smoking might cause health issues. Perceived severity refers to the belief of how serious the condition can become. As an example, one should be aware that not treating a growing birthmark might lead to skin cancer and as a result could cause death. The concept perceived benefits is about how efficient the potential user thinks the action against the disease is. In other words, it explains whether a person can get a benefit out of his actions. For instance, quitting smoking can avoid lethal illness and safe money. Perceived barriers explain the potential “costs” there are when taking the actions. So, if you quit smoking at the beginning you might have negative psychological effects from addiction. These effects are the barriers that can keep you from quitting (Glanz et al., 2008). Over the time some researcher included a fifth element to the four constructs, which is self-efficacy. Self-efficacy refers to the belief of person that he or she is able to execute the necessary measures against the disease (Glanz et al., 2008). In the scope of this research self-efficacy is not taken into further consideration as the task is so simple to execute. Table 3 highlights the four constructs of the HBM model.

Table 3. HBM (Hochbaum, 1958)

Health belief model (HBM)

Perceived susceptibility Perceived severity

Perceived benefits Perceived barriers

(21)

(22)

3 Methodology

The following section provides the methodology for this research.

3.1 Research design

The aim of this research is to investigate people’s perception of the integration of the corona app. Therefore, the concepts of the UTAUT and the HBM model are used to investigate the perception.

Some of the concepts exclusively have already been studied. Combining multiple concepts derived from the different models about acceptance and perception of technology is helpful when it comes to integrating the new technology. Additionally, as the outbreak of COVID-19 is relatively recent and many people around the world are dying from it, new ways of fighting the virus are wanted.

As combining the UTAUT and HBM model in one research is a new approach, an exploratory research is the appropriate method (Barratt et al., 2011). The research can be categorized both as deductive and inductive nature. Deductive because existing theory from concepts is tested and inductive because the correlation between the concepts has not been tested yet and new aspects can result. A case study seems like the right approach, because of the exploratory characteristics and the relevance of the topic. Studying insights in its natural context, facilitates gaining relevant information (Fisher & Aguinis, 2017; Karlsson, 2016)

3.2 Case selection

(23)

Therefore, the two contact tracing apps are at different stages in their integration process and the Dutch version could learn from the German implementation.

In order to build theory out of those networks nine interviews were conducted. This is in line with Eisenhardt (1989) stating that between four to ten interviews are needed for building theory. What is more, it is important to select different cases. This was ensured by choosing both residents and professionals. The differences between the cases and networks enable theoretical sampling. Having different results means theoretical replication as well (Eisenhardt & Graebner, 2007). The majority of the interviews were conducted with residents from either the Netherlands or Germany. Only students were included ranging between the age of 19 and 30. For additional background information on the integration of the app professionals, that are working with the contact tracing apps, were interviewed as well. In the German Network that was one case from the RKI and one from the Grüne in the German Bundestag. For the Dutch network the professional case was a professional working on the adoption of the app for the ministry of health, welfare and sports (VWS). They can provide insights into what is working at the moment, what still needs improvement, what felt easy and what was particularly difficult when implementing the app. To improve validity only people with work experience of more than a year at the organization are included in the interviews. Table 4 provides a list of the interviews with the residents and table 5 shows the professional interview partners.

Table 4. Interview cases of residents

Network Case number Age Gender App-use

Netherlands NL1 29 Male No

Netherlands NL2 21 Female Yes

Netherlands NL3 23 Male Yes

Germany G1 25 Female Yes

Germany G2 27 Male Yes

Germany G3 26 Male Yes

Table 5. Interview cases of professionals

Network Case number Organization Years of

experience

Netherlands P1 VWS 20

Germany P2 RKI 8

(24)

3.3 Data collection and instrument

Most of the interviews were conducted in a semi-structured way, to get the most results out of them. Following an interview protocol enhances the validity and reliability of the data and ensures that no important details are left out. The semi-structured aspect grants space for spontaneous answers that might not have been answered in a structured way. Furthermore, the interviewee can provide more input, which can be relevant (Karlsson, 2016). However, some of the interviewees did not agree on oral interviews. Therefore, some of the interviews were in written form. The interviews were conducted online either via telephone, video messaging or mail. The average duration of the interviews was between 15 and 40 minutes. At the beginning of the interview, personal questions were asked like the function, experience in the organization or age. After that the questions were asked according to the interview protocol. The questions are based on the theoretical concepts such as the UTAUT and the HBM. Additionally, the regulations, technology aspects and initiation of the processes were included as well. Regarding the UTAUT, questions were asked about the different expectancies, the facilitating conditions and the social influence. The HBM model is included using the four aspects of perception. Questions regarding privacy regulations and technology is a combined topic as the use of different technology is bound to the privacy regulations. Construct validity is provided by conducting multiple interviews (Bagozzi et al., 1991). In appendix A there is the interview protocol with the residents. In appendix B the interview protocol with the professionals is listed.

3.4 Data analysis

(25)

concepts (UTAUT, HBM etc.) or if new factors arise new categories will be formed. Table 6 provides an example of the coding process. Appendix C lists the coded material from the paraphrased statements to the second order category.

Table 6. Example of coding

(26)

4 Findings

I this section I will present the findings of the interviews. As there are two different networks (the Netherlands and Germany) this section is divided in two parts as well. The first part of each section presents the findings based on the interviews with the residents. The second part is about the perception of the professionals. The finding of each case will be presented according to the concepts of the UTAUT and HBM models. This provides a clear structure and will make it easier for the reader to follow. The answers about the concept’s performance expectancy, perceived benefits and perceived barriers are in some cases relatively similar. Therefore, those three concepts are presented next to each other.

4.1 Network 1 – The Netherlands

In this network three residents and one professional share their opinions on the use of the contact tracing app. The app is called “CoronaMelder”.

4.1.1 Case 1 – NL1

The first case of this network is the only interview partner that does not actively use the app. He chose not to use it, because of lack of trust in the government. He stated that he believes that the government uses corona in order to gain information of and force rules on the residents.

Facilitating conditions

As the facilitating conditions deal with the infrastructure that is behind the app, it also includes how people get informed about the app. Additionally, they explain whether there is enough information about the app. NL 1 stated that he does not really monitor the likes of the app. He got informed about the by online media.

Social influence

(27)

“I think I am able to make my own decisions.”

Effort expectancy

As NL1 chose not to use the app, there was no statement based on the expected effort.

Performance expectancy

As for performance expectancy he admitted that he is not aware positive or negative personal correlations of the app and fighting the disease. He also stated that it might be beneficial for weaker members of the society, like risk groups and older people for example.

Perceived benefits

The same also holds for the perceived benefits. NL1 mentioned that there are benefits for risk groups and old people. What is more, he hopes that these risk groups are actually using the app.

Perceived barriers

As mentioned before the credibility of the government is the biggest issue for NL1 and one of the reasons why he chose not to use the app. He does not want to cooperate with everything offered by the government. He feels like he is being watched, which is why he does not trust the government.

One major barrier for NL1 was the technical aspect of the app. Firstly, he mentioned that the app might have bugs. Secondly, he states that the location tracking is not always working. As an example, he remarked that the app does not work in apartment buildings when users are on different floors. Then, there would contact messages even though there were no real contacts. Therefore, NL1 does not want to get into quarantine after falsely receiving a message.

Lastly, he believes that people should stay away from the screen more. Additionally, he thinks that technology should not be used for such purposes. Rather he stated that:

“In general, I think technology should be limited to benefit research for medicine instead of

locational services.“ Perceived susceptibility

(28)

Perceived severity

He regards the disease as very severe, mainly for people that are old or susceptible to such diseases.

4.1.2 Case 2 – NL2

NL2 chose to use the app. The reason for that is to get alerted, if she was in contact with an infected person. Additionally, she wants to contribute to preventing the virus from spreading.

NL2 first heard about the Corona Melder app via journals and newspapers. However, she wishes for more advertising regarding the corona app.

As an additional feature she would like to have the opportunity to share her name in the app after a contact message. Therefore, people could double check, if they know about this contact. This is meant to be able to verify, whether you really were in contact with this person. Also, she would want Bluetooth as a feature.

Social influence

The interviewee believes in the benefits of social influence stating:

“I think people are more intended to use or download the app when others in their

environment use it as well,…”

However, it is the choice of each individual to use the app or not. Especially across young Dutch people using the app is not very common. In fact, NL2 does not even know whether her friends and family use the app.

Effort expectancy

The use of the app was very easy for NL2. She quickly knew how to turn on the contact tracing.

(29)

actively use it. The third reason is that people might neglect the warning and avoid testing or going into quarantine.

She also mentioned that the app is not directed to people that are following the government’s rules, as those people do not have close contact with other individuals.

For all the above-mentioned reasons NL2 does not perceive the app as very efficient.

As stated, above NL2 chose to use the app. However, there are also concerns for her. Regarding the privacy regulations she is worried that her data is not safely stored and could be hacked. Furthermore, she believes that her anonymity is not preserved. The reason she chose to use the app is that the health of the public is more important for her than the privacy through the app. Also, she uses apps like social media, which store private data as well.

In relation to the technology, the interviewee’s biggest concern is to receive fake warnings due to false location tracking. For example, when the telephone is too close to the wall of your apartment. Also, because you do not know whom you have been in contact with you cannot verify for yourself whether you actually had close contact. Therefore, her trust in the app is low. What is more, she is not confident that elderly people know how to use the app.

Perceived susceptibility

The interviewee views COVID-19 as highly susceptible as it can be transmitted easily.

NL2 perceives the disease as severe. Especially, older people and people with previous severe illnesses are highly affected by the disease.

4.1.3 Case 3– NL3

NL3 chose to use the app as well, because he wants to know when he gets infected and be able to avoid spreading the virus himself.

(30)

he wished for was Bluetooth of NFC. He informed himself about the technology online and the privacy issues.

Social influence

For the interviewee peer pressure is a tool to persuade people into using the app. He also stated that he could not be influenced against the measures or tools. Yet, he believes that he does not need influence from others to use the app.

Effort expectancy

Regarding the effort NL3 stated:

“No effort besides installing it. My expectancy was met.” Performance expectancy

He expects two things from the app. First, that he does not get infected without even knowing. The second reason is that the number of people that unknowingly spread the virus will be reduced.

For him the benefit is that it is a quick way to know when you were in contact with a person that got tested positive.

NL3 said that he has no issues with the privacy of the app. This is due to the fact that he informed himself about the technology:

“Before this app was published the privacy issues were not completely addressed. This was a barrier for me, until they were addressed.”

(31)

The interview partner views the corona virus as highly susceptible. It can be transmitted through handshakes, touching the same surface or close contact.

The interviewee mentioned that Corona is a deadly disease for him.

4.1.4 Case 4 – P1

Case four is the first interview with a professional from the Netherlands. He works as an organizational consultant for the ministry of health, welfare and sports. He is the project leader for launch and adoption of the app.

Initiation process

The ministry of VWS started the initiation for the Corona Melder. Their responsibility was to develop the app and contact researchers. Additionally, they set the legal frameworks. The app was developed by the ministry with assistance.

Several other organizations worked together on the app. For example, the GGD is responsible for assisting the users with the app. Other expertise was conducted from public and private sectors such as lawyers and programmers, etc. They took the role as facilitators or sponsors.

Technology

User experience was an important aspect for the ministry of health. Therefore, they conducted field tests. They also tested the accuracy, which showed that it is at about 75 percent. The interviewee mentioned that it can be effective to extract data from the app, but it is not ethically desirable.

The main technological challenge is the reliability of the technology, according to P1.

Goals

The app has about 4,5 million users, but the ministry of VWS had no expectations about that. The scope of the app was mainly determined by law. The government even issued an assisting law for the Corona Melder.

Performance expectancy and perceived benefits

(32)

“Supporting the source and contact research of the GGD into the contacts of positively infected people.”

The code of the app is open source, which should benefit the trust. Also, the assisting law builds the organizational infrastructure.

For P1 trust is the most important aspect. Trust in the technology and the government. Via communication the ministry aims to improve trust.

Furthermore, the more the people fear the disease the more they will use the app.

4.2 Network 2 – Germany

This network provides perceptions from three residents and two professionals in this field. The German contact tracing app is called “Corona-Warn-App”.

4.2.1 Case 5– G1

G1 uses the app as she stated that every individual should participate in actions that prevent the virus from spreading further.

She first came across the contact tracing apps via media. Although, she consumes media on a daily basis vial social media and newspapers, the interview believes that there should be more promotion by the government.

Social influence

The social influence plays a big role in G1s opinion. For the corona app it is similar to any other sector and you are influenced by family, friends and close people.

Effort expectancy

(33)

As mentioned before the interviewee mentioned that everyone should do their best to help fight the corona crisis. This can be achieved when everyone that gets a notification, applies for testing and gets into quarantine. She stated:

“It might have a big impact on all of us.”

For G1 the benefit is to get warned when she was in contact with an infected person. The benefit is for every member of the society.

Regarding the privacy her concerns only played a medium role. As she also uses social media, she believes that those means are way more questionable, than the app.

As for the technology, she mentioned that the Corona-Warn-App might have issues or bugs. If there were too many bugs, the trust in the government could decline.

What is more, she stated laziness or people not taking the illness seriously can be barriers as well.

G1 believes that the virus is real, and it is a problem that people deny the disease.

For her the disease is severe. Furthermore, she said that also young people could suffer from it, but the older you get, the more severe the outcome can be.

4.2.2 Case 6– G2

G2 also chose to use the app, because he cares about his fellow citizens, especially elderly people. He believes that it is the interest of everyone to help fight the virus.

(34)

“So, I could check for myself if I was in contact with many people at that time”

Additionally, he wants the app to run on old generations of smartphones as well, as he read that currently they do not.

Social influence

He thinks that social influence only plays a minor role when it comes to using the app. He mentioned that not using the app seems socially acceptable. While, while he believes he was not influenced to use the app, he stated that social influence when fighting the pandemic is important to him. This is because some family members fall into risk groups and he wants to save them by testing regularly and minimizing contact.

Effort expectancy

For G2 the app is quiet user friendly and he believes that this is not a reason why young people are not using the app.

The interview partner wishes for the app to reduce the number of infections. He stated that it is beneficial for the weaker members of society. However, a lot of people must use the app for it to fully function.

As mentioned above the interviewee views the benefits of the app more towards risk groups. He chose to use the app in order to protect these groups.

Perceived barriers

The privacy concerns only play a minor role for G2, as he trusts the government. Also the accuracy problems do not hinder him from using the app. One remark he had was that it is not working on all generations of phones.

What is more, the app does not work in all countries.

(35)

Regarding the severity he does not perceive the disease very severe.

4.2.3 Case 7 – G3

G3 also uses the app, because he wants to enforce and facilitate conventional contact tracing. What is more, he wants the pandemic to end and thus, tries everything to help fighting the virus.

He read about the app online and in newspapers. However, he believes the government could advert the Corona-Warn-App more.

Social influence

In general, G3 believes that the public opinion plays a big role. Also, the credibility of the government is a vital aspect. If the credibility is low, people are less inclined to use it.

For him personally he is not really influenced. He stated that only parts of his friends use the app.

Effort expectancy

The interviewee did not have to put much effort into using the app. He just had to download it and activate it. The effort was as expected.

He expects from the app to have a positive impact on conventional contact tracing and the fight against the virus. He also stated:

“I do not believe that it is the only measure but plays a big role.”

Furthermore, he said that using the app is more convenient than other restrictions.

(36)

The privacy is not an issue for G3 and did not play a role when using the app. Also, he said that he uses social media, which he has more concerns with. Additionally, he trusts the government as their only aim is to end the pandemic as well.

Tracking the contacts might be not always accurate and people might not get warned thereafter, but this still was no issue for him. He mentioned that using the app with minor errors is still better than not using it.

He said that the virus is real.

He believes that especially for old and susceptible people the virus is severe.

4.2.4 Case 8 – P2

Case 8 is an interview with a PR representative from the RKI. He shared information about the acceptance of the app, technologies, barriers and partner organizations and their roles. This case is not solely based on the concepts of the UTAUT and HBM.

The initiation started when a German businessman and the ministry of health had a meeting about technologies during the pandemic. They aimed to work out technical solutions to help fight the virus. After making agreements the Deutsche Telekom and SAP were commissioned to develop the app. Scientists from all different sectors worked together with the RKI. For example, scientist that worked on technology that assists the fight against Ebola. Also, the Fraunhofer Institut worked on evaluating parameters.

Technology

At the beginning of the development the German government followed a centralized approach. However, the changed to a decentralized approach. It was always the plan to build upon existing technologies such as smartphones.

(37)

Regarding other technologies like GPS for example he said that GPS has advantages, but also a lot of disadvantages. As an example, GPS cannot evaluate whether a person is on the same floor of a building. Then there might be fake warnings.

So, in the end Bluetooth is the best option, even though it was not invented for such purposes.

Goals

The RKI did not set prior goals regarding the number of users for example. They were happy for every user. Until now the app has about 25 million users and is among the apps with the most downloads.

Performance expectancy and perceived benefits

The interviewee emphasized that the contact tracing app can only be a part of fighting the virus and is not the only measurement. Primarily, it helps to facilitate conventional contact tracing.

The RKI is not responsible for advertising so P2 could not provide much information on that topic. However, he stated that especially at the beginning there were a lot of advertisements about the Corona-Warn-App.

Privacy regulations play a big role when implementing such a technology. Society wanted the decentralized approach. The interview partner stated:

“… the decision in favor of the decentralized approach was also a decision in favor of data protection.”

They were aware that a data sparing approach would lead to more acceptance of the technology.

4.2.5 Case 9 – P3

The last interview was with a representative from the green party in the German Bundestag.

(38)

approach. At the beginning there was a hackathon of the crisis in order to find technology that can assist with preventing the virus from spreading. Another part of the discussion was to link more tools to the corona app like personal details. Furthermore, they discussed about the need of an accommodating law for the integration of the app.

The discussions took about three months until everyone agreed on the decentralized approach. In general, the collaboration on the contact tracing app worked pretty well.

The green party sees itself as the digital driver of the bundestag. Their role is to be a corrective for the federal government.

Technology

As mentioned above the decentralized approach was chosen. When comparing it to France, where they use a centralized approach and only two million people use the app. Way less than in Germany. The interviewee does not see any benefits, if the used GPS instead of Bluetooth. Additionally, he added that in Germany legally it would not be allowed to use GPS.

He stated that the app alone cannot combat the virus. Even if there was no privacy protection, the app would not be enough to prevent the virus from spreading.

In more recent time the app does not show as many contacts as it used to, because some technical aspects were changed. This leads to less trust by the society. With the last update a contact journal was implemented, where one can provide a list of people and locations he visited. What still is urgently needed is a cluster recognition, especially now that the number of infections is soaring again. What is more, not all laboratories are connected to the corona app, which P3 would really want. So, health institutions should all be digitalized. All these could have been implemented during the summer, where times were calmer.

Goals

The interview partner really emphasized that the app is just a part of solving the problem. The main goal of the app was:

“Basically, the intention was to give everyone a way to individually - That's very important - asses his own risk of infection.“

(39)

The expectancy one should have towards the app, should be limited to contributing a part to fighting the virus. There is a discussion going on at the moment whether the app generates this value.

P3 mentioned that many millions were spent on advertising on the media and on billboards.

Data protection is the biggest concern people have. What is more, a problem is that many people, including politicians, do not know how the app actually works. So, these politicians say that not enough people use the app because of the privacy. P3 however believe, that is only made up because they neglected to further develop the app over summertime. From a privacy standpoint there is no issue with using the app.

Additionally, people tend to feel a bit diverse towards the app at the moment because they do not really see the value of the app.

4.3 Summary of the findings

(40)

Table 7. Summary of the interviews with residents

Case UTAUT HBM

NL1 Hopes risk groups use the app; believes app uses location tracking; personally, not really socially influenced

App can benefit risk groups, no trust in government and technology; privacy concerns;

NL2 No correlation because not enough people use the app; across young people no social influence; more information/advertising wanted

people reject the warnings; relatively severe; privacy concerns; technology skills of old people

NL3 you quickly know when you have been in contact with an infected person; informed himself about app and technology behind it; Social influence important

Highly susceptible; privacy no issue; accuracy of Bluetooth can be an issue

G1 App should be promoted more; Social influence is a vital aspect

Everyone can benefit from the app; severe even for young people

G2 App is efficient if a large majority of people use it; Social influence is marginal; wants a feature to monitor when app did not function properly

Disease not that severe; not usable on all generations of phones; not sure about accuracy; privacy plays a minor role

G3 Social influence not too important; correlation of app use and benefits should be advertised more; positive impact on conventional contact tracing

(41)

5 Discussion

The following section discusses the findings of all cases and compare it to the preexisting theory. In total the discussion incorporates the perceptions of six residents and three professionals from each network.

Across the two networks every participant stated that the effort they had to put in was very low. They just had to download and activate the app. This also coheres with the quantitative study of Walrave and colleagues (2020b). They investigated that effort expectancy had no relation to the app use intention.

Social influence was generally regarded as important by the interviewees. However, the majority mentioned that personally they were not influenced to use the app. This is contradictory to Walrave et al. (2020b) as they found that social influence is among the most important aspects to predict the uptake of the app.

A difference between the networks shows in the knowledge about the technology. NL1 and NL2 did not know that the technology behind contact tracing is actually Bluetooth rather than location tracking. Therefore, their privacy concerns were the highest among all the cases. NL3 did not have issues with the privacy, because he really familiarized himself with the technology before using it. Also Walrave et al. (2020b) found out that the privacy issues have a negative impact on the use. All three German interviewees knew about the way the technology works. Still, they indicated the privacy concerns, but the perceived that the benefits outweighed the barriers.

According to most residents the governments should indicate the positive correlation and app use more. However, all professionals stated that the governments do a lot to communicate the correlation. For example, many millions were spent on advertising.

(42)

older. Kretzschmar and colleagues (2020) investigated that an app uptake of about 10% already makes a positive impact on society.

Shahnazi et al. (2020) found that perceived barriers is among the biggest reason not to use the app. During the interviews the barriers were also the aspects that were discussed by far the most. Various kinds of barriers were detected by the residents. Firstly, as mentioned before the privacy of the user was the most prevalent topic. One of the reasons for that is the credibility of the government. NL1 for example stated he does not want to be watched by the government. Another reason can be that the data is not stored anonymously. Nonetheless, all professionals agree that the app has no conflicts with privacy of the users. Additionally, P3 mentioned that the reason for the lack of trust in the government is, because the government failed to improve the app over summer.

The second barrier can also partly be linked to the privacy which is technology concern. NL2 for example argued that the app might be hacked, and private data can be retrieved from it. Furthermore, several cases had the concern that the technology might not function properly. Reasons for that could be that it does not work in buildings or it has technical problems in general. Those problems make the app less reliable. In fact, the apps do not have an accuracy of 100 percent. Yet, the Corona-Warn-App has an accuracy of 80% and the Corona Melder about 75%. Although, the number could be higher P2 stated that conventional contact tracing also does not work properly all the time. G2 mentioned that he would like to have a feature in the app that indicates at what time the app did not work. Then the user could check for himself if he had a lot of interaction during that time.

NL 2 proposed a feature where one could state his or her name in the app for people to check if they had contact, because she believes that most likely the possible contact is with a familiar person. The German version has a similar feature called the contact journal, which is comparable to conventional contact tracing. Hereby, the user can add the contacts he knows himself.

(43)

Walrave et al. (2020a) investigated that perceived severity and susceptibility do not really have an impact on the use of the app. However, P1 stated that the more people fear the disease they are willing to use the app. Most residents viewed the disease as severe and susceptible.

(44)

6 Conclusion

The aim of this research was to investigate how people perceive the integration of the contact tracing app both in the Netherlands and in Germany. Therefore, the UTAUT and the HBM models were used to determine people’s perception on the app (Hochbaum, 1958; Venkatesh et al., 2003).

In the context of the COVID-19 pandemic several researchers found that not all concepts of the models have an influence on the app use intention (Shahnazi et al., 2020; Walrave et al., 2020a, 2020b). The findings also suggest that some of the concepts are of higher relevance. For example, the perceived barriers payed the biggest role when it comes to the integration. Various reasons account to the perception of barriers. One major reason is the trust in the government. Low credibility of the government leads to people being less inclined to use the app. The trust in the government also contributes to the next barriers which are privacy and technology concerns. The privacy concerns are that data is not stored or handled safely. The technology concerns address the accuracy. People feared that the app might not work. However, the professionals stated that the app has an accuracy of 75 (Corona Melder) and 80 (Corona-Warn-App) percent.

According to most residents, the app mainly benefits weaker members of the society. So, the interviewees did not use the app out of perceived personal benefits.

The lack of perceived benefits as well as the number of barriers are reasons why people might not use the app and the integration is not as effective. The reasons are not always compliant with the perception of the government. Therefore, it is their responsibility to better address those concerns.

A second aspect of this research was to examine whether the Netherlands can learn something from the German integration, as the German app was integrated month before the Dutch app.

(45)

An issue the German app has at the moment is that people feel a bit numb towards the app. They do not get as many notifications and as a result question the benefit of the app. This is partly because the government failed to further develop the app during times with lower infection rates.

(46)

7 Limitation and future research

This section deals with the limitations of this research and provides opportunities for future research.

The first limitation originates from the case selection. For the interviews with the residents only students of similar age were interviewed. While this ensures better comparability across the cases, it makes the results less generalizable. In order to ensure more generalizability interviews with people with more diverse demographics can be interviewed (Karlsson, 2016).

Furthermore, only one of the interviewees did not use the app. If there were more cases that did not use the app, there could have been different perspectives about the concepts.

What is more, in the Dutch Network only one professional interview partner could be found, contrary to two professionals from the German Network. This may provide uneven insight into the considerations when integrating the app.

Additionally, some of the interviewees only wanted to give oral interviews. Therefore, the semi structured nature of the interviews got lost. As a result, it was not possible to elaborate on some questions, so some thought or perceptions might have been lost. For example, some interviewees did not really distinguish between severity and susceptibility.

(47)

8 Implications

This section deals with the implications. The implications are both practical and theoretical.

8.1 Theoretical implications

For the theoretical implications this research extends the studies on the two models investigating the use of the contact tracing apps during the corona pandemic (Shahnazi et al., 2020; Walrave et al., 2020a, 2020b). They use either the UTAUT model or the HBM. However, the UTAUT model does not regard the aspects of this disease. Yet, the coronavirus is a serious disease. Hence, the HBM describes the perception of the disease and actions against it. Contrary to the UTAUT the HBM does not specifically focus on the use of technology to prevent the spreading. What is more, Carpenter (2010) argues that the HBM alone is not enough to investigate the behavior towards an illness.

(48)

Table 8. Combined UTAUT and HBM Model

Combined UTAUT and HBM

(C-UTAUT-HBM)

Facilitating conditions Social influence Effort expectancy Performance expectancy Perceived barriers Perceived susceptibility Perceived severity

8.2 Managerial implications

The managerial implications deal with the perception and behavior towards the different concepts.

Firstly, more emphasize should be put on explaining the technology of the app. This could be indicated better in the app for example. Some of the resident did not know that the app uses Bluetooth instead of GPS. Therefore, they had privacy concerns with the technology.

Speaking of privacy concerns, the governments should address the fact more that there are no privacy issues with the app. No relevant personal data will or can be extracted from the app. As every participant addressed privacy concerns, more clarity on this topic should be expressed.

Master Thesis

Master Thesis

A perspective on the integration of contact

tracing apps during the COVID-19

pandemic

Table of contents

List of figures

List of tables

Abstract

1 Introduction

How do residents perceive the integration of contact tracing apps?

2 Theoretical background

2.1 Pandemic and COVID-19

2.2 The use of Information and Communication

Technology

2.2.1 Information and Communication Technology

2.2.2 Implementation and integration

2.2.3 Usage of technology during the pandemic

2.3 Contact tracing

2.3.1 Conventional contact tracing

2.3.2 Contact tracing app

2.3.3 Different approaches of the app

2.3.4 Threats concerning the app

2.4 Acceptance of the app

2.4.1 Unified theory of acceptance and use of technology

2.4.2 Health belief model

3 Methodology

3.1 Research design

3.2 Case selection

3.3 Data collection and instrument

3.4 Data analysis

4 Findings

4.1 Network 1 – The Netherlands

4.1.1 Case 1 – NL1

4.1.2 Case 2 – NL2

4.1.3 Case 3– NL3

4.1.4 Case 4 – P1

4.2 Network 2 – Germany

4.2.1 Case 5– G1

4.2.2 Case 6– G2

4.2.3 Case 7 – G3

4.2.4 Case 8 – P2

4.2.5 Case 9 – P3

4.3 Summary of the findings

5 Discussion

6 Conclusion

7 Limitation and future research

8 Implications

8.1 Theoretical implications

Combined UTAUT and HBM

(C-UTAUT-HBM)

8.2 Managerial implications