Project Topics
Get Complete Project Material File(s) Now! »
Crowdsourcing for environmental data collection
As mentioned, one of the applications of crowdsourcing is through citizen science. In citizen science, the crowd is asked to help in “collecting large quantities of data across an array of habitats and locations over long spans of time” (Bonney et al., 2009, p. 977). This data can be used to define trends, monitor changes and inform management decisions. Citizens have been involved in data collection for over a century, however the term citizen science has only appeared in the literature more recently (Zheng et al., 2018). With regard to the collection of geophysical data through crowdsourcing, the literature has been significantly increasing since 2012. From their extensive literature review, Zheng et al. (2018) were able to select 217 academic articles on this topic.
The collection of geophysical data for climate through crowdsourcing has been used and studied for meteorological knowledge (Zhu et al., 2019), biodiversity monitoring (Chandler et al., 2016), air temperature (Meier et al., 2017), air pollution (McKercher et al., 2017), paleobiology and phenology (Killen et al., 2017; Willis et al., 2017), oceanographic data (Lauro et al., 2014), water quality and quantity (Willis et al., 2017; Jollymore et al., 2017), flood risks, prevention and monitoring (Degrossi et al., 2014) and climate change adaptation and resilience (Sarker et al., 2020). The content of the literature tends to focus either on the specific crowdsourcing methods or on the broader management of crowdsourcing applications. With regard to management, most attention is paid to engagement strategies, data collection protocols and standards, sample design and the assimilation and integration of the collected data (Zheng et al., 2018).
Most of the crowdsourcing research acknowledges the importance of the motivations for users to join in crowdsourcing (Lenart-Gansiniec & Sułkowski, 2018). In fact, understanding these motivations is seen as crucial for the success of crowdsourcing initiatives as this has a direct effect on their outcomes. Due to the wide range of crowdsourcing applications, motivations differ greatly per initiative (Hossain & Kauranen, 2015). Additionally, Wang et al. (2018) wrote that crowdsourcing participants often are influenced by various types of motivations that can’t always be distinguished as intrinsic or extrinsic, which is why both should be taken into account.
Nevertheless, Hossain (2012) found that within the literature, intrinsic motivations, such as fun and enjoyment, tend to be dominant. However, when a task increases in complexity, motivation tends to become more extrinsic. A common extrinsic motivating factor found within the literature is financial reward. Other motivating factors include altruism, peer pressure, reputation, status and community identification (Hossain, 2012). With a more specific focus on co-creation, Füller (2010) defined ten motivation categories. Namely: “intrinsic playful task, curiosity, self efficacy, skill development, information seeking, recognition (visibility), community support, making friends, personal need (dissatisfaction), and compensation (monetary reward)” (p. 103). Finally, Ye and Kankanhalli (2017) found that crowdsourcing participants were more likely to engage when they expected the benefits of participation to outweigh the costs, such as time and (cognitive) effort spent on the task.
A few factors are found to be important in attracting and sustaining participant engagement in the projects for environmental data collection. First, both Groom et al. (2017) and Kobori et al. (2015) argue that it is important to recognize the participants’ contributions. This recognition should connect to the participation motivations that matter most to the participant (Groom et al., 2017). Volunteer recognition systems or tools that facilitate interaction between participants can be useful in increasing a sense of recognition. Adding, front-end evaluation is important for organizers to understand the needs of participants and use this knowledge to improve their experience (Kobori et al., 2015). Additionally, incentives such as gamification or competition can also increase participation rates (Zheng et al., 2018). Secondly, regular communication and feedback are key in both attracting participants and sustaining their engagement (Donnelly et al., 2014; Kobori et al., 2015). Finally, protocols for data collection should be simple and easy to execute (Kobori et al., 2015). Besides helping to increase the quality of the data, training and workshops can help make participation easier and increase engagement (Donnelly et al., 2014).
Role of technology
The use of technology plays a crucial role in the crowdsourcing process (Ghezzi et al., 2018). Mostly, advances in the internet-enabled virtual environment have strongly supported crowdsourcing organizations in reaching out to possible participants from all over the world through either their own platform or other communication platforms such as email, Skype and social media (Modaresnezhad et al., 2020). Within the literature, it is recognized that the fit of the technology used to perform the task increases the quality of the task performed, as well as the participants’ enjoyment of performing the task. The IT behind the platform is also important in increasing interactivity and communication between the participant and crowdsourcing organization, increasing speed of data collection, reducing costs, improving participant engagement and ensuring anonymity of the users. Moreover, media richness, through the use of multiple media and multiplicity of cues, helps reach a wider public and increases feedback immediacy and clarity (Modaresnezhad et al., 2020). Finally, technology can also be used for the gamification of tasks, which can increase engagement (Fritz et al., 2017).
With regard to environmental data collection, significant improvements in IT have accelerated opportunities for collecting, storing and processing the data both at a greater spatiotemporal resolution and at a lower cost (Muller et al., 2015; Zheng et al., 2018). First, improvements in technology have made it easier for participants to use the instruments for data collection and have increased the accuracy of the data. New types of cheap and robust sensors enabled a greater number of citizens to be involved in monitoring the weather, ecological variables, temperature and other atmospheric variables (Donnelly et al., 2014; McKercher et al., 2017; Muller et al., 2015; Chandler et al., 2016). The use of images and videos taken by citizens through their own devices has also gained popularity within various forms of environmental monitoring. Overall, the simpler the technology used for collecting and uploading the data is, the easier it is to attract participants (Zheng et al., 2018).
IT can also be used to engage participants through online workshops, images, guides, training videos and newsletters (Donnelly et al., 2014). Moreover, technology can even replace participants through the automatic data collection from citizen devices such as GPS systems or smartphones (Kietzmann, 2017). Opportunistic sensing, for instance through microwave links or mobile devices such as smartphones, cars or cameras, and data mining from social media platforms have been used for monitoring precipitation, floods, hurricanes, earthquakes, wildfires, river levels, air pollution and biodiversity (Zheng et al., 2018). Additionally, there is a wide availability of low-cost sensors that can automatically transmit and store the data acquired. Finally, improvements in technology have enabled the crowdsourcing organizations to store large amounts of data and to process these data in various new ways (Zheng et al., 2018). Earth scientists are increasingly using AI for all areas of geology and this helps them prepare the data, process the data and create models. This is a large breakthrough, as this provides a way that the data collected through crowdsourcing can now actually be used, while before inabilities in processing the data were still a large challenge that hindered practical implementation (Sun et al., 2022b; Zheng et al., 2018).
Self-Determination Theory
A central theory within psychology research into human motivation is Self-Determination Theory (SDT) (Ryan & Deci, 2000). As a result, various researchers have connected this theory to the motivations to join crowdsourcing projects (Wang et al., 2018; Zhao & Zhu, 2014; De Vreede et al., 2013). Ryan and Deci (2000) state that: “Motivation concerns energy, direction, persistence and equifinality–all aspects of activation and intention” (p. 69). De Vreede et al. (2013) add that motivation influences the form, direction, intensity and duration in which a task is undertaken. Therefore, understanding these motivations is crucial when aiming to mobilize others to act in a certain way (Ryan & Deci, 2000). Many different forms of motivations exist, however within SDT, the most prominent distinction is made between intrinsic and extrinsic motivation (Deci & Ryan, 1985). Whether one tends to be more intrinsically or extrinsically motivated depends on their degree of self-determination, which is the extent to which they can regulate their own behavior (Deci & Ryan, 1985). When self-determination is high, individuals do not need extrinsic rewards but rather find motivation to act internally. On the other hand, when self-determination is lower, individuals require more external support (Deci & Ryan, 1985). In this case, the motivation is extrinsic as it does not stem from the activity itself, but rather from a desire to achieve or avoid a result that is separate from the activity (Ryan & Deci, 2000).
On the other hand, Ryan & Deci (2000) argue that intrinsic motivation often increases persistence, creativity, engagement, enjoyment and performance. They explain this by the fact that “The construct of intrinsic motivation describes this natural inclination toward assimilation, mastery, spontaneous interest, and exploration that is so essential to cognitive and social development and that represents a principal source of enjoyment and vitality throughout life” (Ryan & Deci, 2000, p. 70). Therefore, people tend to be intrinsically motivated for acts that are of interest to them and that seem novel, challenging or of aesthetic value. As a subtheory of SDT, Cognitive Evaluation Theory (CET) gives three factors, namely competence, autonomy and relatedness, that can explain variability within intrinsic motivation (Ryan & Deci, 2000). Firstly, intrinsic motivation increases when acts are perceived as promoting one’s autonomy and competency (Deci & Ryan, 1985). This sense of competency can be enhanced by receiving positive feedback. Moreover, a sense of autonomy can be increased by opportunities for self-direction and choice (Ryan & Deci, 2000). Secondly, relatedness concerns the social environment in which an act is performed. This environment can enhance intrinsic motivation when it gives a sense of security and relational support (Ryan & Deci, 2000).
Social Exchange Theory
Within crowdsourcing there is a clear exchange between the participants and the crowdsourcing platform. Participants offer something that the platform asks for and expect some sort of reward for their effort (Karachiwalla & Pinkow, 2021). As such, Social Exchange Theory (SET) is deemed useful in examining participation and engagement on these platforms (Wang et al., 2018). SET (Blau, 1964) posits that social relationships are based on the norm of reciprocity, through which both tangible and intangible resources get exchanged (Gouldner, 1960; Molm, 1997). Social exchange ‘refers to voluntary actions of individuals that are motivated by the returns they are expected to bring’ (Blau, 1964, p. 91). During these exchanges, individuals aim to minimize the costs and maximize the benefits for themselves (Blau, 1964). Therefore, SET takes a cost-benefit perspective on human behavior, assuming that individuals will only take part in an exchange when they expect positive net reward (Molm, 1997). This means that the actor that provides the initial favor should believe that the other actor either has possessions or behavioral capabilities that are valuable to them. Besides goods and services, socially valued resources could be status, approval, companionship or psychological gratification such as self-esteem and satisfaction (Molm, 1997). As social exchange is always conducted voluntarily, it is not governed by any explicit rules or agreements (Wang, 2022). This means that the nature of the favor that is gained in return for helping the other is not specified in advance (Blau, 1986). As a result, the intention to participate in an exchange is often based on an expectation of future benefits and therefore connects to long-term relationships instead of singular exchanges. This also means that there is a higher interdependency and a need for trust in the other party (Molm, 1997; Blau, 1986). With regard to the continuance of social exchange, Molm (1997) states that the frequency of exchange will decrease when the produced value of interaction is lower than expected. On the other hand, frequency will increase when the produced value is higher. Finally, if the obtained value declines to zero, the relationship will cease to exist (Molm, 1997).
Social Identity Theory
A final theory that has been found relevant to citizen engagement in crowdsourcing, is Social Identity Theory (SIT) (Boons 2015; Federenko et al. 2017; Sun et al. 2022a). Tajfel (1978) describes social identity as “that part of the individual’s self-concept which derives from his knowledge of his membership of a social group … together with the value and emotional significance attached to that membership” (p. 63). According to Blader and Tyler (2009), social identity is important when aiming to understand individuals’ behavior within organizations. More specifically, their level of engagement is dependent on how being a part of the group or organization makes them feel about themselves. This connects to the group engagement model that is composed out of two components, namely cognitive and evaluative (Tyler & Blader, 2003). First, the cognitive dimension relates to organizational identification, namely ‘the perception of oneness with or belongingness to an organization, where the individual defines him or herself in terms of the organization(s) in which he or she is a member’ (Mael and Ashforth, 1992, p. 104, as cited in Boons et al., 2015). The more one identifies with the organization, the higher the behavioral effort for the group will be (Blader & Tyler, 2009). Secondly, the evaluative dimension relates to how an individual feels about the status of the organization that they are a part of. When they are proud of the organization, this pride will likely reflect positively on their self-concept (Blader & Tyler, 2009). Another evaluative factor is perceived respect, which is the members’ evaluation of how he or she is valued by other members (Boons et al., 2015). Overall, the more respected and proud a member feels within a group, the more they will identify with this group, and thus the more likely they are to engage (Blader & Tyler, 2009).
Theoretical framework
A useful framework for analyzing crowdsourcing is proposed by Ghezzi et al. (2018) (See figure 1). By using the Input-Process-Output (I-P-O) model by McGrath (1964). They took a process view on crowdsourcing to prevent fragmentation. This approach connects to literature on open innovation and co-creation, which are also often presented as processes (Ghezzi et al., 2018). As depicted in figure 1, ‘Input’ concerns the task that has to be undertaken by the crowd. When complexity of this task for the user increases, motivation to participate decreases (Karachiwalla & Pinkow, 2021). Following, the ‘Processes’ can be guided through session management (how the crowdsourcing platform manages the project), people management (how the crowdsourcing platform attracts and motivates the participants), knowledge management (what is done with the collected data by the organization; both during the collection and after) and the technologies that are used to manage the process. Finally, ‘Output’ includes the benefits that the completed task holds for both the crowdsourcing platform and the participants (Ghezzi et al., 2018).
Case and sampling
The organization that was studied was selected due to their success in attracting participants for climate data collection all over the world. In fact, their climate data collection project was so successful that they found individuals from all seven continents, even Antarctica, to participate. Additionally, this led them to break a world record. By examining how they did this, advice can be formulated for other organizations that want to use crowdsourcing for a similar purpose. Within this project, most of these participants collected the data voluntarily, which makes it particularly interesting to investigate what motivated them. Emails were sent to an available selection of 29 individuals that participated in the project. Only in the first round, there was already a response rate of 13 people. One of these was excluded since he did not participate personally, but only helped find someone else for the job. Another participant did not reply to the follow-up email that was sent to schedule the interview. As one of the interviews was with two pilots at the same time, a total of 9 interviews with the participants were conducted.
The relatively small size of the sample can be justified by the fact that this is a relatively specific case study and that there was a limited amount of time for gathering and analyzing the data. Furthermore, because the project and the interview questions were quite straightforward, and the process of participating was very similar for every participant, no significant new findings were found from the final interviews. This indicates that data saturation might have been realized despite the small sample. Additionally, organizational documents and one internal interview were also collected through personal connections within the organization, as one of the researchers works there. Documents were only taken into account when they were deemed relevant for answering the research question. Adding, the interview with one of the employees of the organization helped the researchers gain a deeper understanding of the project.
Interview topics
Following the inductive approach of this thesis, the interview topics were based on a theoretical framework formed by existing research (Rowley, 2012). An interview guide was developed with some main questions to be asked, but since the research is exploratory, the researchers also wanted to leave room for other questions that came up during the interview itself, including follow-up and probing questions. Structuring the interviews too tightly would have prevented interviewees from being able to share their personal experiences, feelings and motives, which is precisely what was aimed to learn more about (Silverman, 2015).
Based on the I-P-O model (Ghezzi et al., 2018), the questions that were asked during the interview were divided into questions about how they felt before participating (input), during participating (process) and after participating (output). Questions about the process before participation focused on how they found out about the project and what made them decide to apply. Following, questions about the participation process focused on the experience of participation, the complexity of participating, how the experience could be improved and the quality and quantity of communication with the organization and other participants. Finally, respondents were asked what they, looking back, felt that they had gained and lost from participating and if they would participate again in similar projects or in longer term projects. Table 2 shows some examples of how the interview questions related to the theoretical framework used in this paper. A full interview guide can be found in the appendix (see Appendix A).
Presentation of the project
Normally, drone operators registered in the platform are invited to paid missions nearby their location. However, from December 10th 2021 to February 10th 2022, the company organized a volunteer-based project to break a world record. The project aimed to gather “the largest online photo album of aerial photographs”. Specifically, drones were used to collect the photos and the images showed the effects of climate change on local ecosystems, worldwide. The pictures showed natural areas that are at risk or that were relevant to environmental monitoring in a different way. When describing the goal of this project, the flow manager mentioned:
“I mean, the world record in itself wasn’t really the goal. It was more about showing how, what drones can do. And like, what, pretty much what anyone can do with them with the standard commercial drone, like, anyone can go out, monitor or map a piece of land with their drones. And that can be used for research for researchers who know how to use that data. So it was basically just showing the, yeah, the power of crowdsourcing, and how a even the simple drones can be really used for something good when it comes to climate change monitoring” (R11).
Additionally, the reason that they decided to make the project voluntarily-based instead of paid was “because we really just wanted to show that people come together as volunteers for the cause of climate change. And basically, the drone community comes together.” (R11).
The company invited the drone operators mainly with two channels, their own platform and social media. They reached out to all their registered members through their platform and invited other operators outside it through social media and the internet. Over the course of the two months, 31 drone operators collected 22,909 images, covering 22 countries and all 7 continents, resulting in setting a new world record. After the project ended, all the pictures were made publicly available on the organization’s website and the participants’ names and photos were also published to give them recognition.
Table of contents :
1. Introduction
1.1 Background
1.2 Research problem
1.3 Research aim and questions
1.4 Thesis structure
2. Previous research
2.1 Conceptualization of crowdsourcing
2.2 Crowdsourcing for environmental data collection
2.3 Role of technology
3. Theory
3.1 Self-Determination Theory
3.2 Social Exchange Theory
3.3 Social Identity Theory
3.4 Theoretical framework
4. Methodology and methods
4.1 Methodolog
4.2 Research design
4.2.1 Methods of data collection
4.2.2 Case and sampling
4.2.3 Interview topics
4.2.4 Methods of data analysis
4.3 Reliability and validity
5. Presentation of object of study
5.1 Presentation of the company
5.2 Presentation of the project
5.3 Presentation of the respondents
6. Analysis
6.1 Engagement Strategy
6.1.1 Channels of communication
6.1.2 Participation of the drone operators
6.2 Self-Determination theory
6.2.1 Intrinsic motivation
6.2.2 Extrinsic motivation
6.3 Social Exchange Theory
6.3.1 Costs of participation
6.3.2 Benefits of participation
6.4 Social Identity Theory
7. Discussion
7.1 Initial motivations to participate
7.2 The process of participation
7.2.1 Clarity of instructions
7.2.2 Sense of communit
7.3 Benefits of participation
8. Conclusion
8.1 Conclusion and implications
8.2 Limitations and future recommendations
References
