Conventional citizen scientists

One reason that citizen science projects rarely impact volunteers’ attitudes and behaviors is that volunteers often enter into projects already highly engaged with science and/or the environment and so are unlikely to show gains in these metrics; this phenomenon, often referred to as self-selection bias, can be compounded by a “ceiling effect” in measurement that makes differences in key outcome variables difficult to detect [15–18]. For instance, Brossard et al. [19] found that citizen scientists in the avian citizen science project “The Birdhouse Network,” were much more likely to express eco-centric, pro-conservation attitudes than people who weren’t citizen scientists; as a result, they concluded that “it may be difficult, and even unnecessary, to change project participants’ attitudes toward the environment” (p. 1113).

In addition to being more engaged in science and the environment than most people, the sociodemographic characteristics of citizen science volunteers also differ from those of people who are not citizen scientists. Specifically, with a few exceptions [20], when compared to the general public citizen scientists are more likely to be well-educated, relatively affluent, able-bodied, employed in science-related fields, and White [5,21–23]. To take just one particularly striking example, more than 99% of participants in a field-based avian citizen science project operating in nine countries in southern Africa identified as White [24].

The lack of diversity in citizen science has many negative consequences, including patchy data [21,25], the inequitable distribution of participation benefits [23,26], and a homogeneity of perspectives within the field of science [27].

Facilitator organizations

One means of reaching new sources of citizen science volunteers is through “third party” or “facilitator” organizations already embedded in communities, such as schools, churches, corporate social responsibility programs, and community groups [10,23,26,28–32]. Because these organizations focus on volunteering or member engagement outside the specific context of science or the environment, their volunteers may have different characteristics than typical, self-selecting citizen science volunteers who independently seek out experiences with science and nature [28].

There are a growing number of facilitator organizations that engage volunteers in citizen science. For instance, libraries in the United States, Europe and the Middle East are now organizing successful opportunities for patrons to participate in citizen science [33,34]. Schools, another type of facilitator organization, are also engaging students in citizen science. For example, a camera trapping project has partnered with classrooms in India, Kenya, Mexico and the US to collect data on mammals [35], and an ecology project engaged primary and secondary school children in Austria in monitoring bees, hedgehogs, and other animals [36]. An environmental health citizen science project in the US state of Arizona engaged socio-demographically diverse volunteers in rainwater monitoring in part through partnerships with community health workers called promotoras [37]. An avian citizen science project called “Celebrate Urban Birds” successfully engaged diverse volunteers living in urban centers through community partnerships with faith-based groups, community centers, youth groups and rehabilitation programs [38].

One type of facilitator organization that has been little-studied in the context of citizen science participation (though see [28]) is corporate volunteer programs. The goal of these programs, which are one type of corporate social responsibility, is to encourage employees of a company to contribute time to their communities, non-profits or other charitable groups through planned volunteer activities [39]. These programs support important service activities, can improve employee morale, and may improve a company’s reputation [40]. Corporate volunteer programs sometimes involve citizen science. For instance, companies such as HSBC, Starbucks, Alcoa and Syngenta have partnered with the environmental charity Earthwatch to provide opportunities for cohorts of employees to work as field assistants for environmental scientists [41]. Partnerships such as these can have a positive impact on participants’ commitment to and efficacy for supporting sustainability initiatives at work and home [42,43]. Importantly, Earthwatch expeditions are uniquely immersive experiences within the context of citizen science, often costing hundreds of dollars and requiring international travel; most citizen science projects have far fewer barriers to entry.

Motivations to participate

Some authors have pointed out that volunteers from facilitator organizations may have different motivations for participating in citizen science than typical volunteers who are often intrinsically motivated to participate [44,45]. For instance, students sometimes participate in citizen science because it is required by a school [46–49]. Other motivations potentially unique to facilitator organizations might include a member of a corporate volunteer program who is enticed to volunteer because prizes like paid time off are awarded to participating employees, a member of a church volunteer group who may feel group pressure to participate in a program, or Girl Scouts who join a citizen science program because their troop leader signs them up to participate [31]. Volunteers who seek out citizen science experiences independently of a facilitator organization are unlikely to share these group-oriented motives.

Self-Determination Theory (SDT) is an empirical model of human motivation that is well-suited to addressing questions of motivations in citizen science volunteers from facilitator organizations [45]. SDT models motivation on a continuum of increasing “self-determination,” that is, from behaviors that one is compelled to do by external forces to behaviors whose motivations are experienced as emanating entirely from the self [50]. Specifically, SDT identifies different types of motivation along this continuum of increasing self-determination. The least self-determined motivation is amotivation—doing an activity despite any internal motivation to do so. Amotivation is followed, along the self-determination continuum, by extrinsic motivation. SDT identifies different types of extrinsic motivation embodying increasing levels of self-determination, from external regulation (doing things to get a reward or avoid punishment) to introjected regulation (doing things out of a sense of guilt or ego) to identified regulation (doing things because they are identified as important). The most self-determined motivation is intrinsic motivation—doing things for fully internalized reasons, such as because the behavior is simply enjoyable ([50]; Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Motivations arranged along a spectrum of increasing self-determination.

More self-determined motivations (i.e., those regulated by factors internal to the self) are facilitated by meeting individuals’ needs for competence, relatedness and autonomy and are linked to longer-lasting and more impactful behaviors. Figure adapted from Ryan & Deci [50] and does not include integrated regulation, which is sometimes situated between identified and intrinsic regulation.

https://doi.org/10.1371/journal.pone.0331221.g001

A robust body of literature, including in the contexts of education, health behavior change, psychotherapy and prosocial behavior, finds that people are more likely to learn from and/or sustain engagement in activities that they do for more self-determined reasons [51]. On the other hand, activities performed for less self-determined reasons (i.e., amotivated, external, and introjected motivations) are likely to be short-lived and have little impact on participants [52]. For instance, people who engage in pro-environmental behaviors like recycling or saving energy for highly self-determined reasons (e.g., because they enjoy doing so, or because doing so aligns with their internalized vision of themselves as a conservationist) engage in more pro-environmental behaviors, and are more likely to sustain this behavior, than people who engage in these behaviors for less self-determined reasons (e.g., because they would feel guilty if they didn’t recycle [53–55]). In the context of education, students taught using an externally motivating pedagogy that relies on threats, rewards, or shame lose interest, initiative and persistence in learning, and learn less effectively, compared to students taught in a manner that supports more self-determined motivations [56–58].

Importantly, SDT posits that motivations are malleable, and that a person’s motivations can become more self-determined if they find that an activity meets what SDT proposes are three fundamental human needs: competence, relatedness, and autonomy [50]. Competence refers to the need to feel effective and capable of meeting challenges. Relatedness refers to the need to belong and feel valued by others in a community. Autonomy refers to the need to feel that your behavior is self-directed and that you have freedom in the choices that you make. If an activity supports these three fundamental needs, a participant is more likely to experience self-determined motivation and, thus, to gain more from the experience.

A number of studies have examined motivation in the context of citizen science [59]. This body of work finds that most citizen scientists participate for highly self-determined reasons, including intrinsic motivations such as participating because it is exciting, interesting, or enjoyable, and identified motivations such as participating because one values getting to contribute to science or help the environment [60–65]. Less self-determined motivations are rarely measured and sometimes deliberately excluded from analyses of volunteers’ motivations; one study excluded external “pressure” as a motivation because the researchers doubted that this motivation was relevant to volunteer citizen scientists [45], and another study excluded motivational data from students who participated in a project because they were required to [36]. When less self-determined motivations are measured, they are typically, though not always, rated by volunteers as less important than more self-determined motivational factors. This includes externally regulated factors such as participating in order to compete on leaderboards [66,67], to receive something in return [68,69], to improve job prospects [70], to gain knowledge [71], and to avoid negative consequences [72]. It also includes introjected factors such as to gain recognition and reputation [73–75], or because one feels obligated to contribute [76–78]. Studies of associations between citizen scientists’ motivation and participation patterns find that volunteers with more self-determined motivations to contribute to citizen science tend to engage more deeply in projects [79] and provide more sustained – and more novel – contributions [45,80].

Importantly, few of these studies used the SDT motivational continuum explicitly when measuring motivation and so we have applied these labels a posteriori based on our interpretation of the factors and items used. Studies of citizen science motivation that do focus on SDT include various studies evaluating the three fundamental needs predicted by SDT: whether citizen science fulfills these needs [45], how failing to fulfill these needs leads to participant drop-out [81], and whether making design changes to meet these needs improves a project [82]. Perhaps the most comprehensive treatment of SDT in the context of citizen scientists is Rutten et al. [48], who reviewed the literature on motivation to map citizen scientists’ reported motivations onto the SDT continuum. Recognizing this gap in citizen science motivational studies, researchers have called for more studies investigating citizen science volunteer motivations using the framework of SDT [83].

Objectives

While various researchers have suggested that partnerships between citizen science projects and facilitator organizations could facilitate deeper engagement by diverse volunteers [28], few studies have tested this explicitly and none have evaluated the importance of different types of motivation, particularly less self-determined motivations, for members of facilitator organizations. Studying members of facilitator organizations provides a unique opportunity to test enduring questions about citizen science volunteer retention, participation equity, and motivation.

For this research, we studied volunteers who participated in citizen science through a corporate social responsibility program. Specifically, in 2019, SciStarter.org, an online citizen science project clearinghouse, began a partnership with a large Fortune 500 company to create a customized online citizen science portal for their corporate volunteer program (to protect the anonymity of participants, we are unable to share the name of the company in this publication). Through this partnership, employees are able to find and contribute to citizen science projects that are hosted on the SciStarter portal. Participation in the program is voluntary, but highly encouraged by the company in several ways. For instance, certain employees are selected as leaders in promoting participation in the program and progress towards program goals (total number of volunteer hours, etc.) are tracked and publicized by the program. Within the context of this corporate volunteer program, our study had four objectives, with the overarching goal of better understanding the role of facilitator groups and motivations in connecting new audiences with the benefits of citizen science participation. We achieved each of these objectives in this study.

1. Characterize the overall participation of volunteers in the citizen science program.
2. Evaluate whether these corporate volunteers are more diverse than typical citizen scientists.
3. Evaluate the motivations of these corporate volunteers using the SDT motivational continuum and explore socio-demographic differences in motivations.
4. Identify factors, including motivations, associated with volunteers’ depth of engagement in the program.

Digital metadata

SciStarter collects digital metadata about volunteer participation in the projects on its corporate volunteer portal using an Application Programming Interface (API). We downloaded this contribution data for each corporate volunteer from the start of the citizen science initiative in November 2019 through August 2022, when we began analyzing the data (i.e., nearly three years of contributions). From this data, we were able to derive detailed quantitative data about each corporate volunteer’s participation in the citizen science program. All digital metadata was shared with the authors in accordance with SciStarter’s privacy policy.

Survey data

We surveyed a sample of the participants in the corporate citizen science volunteer program between April 1, 2020 and January 31, 2022 in order to collect socio-demographic data and data on volunteer motivations. The survey distribution protocol prioritized surveying volunteers before they began participating in the citizen science program. Links to the survey were posted on SciStarter’s online portal for corporate volunteers in April 2020 (approximately five months after the citizen science program launched) and remained there until April 30, 2021. Additionally, in May 2020, we sent emails to 490 employees who had already created a SciStarter account encouraging them to take the survey. We also sent automated emails promoting the survey to any employees who signed up for a SciStarter account between May 2020 and January 2022. A maximum of two reminder emails were sent to employees who did not complete the survey. Lastly, we shared a link to complete the survey during webinars introducing employees to citizen science. Employees who completed the survey were entered into a drawing for a $100 gift card in order to encourage participation. Written informed consent was required from participants before starting the survey.

In total, we received 459 responses to this survey. We eliminated substantially incomplete surveys (respondents who answered <10% of questions), survey responses that were clearly rushed (completed in less than 60 seconds), surveys with substantial straightlining, and three duplicate surveys (we eliminated the later response for each of these) for a final survey sample of 388. We were unable to calculate a precise response rate for this sample, as we did not know exactly how many volunteers saw the invitation to participate in the survey given the variety of channels through which the survey was promoted. However, given the total population of corporate citizen science volunteers (13,425; see below), this sample represents approximately 3% of the volunteer population, though 12% of employees who completed the survey did not go on to participate in the citizen science volunteer program (see below).

The survey had three primary goals. First, we evaluated participants’ prior citizen science experiences. To do so, we asked respondents whether they had experience participating in citizen science and, if so, to indicate what project(s) they had participated in and how frequently they participated. Second, we collected socio-demographic data from participants. Specifically, we asked respondents to indicate their gender, race/ethnicity, age, the level of education they had completed, whether they worked in science and technology-related fields (STEM), and their political ideology on a five-point scale from very liberal to very conservative. Third, we evaluated participants’ motivations for contributing to citizen science. We did so by adapting established instruments for assessing motivation along the SDT continuum to the context of citizen science. Respondents were asked why they were “participating in (or considering participating in) citizen science?” Fifteen items assessed volunteers’ motivations, with three items each for intrinsic regulation (e.g., “because I simply enjoy participating in the projects”), identified regulation (e.g., “because I really value getting to contribute to science”), introjected regulation (e.g., “because I’d feel bad if I didn’t do citizen science”), extrinsic regulation (e.g., “because it’s what I’m supposed to do”), and amotivated regulation (e.g., “I feel that contributing to citizen science is a waste of time”). The items used in the survey were adapted from the Learning Self-Regulation Questionnaire [84] and the Motivation Toward the Environment Scale [54]. Items were presented in a random order. Each item was assessed using an 11-point Likert scale from “Not at all agree” (0) to “Strongly agree” (10).

Respondents also answered a free-response question assessing their motivation for participating in citizen science (“Why are you participating in (or considering participating in) citizen science?”). This question preceded the Likert-type questions in order to avoid biasing responses. The survey also asked a variety of questions unrelated to this study (volunteers’ connection to nature, etc.). We used volunteers’ email addresses from the survey to link survey data to the digital metadata.

Analysis

To assess volunteer participation in the program (Objective 1) we used the digital metadata to calculate the number of volunteers who participated in the corporate citizen science program, the number of volunteer hours spent on each project, and the number of projects volunteers participated in. We also calculated individual-level participation in the program in terms of each volunteer’s hours of participation, number of projects, and number of contributions (i.e., the number of times a volunteer participated in a given project on a given day). Next, we categorized each project according to its discipline (i.e., ecology/environment, health/medicine etc.) and mode (online or offline) using multirater-consensus categorizations employed in a previous publication analyzing projects hosted by SciStarter [5]. We then calculated the number of projects from each discipline and mode, and the number of volunteer hours spent on each discipline and mode. Lastly, in order to evaluate the representativeness of the survey sample, we used Wilcoxon signed-rank tests and Chi-square tests to compare the participation patterns of the full population of corporate volunteers to the participation of volunteers who took the survey [85].

We used the survey data to report descriptive statistics of the socio-demographics of the volunteers in the corporate citizen science program (Objective 2). We used Chi-square tests to compare these socio-demographics to the socio-demographics of SciStarter volunteers who aren’t members of a facilitator organization, which we assessed in a prior study [5]. We also compared these socio-demographic factors to the socio-demographics of the United States population (most of these corporate volunteers are based in the United States, as are most projects hosted by SciStarter).

We used the survey to conduct a quantitative assessment of the motivations of these corporate volunteers (Objective 3). We performed exploratory factor analysis (EFA) on the 15 motivation items in order to assess the validity of the proposed constructs using the “psych” package in R [86]. We opted for EFA rather than confirmatory factor analysis because these items were modified from their original instruments and had not been used in the context of citizen science before. We used the Kaiser Criterion (counting the number of eigenvalues greater than one) to establish the number of factors specified in the EFA [87] and used oblique rotation and Principal Axis Factoring. We calculated the aggregate mean score for each factor resulting from EFA by averaging the items that made up each factor for each respondent. We then calculated a “Relative Autonomy Index” (RAI) score for each respondent. The RAI is a single measure of self-determined motivation made by collapsing multiple motivation factors and is often used in studies utilizing SDT [88]. We calculated the RAI by summing the mean motivation scores for each volunteer and weighting each factor according to its degree of self-determined motivation (i.e., Autonomous – Controlled – Amotivated*2; see below). We then rescaled the RAI so that its scale was the same as the three individual motivation factors (i.e., had a range of zero to ten). Respondents who completed fewer than half the items in a given factor were not given an average score (~8% of respondents for each factor). We use quotes from the free-response motivation question to provide more context to the quantitative data. Next, we used multiple regression to explore whether the socio-demographic factors measured in the survey (race/ethnicity, gender, education, political ideology, STEM occupation and age) predicted volunteers’ self-determined motivation as assessed by the RAI motivation index using the “lm.beta” package in R [89].

To explore links between self-determined motivation and participation levels among corporate volunteers (Objective 4), we first calculated the association between volunteers’ mean score for each motivation factor and the number of hours they spent contributing to the citizen science program. We also calculated associations between volunteers’ RAI score and the number of hours they spent contributing to the program. Next, we used multiple regression to assess whether associations between motivation and participation were robust to other factors that might influence engagement in the citizen science program [89]. For this regression, the outcome variable was the total number of estimated hours a volunteer spent doing citizen science over the sampling period (November 2019 – August 2022). Because a small proportion of volunteers contributed far more hours than most volunteers, we log-corrected this duration data to reduce skew as has been done in other studies of citizen science participation with highly skewed participation data [74]. We included the motivational factors identified in the EFA from Objective 3 as predictor variables. We controlled for the socio-demographic factors collected in the survey (race/ethnicity, gender, education, political ideology, STEM occupation and age) since socio-demographic factors are known to influence participation in citizen science [5] and motivation [90]. We also controlled for the date a volunteer took the survey, given the wide date range over which volunteers completed the survey, and controlled for whether or not a volunteer had experience participating in citizen science prior to taking the survey.

Goal 1: Participation patterns

Over the nearly three year sampling period, 13,425 volunteers contributed 31,143 hours to citizen science through this corporate volunteer program. This volunteer total represents approximately 11% of the total workforce of this company. These volunteers’ total hours of work equate to approximately $990,000, according to the US national volunteer hourly in-kind rate [4,91]. Individual volunteers’ hours of contributions ranged from 15 minutes to 381 hours. As with much citizen science [79], participation was highly skewed; the average volunteer participated for just 2.3 hours, the median volunteer participated for 1 hour, and the volunteers in the top 20th percentile of volunteer hours contributed 64% of the total hours. Approximately 59% of volunteers participated for an hour or less (N = 7,951) and just 6% of volunteers contributed more than five hours total (N = 826).

Volunteers who participated in the program made between 1 and 535 individual contributions to citizen science, with the average volunteer making 2.9 contributions and the median volunteer contributing once (“contribution” in this case means a volunteer participated in a given project on a given day; this definition is used because the amount of effort needed to submit a single data point varies widely from project to project). Contributions were inconsistent across the sampling period; for instance, contributions spiked in April of each year, presumably in conjunction with “Citizen Science Month” initiatives organized by SciStarter each April. Individual corporate employees who participated in the program contributed to between 1 and 18 different projects, with the median volunteer contributing to one project and the mean volunteer contributing to 1.4 projects. Approximately 62% of volunteers contributed just once, 11% of volunteers made multiple contributions to a single project and 27% of volunteers contributed to multiple projects (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Citizen science participation patterns of corporate employees according to population-level data for the entire volunteer sample (N = 13,425) and survey data from a smaller subset of volunteers (n = 343).

https://doi.org/10.1371/journal.pone.0331221.t001

Volunteers participated in 49 different citizen science projects overall. Approximately half of these projects could be completed entirely on a computer and approximately half of the projects were related to ecology and the environment. There was a large amount of project-specific participation skew, with volunteers contributing fewer than 10 total hours to 17 of these projects and contributing more than 1,000 hours to seven different projects (Table 2). Health-related projects were the most popular (53% of all volunteer hours), followed by ecology and environment-related projects (22% of hours) and then geology and earth science-related projects (14% of hours). Approximately 65% of volunteer hours were spent on online-only projects, while 35% of volunteer hours were spent on projects with an offline component.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. The ten most popular projects for participants in the corporate citizen science volunteer program.

https://doi.org/10.1371/journal.pone.0331221.t002

Among the 388 employees who completed the survey, 85% had no experience with citizen science prior to taking the survey, and just 3% of volunteers had done citizen science more than once prior to taking the survey. Of these 388 surveyed employees, 88% would go on to participate in citizen science (per the digital metadata). Volunteers who took the survey participated in more citizen science, and were more likely to participate in multiple projects, than volunteers who did not take the survey (Table 1). However, the effect size of taking the survey on citizen science participation was small (Cramer’s V ≤ 0.04).

Goal 2: Volunteer socio-demographic diversity

The demographic summaries reported below only include the 343 survey-takers who would go on to participate in the citizen science program (Goal 2 focuses on the diversity of citizen science participants rather than prospective participants; in addition, there were no statistically significant differences between volunteers who did and did not go on to participate in citizen science according to any of the socio-demographic variables we collected). Among these respondents, 60% identified as non-Hispanic White, 55% identified as women and the average age was 44 years old. Approximately 28% of survey-takers reported having a graduate, professional or other advanced degree, 53% of survey-takers had a bachelor’s or associate’s degree as their highest degree, and 19% of volunteers had at most a high school degree (this category includes individuals with some postsecondary education but no postsecondary degree). A plurality of respondents had moderate political views (46%) and a majority of respondents reported working in STEM-related fields (62%; Table 3). Corporate volunteers were more representative of the US average in terms of their race/ethnicity, gender and level of education than typical SciStarter volunteers (i.e., those not involved in the corporate volunteer program), who tend to be White, highly educated women (Table 3). In addition, while there is no existing data on the political beliefs of typical SciStarter volunteers specifically, these corporate volunteers’ political beliefs were more representative of the US average than those of citizen scientists from other domains, who tend to identify predominantly as Liberal [5,92]. Corporate volunteers were less representative of the US average than typical SciStarter volunteers in terms of their age and whether they worked in science-related fields, with these corporate volunteers less likely to be above retirement age and more likely to work in STEM.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Summary socio-demographic data for corporate volunteers, typical citizen science volunteers on SciStarter [5], and the general US population.

https://doi.org/10.1371/journal.pone.0331221.t003

Goal 3: Volunteer motivations

EFA indicated that the motivation items mapped to three factors rather than five, according to the Kaiser Criterion. One item from the proposed external regulation factor had a loading of only 0.3, below the threshold of 0.4 for interpretability established by Stevens [98], and didn’t load on any of the other factors, so we removed it from the analysis. All other items had factor loadings of at least 0.4. We repeated the EFA with the problematic item removed and the items again mapped onto three factors (Table 4). The first factor, “Autonomous Motivation” (Cronbach’s α = 0.91), included all three items relating to intrinsic regulation, all three items relating to identified regulation, and one item from introjected regulation. The second factor, “Controlled Motivation,” contained two items from the proposed introjected regulation factor and two items from external regulation (α = 0.70). The third factor, “Amotivation” (α = 0.67) contained all three remaining items relating to amotivated regulation.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Means and loadings across three factors for volunteers’ motivation to participate in citizen science (n = 388).

https://doi.org/10.1371/journal.pone.0331221.t004

Overall, responses to the quantitative motivation items indicated that corporate volunteers have relatively high levels of self-determined motivation to participate in citizen science. On the 0–10 scale used in the survey, scores for the Autonomous factor (high self-determination) were high (x̄ = 7.82), scores for the Controlled factor (low self-determination) were moderate (x̄ = 4.31) and scores for Amotivation were low (x̄ = 1.83; Table 4). The mean score on the RAI was 6.61 (range = 0–9.86).

Responses to the qualitative motivation question indicated a range of motivations. Some volunteers described highly intrinsic motivations, such as: “it seems interesting”, “Honestly, the study sounds pretty neat!” and “Enjoy doing it”. Likewise, many responses indicated that volunteers participated because doing so aligned with their internals values (i.e., identified regulation) such as: “Interested in helping scientists solve existential world problems” and “Looking for ways to volunteer and give back to my community”. Many of these volunteers identified specifically with the health-related or environment-related focus of many of the projects, for instance: “I joined the first study as part of a volunteer program through my employer and since I am a high risk individual, I felt i could contribute quite a bit to the study of the virus”, “Was interested due to the dementia part. I have family that suffer and suffered from it.”, and “Pollinators are important to our ecosystem. I want to help provide a read of the pollinators in my area.” A few volunteers said they participated for introjected reasons, such as one respondent who said, “My daughter is fascinated by science and I’m trying to model good behaviors of involvement for her.” Finally, some responses indicated highly external motivations for participating, such as: “It’s part of my volunteer day with work.”, “assigned to do so through my employer”, “for work requirement”, and “Getting credit at work for volunteering”. A number of volunteers indicated that multiple different motivations along the SDT motivational continuum were operating simultaneously, such as a respondent who said “Need to commit a specific set of hours today, and the project sounded interesting.”

The only socio-demographic variable that was significantly predictive of volunteers’ self-determined motivation was gender, with women scoring, on average, approximately 0.9 points higher on the 0–10 RAI scale (S1 Table).

Goal 4: Motivations and other determinants of depth of engagement in citizen science

Self-determined motivation was associated with more hours of participation in citizen science (Fig 2). Specifically, higher scores on the Autonomous Motivation factor (2a) and the RAI (2d) were associated with more participation, while higher scores on the Controlled motivation (2b) and Amotivation factors (2c) were associated with fewer hours of participation. For these analyses, we binned total hours of participation to increase readability owing to the right-skew of the variable.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Associations between volunteers’ motivations for participating in the citizen science program and the number of hours they spent contributing to projects.

Figure is separated into panels for autonomous (A), controlled (B), and amotivated (C) motivations. Panel (D) depicts the association between hours of participation and a “Relative Autonomy Index,” which combines the three motivation factors into a single measure of self-determined motivation. Vertical black line denotes median, blue dot denotes mean (n = 356).

https://doi.org/10.1371/journal.pone.0331221.g002

When controlling for socio-demographic variables, survey completion date, and volunteers’ prior citizen science experience, higher scores on the Controlled motivation factor had a significant association with fewer hours of participation in citizen science (Fig 3). Specifically, a one-point increase in mean controlled motivation, as measured on the 0–10 scale, predicted approximately 4% fewer hours of participation in the citizen science program. Thus, a volunteer who “completely agreed” that they participated in citizen science “because it’s what I’m supposed to do” (i.e., scored a “10” on this motivation item) and had similar scores for the other Controlled motivation items would be expected to spend 1.2 fewer hours contributing to citizen science than a volunteer who did not agree at all with these motivations (i.e., scored a “0” on these items) and participated for three hours. Higher scores on the Autonomous factor were associated with more participation in citizen science, but the relationship was not statistically significant. The Amotivation factor did not appear to have a relationship with the number of hours a volunteer spent contributing to citizen science.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Model of the effects of volunteer motivation on the number of hours spent contributing to citizen science (log-corrected).

Regression model controls for volunteers’ race/ethnicity, gender, education, political beliefs, age, whether they worked in STEM, when they completed the survey and whether they had prior citizen science experience. Figure shows 95% confidence intervals for the three motivation variables (vertical black lines). Degrees of freedom = 221; adjusted r² = 0.11.

https://doi.org/10.1371/journal.pone.0331221.g003

Two of the socio-demographic variables in the model also had a significant relationship with the number of hours a volunteer spent participating in citizen science. Specifically, Black volunteers participated in, on average, 25% fewer hours of citizen science than White volunteers, and a one-year increase in volunteer age was associated with an average 1% increase in the time volunteers spent participating, controlling for the other factors in the model. In addition, the 15% of volunteers with citizen science experience prior to taking the survey participated in 57% more citizen science than volunteers without this experience. Standardized beta coefficients (β) show that having prior citizen science experience was the strongest predictor of total hours of participation in the citizen science program, followed by age, controlled motivation, and identifying as Black (Table 5).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Results of regression predicting hours of citizen science contributions (log-corrected) from volunteers’ motivations, socio-demographics and other factors.

https://doi.org/10.1371/journal.pone.0331221.t005

Implications for practice

Facilitator organizations such as corporate volunteer programs offer a promising source of new citizen science volunteers who can provide important contributions to science, who have more to gain from their experiences (e.g., new experiences with science) than conventional self-selecting citizen scientists, and who may improve the practice of citizen science by bringing new, more diverse perspectives and approaches. However, achieving these goals requires sustained engagement in citizen science, something that our study found was relatively rare among a cohort of corporate volunteers. Project owners and facilitator organizations should engage in reflective practice to assess whether their projects are supporting self-determined motivation [107].

In cases where volunteer groups have low self-determination, steps should be taken to change this. According to Self-Determination Theory, people develop self-determined motivation when their needs are met for competence, relatedness and autonomy [50]. We suggest a number of tangible ways that projects might meet these needs in Table 6. Meeting volunteers’ need for competence, for instance, might be met through comprehensive trainings and opportunities to track progress and demonstrate efficacy [82,108]. Competence might be a particularly salient factor for facilitator group volunteers, who, as we demonstrated in this study, may have less experience with citizen science and may lack confidence in their abilities to participate in science-related activities. Gamification can be another way to meet volunteer needs for competence, such as using points, leaderboards and digital badges, however this can have unintended negative consequences. If volunteers’ primary motivation revolves around the project’s gamified elements, then their intrinsic motivation can actually be undermined by these extrinsic factors through the process of “crowding out” [109]. Used effectively, gamified and/or competitive elements of participation reinforce interest in and commitment to project goals rather than distract from them.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. Suggestions for designing citizen science projects to facilitate self-determined motivation by meeting volunteers’ needs for competence, relatedness and autonomy, using examples from popular projects among corporate volunteers.

https://doi.org/10.1371/journal.pone.0331221.t006

Volunteer needs for relatedness can be met by fostering volunteer community within projects such as through forums and other opportunities to interact with other volunteers and project scientists [110]. While many projects use reportbacks to communicate project results to participants (one-way communication), the need for relatedness is best met by opportunities to build relationships, such as through chat features, forums, in-person meetings and other forms of two-way communication. Relatedness needs might also be met by making projects more inclusive to people from different backgrounds, as discussed above, such that more volunteers see their identities reflected in other project volunteers and facilitators.

Meeting volunteer needs for autonomy can be met by providing volunteers access to project data, opportunities to follow their own scientific curiosities, and allowing for flexibility in how volunteers complete tasks and follow protocols [49,111]. Projects that rely on rigid protocols and that limit access to data are unlikely to meet volunteer needs for autonomy. Among those needs identified by Self-Determination Theory, the need for autonomy may be the least fulfilled by modern citizen science that often embraces a crowdsourced approach to data collection. While many historical citizen science projects involved dedicated amateurs pursuing their own curiosities through projects they created themselves (e.g., birdwatching enthusiasts initiating monitoring projects) most of today’s projects (and all the projects these corporate volunteers participated in) involve top-down protocols designed by and for professional scientists and scientific institutions [112]. These “contributory” style projects [113] inherently stifle some amount of volunteer autonomy, since volunteers are answering someone else’s research question rather than following their own curiosities or pursuing issues that are relevant to their lives. iNaturalist, which allows volunteers to create sub-projects based on local or taxonomic-based interests, as well as EteRNA, which poses previously unsolved challenges that volunteers solve creatively, provide some examples of how a few popular projects are fulfilling volunteers’ autonomy-related needs. Citizen science would better support volunteer autonomy if it served more effectively as a space where people can develop and test their own scientific questions, rather than solely contributing data to serve the research interests of professional scientific institutions. These suggestions should serve as a starting point for thinking more deeply about meeting volunteer needs to support self-determined participation in citizen science.