Building motivationally supportive course-based research experiences for undergraduates: a self-determination theory perspective

ABSTRACT

Many institutions have turned to course-based undergraduate research experiences (CUREs) to involve more students in authentic research. However, research is lacking on how to best support and nurture student motivation in CUREs. Motivation is a critical construct to understand at the undergraduate level as lack of motivation has been identified as contributory to lower achievement as well as reduced interests in STEM college degrees and careers. Using mixed methods and self-determination theory (SDT), the purpose of the current study was to determine how students’ basic psychological needs were supported or frustrated during CURE activities. Findings suggested that student motivation was potentially reduced during all CURE activities (i.e. scientific practices, collaboration, discovery, and iteration). Furthermore, students reported that autonomy was the least supported motivational construct during the CURE. Considering these challenges, the authors recommended specific strategies to increase choice, support competence, improve collaboration, inspire discovery, and create more opportunities for iteration during CUREs.

KEYWORDS:

• CURE
• undergraduate science education
• motivation
• SDT
• mixed methods

Introduction

Historical approaches to undergraduate education in the sciences have characteristically been associated with either a teaching or laboratory emphasis. Over time, some educators adopted a combination of teaching and laboratory experiences delivered in a research-based learning format (Bauer & Bennett, 2003). Customarily, these research-based learning experiences were conducted in an internship format. However, many institutions are limited in their ability to provide internships on a scale large enough to accommodate the majority of their students. Some institutions have consequently turned to course-based undergraduate research experiences (CUREs) as a way to involve more students in authentic science (Brownell, Kloser, Fukami, & Shavelson, 2012; Dolan & Weaver, 2021; Hanauer, Dolan, & Smith, 2014). CUREs engage more students in research-centered learning (Auchincloss et al., 2014), and this style is actively promoted by many organizations including the Howard Hughes Medical Institute.

Past research suggests CUREs can positively influence student achievement (Jordan et al., 2014), enhance feelings of ownership and identity in science (Cooper, Knope, Munstermann, & Brownell, 2020; Staub et al., 2016), promote continued engagement in science (Hanauer et al., 2017; National Academies of Sciences, Engineering, and Medicine, 2015), and improve achievement for historically underrepresented students (Ing, Burnette, Azzam, & Wessler, 2021). In some cases, CURE outcomes have been shown to rival those of research internships (Olivares-Donoso & González, 2019), considered by some to be the gold-standard of college science learning. However, in spite of these positive findings, research on CURE effectiveness has been called into question, as assessment approaches have been challenged for many reasons including but not limited to disagreements about how best to go about assessment (National Academies of Sciences, Engineering, and Medicine, 2015), challenges with creating uniform assessments across disciplines (Bhattacharyya et al., 2020), and lack of considering psychosocial perspectives (Auchincloss et al., 2014). Although there are emerging studies that consider how specific CURE activities (e.g. Corwin, Graham, Dolan, & Ledbetter, 2015; Corwin et al., 2018) and pedagogical practices (Hanauer et al., 2022) impact student outcomes, there is still a gap in terms of establishing links between CURE dimensions and student affective outcomes.

All CUREs are not created equally and often differ markedly from institution to institution (Staub et al., 2016). This fact makes assessment challenging. However, focusing on student motivation in general as well as satisfaction or frustration of basic psychological needs within the CURE dimensions can provide a more robust and consistent assessment tool, irrespective of CURE contexts. Equipped with this new knowledge, CURE developers can create experiences that consistently target student psychological needs and improve student motivation in STEM.

The purpose of the current study was to: (a) determine how specific activities within CURE dimensions supported or frustrated students’ motivational needs; and, (b) discuss specific strategies to facilitate motivational support, particularly in the most challenging areas as identified in this study. As lack of motivation has been recognized as one of the causal agents in failing to retain aspiring STEM majors (Cromley, Perez, & Kaplan, 2016), more voices are calling for studies investigating the links between CURE activities and motivation (Corwin et al., 2018; Dolan & Weaver, 2021).

Defining CURE activities

The Course-Based Undergraduate Research Experiences Network (CUREnet) defined CUREs as experiences including five specific dimensions (Auchincloss et al., 2014). In their recent book, Dolan and Weaver (2021) combined two of those dimensions and presented the following four activities as definitive of a CURE: (1) Using scientific practices (e.g. asking questions, proposing hypotheses, selecting methods, gathering and analyzing data); (2) Making discoveries (e.g. investigating questions with largely unknown outcomes and somewhat novel findings) that have relevance outside of the classroom; (3) Collaborating (e.g. students working with each other and with their instructors); and (4) Iteration (e.g. designing, testing/retesting, and revising research questions and experimental procedures). This structural framework has proved useful, as research begins to connect CURE activities, instructional models, student practices, and student outcomes (Corwin et al., 2015; Hanauer et al., 2022).

According to past research, not all CURE activities produce the same outcomes. For example, Corwin et al. (2018) reported iteration had a greater impact on students’ feeling of ownership about a project than did discovery. In other examples, even variations within a CURE activity (e.g. scientific practices) have impacted student outcomes. In a study involving students in the SEA-PHAGES program, researchers discovered that students who analyzed their own data showed greater gains in scientific identity and emotional ownership than students in the same program who analyzed data collected by professionals (Cooper et al., 2020). Furthermore, the importance of pedagogical support provided by instructors has been shown to affect CURE outcomes. For example, Hanauer et al. (2022) emphasized the importance of instructors supporting students’ independence and fostering a sense of project ownership in order to successfully facilitate CUREs.

Basic needs satisfaction in CUREs

Despite that fact that many of the previously-mentioned studies reported results not explicitly tied to motivation, authors like Corwin et al. (2018) called for a deeper look at the role of student motivation and expressed how motivation could be used to better explain CURE effectiveness. Stakeholders in undergraduate education should not assume all occurrences of ‘doing science’ are equally motivating. In fact, many students are turned off by science learning (Smith, Deemer, Thoman, & Zazworsky, 2014), including many high-achieving students who sometimes find introductory science classes to be uninspiring and boring (Walker, Sampson, Southerland, & Enderle, 2016). It is not productive to assume intrinsic motivation is enough to draw and retain students to the STEM fields; it is equally not productive to assume students will find the activities associated with CUREs to be motivating. Instead, research is needed to explore the links between student motivation, basic needs satisfaction, and specific CURE learning activities.

Motivational theories become useful in this regard, as filtering findings through a theoretical lens can illuminate key factors (Miles, Huberman, & Saldaña, 2014). One such theory is self-determination theory (SDT; Deci & Ryan, 1985), which is one of the most used and empirically-verified motivational theories in education. In education contexts, SDT can help educators identify factors that enhance and/or diminish student motivation (Deci, Vallerand, Pelletier, & Ryan, 1991). SDT theorizes humans are active organisms with basic psychological needs including autonomy, competence, relatedness. According to basic needs theory, a sub-theory of SDT, when these needs are satisfied, a person experiences health and well-being (Deci & Ryan, 2000). Although studies involving motivation and/or SDT have been conducted in college science contexts, they have not typically: (a) been applied specifically to CUREs, and/or (b) focused on basic psychological needs satisfaction.

For example, Black and Deci (2000) conducted an early SDT-based study of a non-CURE organic chemistry course using several motivation surveys. Their findings highlighted the important link between students’ feelings of autonomy and instructor autonomy support on student outcomes such as grades. Specifically, this study revealed that instructional styles low in autonomy support were ‘likely to be related to students’ feeling bad, and possibly to performing badly’ (p. 754). Furthermore, the authors suggested that ‘more autonomy-supportive instruction would likely be helpful in other college-level natural science courses’ (p. 755).

Since then, studies linking motivation and CUREs have emerged. For example, Olimpo, Fisher, Dechenne-Peters, and Hatfull (2016) study of an introductory cellular/molecular biology CURE used the Biology Motivation Questionnaire (BMQ; Glynn, Brickman, Armstrong, & Taasoobshirazi, 2011) and open-ended responses to investigate links between CUREs and motivation. Results indicated CURE students exhibited more positive shifts in self-determination, self-efficacy, and intrinsic motivation than students in non-CURE courses. Furthermore, 73% of students shared that autonomy was linked to their positive feelings about the CURE. Similarly, Esparza, Wagler, Olimpo, and Hatfull (2020) used the BMQ (Glynn et al., 2011) and found that CURE-based sections of introductory cellular/molecular biology laboratory courses were more likely than non-CURE sections to enhance students’ motivation toward a career in biology. Although SDT was not cited in these studies, the findings suggested that the theory could have been useful for revealing more about why these CUREs led to these specific outcomes.

Other studies have been more explicit about how SDT could improve understanding of CURE outcomes. Using interviews and essays, Wooten, Coble, Puckett, and Rector (2018) reported that CUREs increased students’ self-efficacy and motivation. Recognizing that the trends in their qualitative data corresponded to SDT constructs, these authors used SDT as an interpretive framework to provide a deeper analysis of their findings. Upon mapping their qualitative findings to SDT, Wooten et al. suggested that students in the CURE experienced some basic needs satisfaction (although the authors admittedly ‘did not directly assess students’ motivation’ (p. 14)). Similarly, Gin et al. (2018) compared high-challenge and low-challenge seafood forensics CUREs and discovered themes like autonomy, self-efficacy (i.e. competence), and sense of belonging (i.e. relatedness) from their qualitative analyses. These authors suggested that, ‘Future research could further leverage self-determination theory … to examine how students deal with scientific obstacles during and after CURE … participation’ (p. 17).

In sum, these past studies show a progression toward using SDT as a framework in CURE studies. The current study incorporated SDT throughout the research process, from design to interpretation of results. In doing so, this research provides a valuable contribution to the CURE literature by explicitly linking CURE dimensions to basic needs support and/or frustration. Making the connection between students’ psychological needs and CURE activities is critical because environmental conditions (e.g. CURE activities) are essential to understanding how people perceive satisfaction and/or frustration of their needs (Jeno et al., 2017). For example, when environmental conditions allow people to regulate and control their own behavior through purposeful choice, people feel autonomous and more motivated. When conditions permit people to engage in challenging tasks and still feel somewhat effective, competence is supported and motivation increases. Finally, conditions that facilitate people connecting with each other support relatedness (Ryan & Deci, 2017) and subsequently, motivation. Supporting autonomy, competence, and relatedness leads to increased motivation for all who operate within that environment (Deci & Ryan, 2002).

Therefore, in cases where students may not be intrinsically motivated by STEM (i.e. engage in STEM out of joy), structuring CURE activities to support rather than frustrate the basic needs (Cromley et al., 2016) can be a way to appeal to a broader spectrum of students. Moreover, focusing on basic needs is a way to generalize CURE assessment and link specific CURE activities to motivation. The purpose of the current study was to determine how specific CURE activities satisfied or frustrated students’ motivational needs. Based on this information, specific strategies are recommended to better support student motivational needs during CURE activities.

The current study

The sample institution in this study was a research-intensive, PUI liberal arts college in the Midwest USA. Total enrollment typically exceeds 3,000 students, with about 40% of students enrolling in an introductory STEM course during their first year and 23% ultimately majoring in the sciences. The course in this study was a first-semester, general biology class typically taken by prospective science majors/minors and students with pre-health interests. The course followed a traditional lecture/lab schedule, but the laboratory portion was restructured as a CURE. Laboratory sections met once per week for three hours in a combination of field and laboratory settings. In terms of classification, the CURE followed the ‘modular’ definition of Mader et al. (2017), as it was a sequence of shorter research experiences within a single laboratory course. The CURE consisted of a series of related experiments organized into three modules focused on ecology, plants, and animals.

During the ecology module, students investigated processes that contributed to forest functions (e.g. tree growth, herbivory, succession, decomposition). Students worked in teams to gather data by identifying tree species, measuring annual growth rates for juvenile trees (via twig analysis), and estimating herbivory damage on leaves using grids. Students also measured decomposition rates of leaves using litter bags (placed in the forest the semester before). Invertebrates from loose leaf litter on the forest floor were identified to morphospecies, with students also calculating various measures of species richness and diversity.

In the plant module, students investigated the impact of elevated CO₂ on plant growth, physiology, and resource allocation (see Muilenburg et al., 2018). Specifically, students examined leaf stomata and calculated stomatal density as they compared Capsicum annuum (pepper plants) grown in elevated CO₂ conditions to those grown at ambient CO₂ concentration. Students also determined the photosynthesis and respiration rates of plants grown in elevated and ambient CO₂ conditions using growth chambers and probes, ultimately comparing the two CO₂ conditions using statistical techniques. Finally, students extracted foliar protein from leaves and measured total plant growth and the root/shoot biomass to determine how CO₂ conditions affected growth and resource allocation in plants from the two conditions.

In the animal module, students considered how the environment affected the physiology and behavior of animals. In a series of experiments, students used fish to determine the potential effects of fluoxetine on circulatory function, respiratory rates, and display of behavior. All experiments were conducted in a blind fashion, meaning students did not know which tanks were treated and which were not. During the experiment, students measured swimming activity, aggression, and opercula flaring. The activities of fish exposed to fluoxetine and those that were not were compared using appropriate statistical measures.

Throughout these modules, students did the work of scientists by keeping lab notebooks, reading primary literature, analyzing data using both descriptive and inferential statistics, and using tools such as taxonomic keys and compound light microscopes. They also worked collaboratively and created final products including reports and research posters.

Methods

Researchers used mixed methods to analyze student motivation as a result of participating in the introductory CURE. A convergent mixed methods design was used as it offered a balanced and pragmatic approach (Creswell & Plano Clark, 2011). Using a quantitative survey allowed for a larger sample group while the qualitative feedback provided detail and depth not possible using survey data alone. In general, researchers found this approach useful for uncovering the complex interactions in this educational context (Butler, 2002).

In addition to using mixed methods, CURE activities were analyzed through the lens of SDT to determine how motivation was enhanced or diminished. Quantitative data from an SDT-based motivation survey provided important feedback on how basic psychological needs were being met in relation to each other. Qualitative data from interviews were useful for extracting rich, meaningful contexts related to how CURE activities supported or frustrated autonomy, competence, and relatedness. Although quantitative and qualitative data were both considered, qualitative data were given priority (a practice supported by Creswell & Plano Clark, 2011) in the final analysis as they were best suited for linking the basic needs to specific CURE activities. Finally, it is important to note that researchers did not control or manipulate the conditions in the course, nor participants’ behaviors or actions.

Positionality and trustworthiness

The research team consisted of two professors and two undergraduate research students. One of the professors (PI) had a half-time appointment in the biology department at the institution. However, there was no conflict of interest as this professor did not teach biology majors/minors at the time and had no direct contact with students in the course outside of this study. The other professor (co-PI) had no biology department affiliation nor contact with the students in the course. Neither of the student researchers were biology majors or minors. One of them had taken the introductory biology course before, but in a different semester.

All researchers were trained in human subjects research and appropriate methodology before participating in the project, with methods and participant recruitment protocols being approved by the IRB of the institution. The PI’s role in the project included setting project priorities, training student researchers, scheduling participant interviews, conducting interviews, administering surveys, reviewing literature, coding/interpretation, and writing. The co-PI was involved in literature review and writing. Research student roles included analyzing survey data, transcribing interviews, coding, interpreting data, reviewing literature, and writing. As the PI and student researchers coded, they often met together to negotiate agreement regarding categorization (researcher triangulation per Johnson, 1997). This triangulation was complemented with reflexivity, the practice of each researcher openly discussing preconceptions, assumptions, and systematic biases (Johnson, 1997). During the coding and categorization processes, data were verified using constant comparison (Boeije, 2002). Researchers also recorded extensive notes and memos (Strauss & Corbin, 1990). Finally, as initial categories were established, researchers made predictions about how new data would fit the model and tested those predictions with the new data (i.e. pattern matching; Johnson, 1997).

Participants

An exemption for this study was granted by the Hope College Human Subjects Review Board (HSRB) under U.S. Federal Common Rule 45 CFR 46.101(b)(1) and 45 CFR 46.101(b)(2). Rigorous informed consent procedures were still used, and all participants were made aware of study details and their rights as volunteers. All participants gave consent before completing the surveys or engaging in the interviews or focus groups. The PI solicited participants in the early days of a fall semester. Emails were sent to all enrolled students, and the PI made personal solicitations during several class meetings. A total of 111 students volunteered for the quantitative portion of the study (approximately 55% of course enrollment). For the qualitative portion, 18 students volunteered (9 for interviews, 9 for a focus group; approximately 9% of course enrollment). Finally, two teaching assistants (TAs) participated in interviews. Table 1 summarizes participant characteristics.

Table 1. Participant information for quantitative and qualitative portions of study.

Participants	Gender Identification	Classification
Survey Respondents (n = 111)	65 Females 46 Males	82 Freshmen 19 Sophomores 7 Juniors 3 Seniors
Interviewees (n = 9)	8 Females (referenced as Females A-H) 1 Male (referenced as Male A)	6 Freshmen 1 Sophomore 2 Juniors
Focus Group (n = 9)	7 Females 2 Males	6 Freshmen 3 Sophomores
Teaching Assistants (n = 2)	1 Female 1 Male	2 Seniors

Data collection

All data were collected post-course (within the last two weeks of classes).

Quantitative survey

Beck, Butler, da Silva, and Sundberg (2014) reported in a review article that only 12% of studies including some sort of assessment of inquiry-based learning at the undergraduate science level (including CUREs) used validated and/or published assessment instruments. In this study, researchers intentionally selected the Basic Psychological Need Satisfaction at Work Survey (BPNSS), a well-published and validated instrument (Deci et al., 2001). To fit the context, slight modifications were made to some items by adding phrases like ‘in lab’ (e.g. the original item, ‘I really like the people I work with’ was modified to, ‘I really like the people I work with in lab’) or replacing an equivalent phrase with the phrase ‘in lab’ (e.g. the original item, ‘People at work care about me’ was modified to, ‘People in lab care about me’). Although more general basic needs surveys had been used in schools in the past (see Tian et al., 2014), researchers determined the wording of the BPNSS work version required less alteration and was less likely to confuse participants. Moreover, other studies such as Herman, Goldberg, Trenshaw, Somerville, and Stolk (2017) have successfully used items from the BPNSS work survey in education contexts.

The BPNSS was administered via Google Forms and measured satisfaction of students’ basic psychological needs as a result of the CURE. The instrument had 21 items, and participants responded using a 7-point Likert scale (1 – Not at All true; 7 – Very True). Appropriate items were reverse-scored, and constructs were combined into measurements for autonomy (7 items), competence (6 items), and relatedness (8 items). Cronbach’s alphas indicated reliable survey responses: autonomy (α = .76); competence (α = .69); relatedness (α = .80).

Qualitative data

In an effort to accommodate as many students as possible and to make all participants comfortable, opportunities were extended for one-on-one interviews or group meetings. Nine (9) participants volunteered to engage in an individual, semi-structured interview with the PI. Interview questions related to: (a) surprising encounters, (b) if (or how) the experience was different than other science experiences, (c) what impacted students’ attitudes about the experience, (d) how independent and/or confident students felt, and (e) the most useful parts of the CURE. Interview times ranged from just under 15 minutes to 33 minutes. To gain additional context into CURE implementation, two (2) teaching assistants from the course also volunteered for interviews. Both TAs were upper classmen who had already gone through the course, although the course was somewhat different because many CURE components were implemented later.

In addition, nine (9) other students (different than interviewees) participated in a small focus group. Focus group participants were asked the same questions as the interviewees. This session lasted approximately one hour, and all participants were involved in the conversation at some point. In total, 4.5 audio-hours of conversations from interviews and the focus group were transcribed verbatim.

Data analysis

Descriptive statistics were computed using SPSS 25. Researchers compared BPNSS scores for autonomy, competence, and relatedness to past studies involving introductory biology students at the institution (see Scogin et al., 2020). Qualitative analysis followed processes outlined by Miles et al. (2014). During first-cycle coding, researchers identified all phrases related to like or dislike of the CURE (Figure 1; see A1 for first-cycle coding) and the reason(s) behind the response (Figure 1; see A2). This information was second-cycle coded in a deductive fashion by case in two different ways: (a) first, reasons for like or dislike were coded into the construct categories of autonomy, competence, relatedness, or other (Figure 1; see B1 for construct coding); (b) second, reasons were related to a specific CURE activity (Figure 1; see B2 for activity coding), as applicable. Next, a descriptive meta-matrix was built to show the relationships between satisfaction/frustration of the basic psychological needs and CURE activities.

Figure 1. Brief overview of coding process (with examples) using the following participant excerpt.

Findings

Table 2 results indicate that overall, students in the current study felt more related than competent and more competent than autonomous when engaging in the CURE. These findings were similar to past CURE research conducted in introductory biology courses at the institution in that relatedness scores were always highest, followed by competence and then autonomy (Scogin et al., 2020). All construct scores in the current study were lower than in past studies. Researchers attributed this to the general nature of the course and the student population in this particular study. In past studies, the CUREs were either highly specialized to fit an active and ongoing research agenda (which could have enhanced motivation) or were more selective about the students who qualified to take the course (i.e. these students were geared more toward STEM).

Table 2. Means and Standard Deviations for Basic Needs Satisfaction Survey (BPNSS) for Current Study (n = 111) and Previous CURE Studies Conducted at Same Institution But in Different Courses.

	Current Study		Phage (see Scogin et al., 2020)		Watershed (see Scogin et al., 2020)
Psychological Need	M	SD	M	SD	M	SD
Autonomy	4.22	0.95	5.34	0.57	4.50	0.72
Competence	4.82	0.92	5.73	0.72	5.10	0.63
Relatedness	5.30	0.95	6.05	0.67	6.10	0.64

Notes. Means represent averages of constructs compiled from BPNSS survey (7-point Likert responses with 7 = Very Much and 1 = Not at All).

Table 3 includes summaries from coding about why students felt their needs were satisfied or frustrated as they engaged in CURE activities. During all four CURE activities, students shared ways in which autonomy and competence were sometimes frustrated. In contrast, relatedness was only frustrated during Collaboration (Table 3). Similarly, relatedness was the highest construct from students’ perspectives (Table 2).

Table 3. Descriptive meta-matrix showing relationships between CURE Activities and Satisfaction or Frustration of Students’ Basic Psychological Needs.

	Basic Psychological Need
CURE Activity	Autonomy		Competence		Relatedness
Scientific Practices	Satisfying Some choice in which practices to engage in as they divided workload	Frustrating Little choice in overall project direction	Satisfying Gained basic competence related to quantitative thinking, lab reporting, lab skills, etc.	Frustrating Lacked competence to drive project direction	Satisfying Worked in teams and made connections with others in course	Frustrating [No codes]
Collaboration	Satisfying Some choice in checking work with others and dividing tasks and workload	Frustrating Sometimes felt forced to collaborate	Satisfying Gained some competence from group sharing	Frustrating Competence sometimes waned because of pressure from other group members	Satisfying Worked in teams with like-minded people that sparked excitement	Frustrating Connections sometimes disintegrated due to disagreements
Discovery with Broader Relevance	Satisfying In labs with unknown outcomes, students empowered to find answers	Frustrating In labs with known outcomes, students felt controlled toward an expected end	Satisfying Unexpected results provoked thought and challenged students; Experiences contributed to knowledge of big-picture issues and phenomena	Frustrating Students not challenged by predictable results; Sloppy results led to limited broader application and waning feelings of competence	Satisfying [No codes]	Frustrating [No codes]
Iteration	Satisfying [No codes]	Frustrating Students felt pressured to get answers in a hurry	Satisfying Some competence gained from considering ways to improve experiments	Frustrating Not being able to redo experiments hampered student competence	Satisfying [No codes]	Frustrating [No codes]

Note. Direct quotes supporting many of the above summary statements are located in the main text.

Results in Table 3 also revealed 3 of 3 basic needs were sometimes frustrated during Collaboration activities, while students reported only two needs were frustrated during Scientific Practices, Iteration, and Discovery (2 of 3). In terms of satisfying basic needs, students reported ways in which all basic needs were satisfied by Scientific Practices and Collaboration (3 of 3). The Discovery process often satisfied 2 of 3 needs, and Iteration satisfied 1 of 3. In order to better understand how basic needs satisfaction or frustration was perceived by students during the CURE, the following section is organized by CURE activities.

Scientific practices

Students felt they had some choice about which scientific practices to engage in as they completed lab activities (Table 3). Furthermore, all students reported increased competence in at least one scientific practice (e.g. conducting statistical analysis). However, students did not always feel in control of the project, primarily due to a lack of empowerment to generate research questions or methods on their own. ‘We never had the opportunity to … devise an experiment’ (Female B). Also, students recognized they lacked competence in some other critical areas, and this limited their autonomy: ‘We didn’t work at all on how to develop our own experiments … Which at this point makes sense, because we don’t have that much experience’ (Female H). Quantitative data (Table 2) also reflected that autonomy satisfaction was lower than the other motivational constructs.

Collaboration

Seven of the nine students who were interviewed reported how working as a team was important to their success in the course (along with the majority of the focus group students). ‘[I] definitely learned to work with other people. That was a big thing in lab, working as a team … ’ (focus group). In some cases, collaboration satisfied basic motivational needs. For example, collaborating to verify responses bolstered students’ feelings of competence. ‘I like [working in groups] because … you can check with your friends and make sure you got the same … like if you did the [statistics] wrong, you wanted to make sure’ (Female C). Some connected with their group on deeper levels, and this made the course more exciting. ‘This was also the first time that I’ve ever been with a group of people who are all excited about science’ (focus group).

In other cases, motivation was diminished as students felt they were forced to collaborate in order to meet course expectations: ‘[Collaboration] is what we are supposed to be doing’ (Female H). Under these controlling conditions, motivation sometimes decreased: ‘I don’t like [what my group] did, and I just want to fix it’ (focus group). In rare cases, collaboration completely failed. ‘It didn’t really work with my group because we were all different, and we all would … not agree. And then we ended up agreeing just to be done with it’ (Female B).

Discovery

Six of eight interviewed students and the majority of focus group students felt they were engaged in discovery learning, as they (and often their instructors) did not know the final outcomes. In fact, some participants expected to experience discovery. ‘I feel like in science, you’re not really supposed to know. Normally you don’t know’ (Female C). Participants felt most like scientists and were highly motivated when they were investigating unpredictable outcomes. ‘When we got unexpected results, we were excited!’ (focus group); ‘I mean, when [the instructor] was surprised a couple times with our results, that was cool … . That kind of provoked thought, so that was a good thing’ (Female A); ‘I just like the fact that they don’t know the results, and we don’t know either. So that makes it more exciting … ’ (Male A).

On the other hand, participants sometimes easily predicted outcomes: ‘ … you kind of can predict which result you’re gonna get’ (Female F). When results revealed ‘nothing new’ (Female E) or ‘nothing truly surprising’ (Female H), students lacked excitement and inquisitiveness. They also felt controlled toward an expected end and were not challenged when ‘getting the right answer’ seemed to be the logical end.

In terms of broader relevance, students were split on how much of their work applied to a larger audience. For example, Female A talked about how she appreciated ‘looking at the bigger picture … that’s science, and I like seeing the big picture with all the little interactions within it.’ Other participants saw limited relevance to their work, often because of careless lab habits (often attributed to others in their group). ‘To be honest, if someone wanted to publish our results as a class from experiments we have done, I don’t think anyone would take them very seriously’ (Female F). One TA reinforced this notion and provided additional reasoning: ‘And potentially, if we’re going to call this a research-based lab, we should be able to publish these … data in scientific journals. But that won’t be happening, because students don’t adhere to very strict protocols consistently.’

Iteration

Participants expected iteration to be one of the differentiating factors between a research experience and a traditional lab activity. ‘I think scientific research, more often [than classroom activities], has … more tries … So I feel like research is more like, fail, try again, fail, try again’ (Female C). One participant (Female D) reported that thinking about iteration was motivating: ‘[Thinking about redoing the experiment] made me want to do it again, but the time was out so we just left it.’

Students often felt they were not allowed to try, fail, critique, and retry because of time constraints. In fact, some students, like Female B, felt pressured to just collect enough data to analyze. ‘I still just feel like we would look at something, and that would be the end of it. Just like, off to the side kind of, let’s move on, let’s get these data, and we’ll just collect it. You’ll do a t-test, and we’ll move on.’ TAs also recognized how the experience sometimes fell short in terms of iteration. ‘No real-world research experience is going to have you do one experiment, and then that’s it.’

Discussion

As SDT posits motivation occurs on a continuum and is not static, making appropriate instructional adjustments to areas that diminish motivation can increase overall motivation in an educational setting (Ryan & Deci, 2017). Similar to the concept of scaffolding when teachers provide learners with cognitive supports so they can handle challenges currently above their academic level (Puntambekar & Hubscher, 2005), SDT predicts supporting basic psychological needs enhances student motivation. Consequently, the following discussion is based on the major threats to motivation uncovered in this study. Specifically, the following strategies focus on increasing student choice, supporting competence, improving collaboration, inspiring discovery and relevance, and creating more opportunities for iteration (Table 4). Some suggested strategies target specific basic needs, while others relate more generally to CURE activities.

Table 4. Strategies to support and satisfy students’ basic motivational needs during CUREs.

Goal	Strategies
Increase choice and support competence	Structure CURE such that small groups work on different parts of the project Facilitate interdependent tasks Scaffold with primary literature Build CUREs on longitudinal research
Improve collaboration	Purposefully create spaces for collaboration (structured spaciousness; Herman et al., 2017) Facilitate interdependent tasks Leverage distributed expertise within group dynamics
Inspire discovery and relevance	Connect CUREs to existing research agendas Link CUREs to local issues
Create opportunities for iteration	Alter format of CURE Cut material and focus on replication

Increasing choice and supporting competence

Quantitative data indicated autonomy and competence were the most threatened basic needs in this CURE (Table 2). Qualitative data confirmed that students felt frustration of autonomy and competence during every scientific practice (Table 3). In many cases, these two needs were entangled and therefore difficult to discuss in isolation from one another. For example, students often felt they were not allowed to make more choices because their knowledge was limited. The literature contains excellent suggestions on how to navigate situations when students face challenges due to inexperience. For example, instructors or graduate students could do some of the more difficult analyses and/or use the complex equipment, then report the findings back to CURE students (Dolan & Weaver, 2021). However, these approaches could further undermine autonomy as students do not have the choice to conduct the analyses themselves. So, how can purposeful choice be preserved (or increased) when competence is a limiting factor? This is a challenging conundrum, as students who are given too much choice without feeling competent can actually become less motivated due to feeling overwhelmed (Katz & Assor, 2007). Therefore, strategies should scaffold competence while gradually increasing autonomy.

Strategies to achieve this balance include: (a) structuring work so each group focuses on slightly different parts of a project, as opposed to all groups working synchronously on the same parts; (b) providing interdependent tasks within each group; (c) scaffolding student understanding with literature; and, (d) building CUREs around on-going research (i.e. longitudinal approach). It may seem odd to begin a discussion about improving autonomy by talking about adding structure and promoting group work as both of these strategies seem contradictory to autonomy support. However, past SDT-based research conducted in high schools has shown that teachers who provide supportive structure do not necessarily infringe upon student autonomy. If structure is defined as outlining clear expectations and delivering explicit direction and guidance to students, Jang, Reeve, and Deci (2010) revealed that teachers can provide structure yet remain supportive of students’ autonomy. These authors discovered that moderate structure (i.e. ‘some supervision but not totalitarian supervision’; p. 590) actually increased students’ feelings of competence as they learned required skills; and, this newfound competence helped students feel they had the agency to reach the instructional goals (i.e. made them feel more autonomous).

In terms of promoting group work, it is important to mention that autonomy is not the same as independence, as autonomy means volitional action, not being separate from others (Deci & Ryan, 2000). This distinction is important because all people need autonomy to some degree (according to SDT), but the specific conditions fulfilling autonomy differ (Chen et al., 2015). For example, some might feel more autonomous when following the directives of another if the rationale for completing the directive makes sense, while others may always perceive a directive as controlling and therefore not supporting autonomy (Soenens et al., 2007).

A specific strategy to increase choice while appropriately supporting competence is to have groups work on disparate parts of a project. Students could work on the same research question, with each group being responsible for a different part of the study (e.g. methods, analysis, interpretation, etc.). This format could potentially allow individuals within groups to exercise more purposeful choice as the number of students involved in that stage decreases. Furthermore, as each group is responsible for a reduced portion of the project, this method also scaffolds student competence. This arrangement could also increase feelings of ownership, which is a desirable outcome for CUREs (Cooper et al., 2020; Corwin et al., 2015) because ownership can promote more autonomy and increased motivation (Ryan & Deci, 2017).

A second strategy to promote autonomy and competence is to break up the work within groups into interdependent tasks (Langfred, 2000). Interdependent tasks are those requiring interactions from all group members (Shea & Guzzo, 1987). According to Langfred, working on interdependent tasks is important because if people are given autonomy but the tasks do not require true collaboration, they will often resort to doing things independently. In contrast, interdependent tasks require group members to defend their ideas and choices as they negotiate decisions within the group. Through this negotiation process (which promotes autonomy), feelings of competence also increase.

A third way to support competence is by expanding how literature is used in introductory experiences. Reading and evaluating literature is a common CURE practice (Corwin et al., 2015), but introductory courses sometimes limit literature to explaining and/or confirming findings from lab activities. While this skill is important and should continue, allowing students to expand how they use literature could be useful. For example, instructors can provide articles from which students could choose different approaches for a research question. These studies would scaffold students in research design and allow them to exercise choice in experimental methods. A previous study revealed using literature in this way was challenging for introductory biology students, but with properly-selected articles and guidance, modest gains were documented (Krontiris-Litowitz, 2013).

A final strategy is allowing students to build upon research conducted by previous students in past semesters (as opposed to starting from scratch every semester). This strategy provides a more authentic research context, encourages students to choose which direction to take a project, and perhaps increases feelings of competence as students work on research that they know was done by previous students of their own level.

Improving collaboration

Although quantitative data indicated relatedness was the most supported basic need (Table 2), qualitative findings indicated forced collaboration sometimes frustrated student autonomy (Table 3). Feeling forced to connect with others and/or engaging in collaboration strictly for self-gain (e.g. for better grades) is not as motivating as collaborating out of shared responsibilities (Ryan & Deci, 2017).

In their work describing the Intrinsic-Motivation Course Design Method, Herman et al. (2017) called for course designers to create ‘structured spaciousness.’ The concept of structured spaciousness, in part, combines ideas about clearly communicating purpose and expectations to learners while purposely providing space for them to cultivate relationships within their respective learning communities. The context of the Herman et al. paper was engineering, where carving out collaborative space was particularly important for developing students’ collaborative problem-solving skills (as engineers do this often). Likewise, in science, it is imperative for students to have opportunities to build collaborative skills as scientists increasingly work in diverse teams to tackle complex problems (Auchincloss et al., 2014). As SDT posits humans seek quality connections between one another (Deci et al., 1991), purposefully creating collaborative space within a course is an important motivational support.

A previously-mentioned strategy that could increase authentic collaboration and better support relatedness is providing groups with interdependent tasks. Interdependent tasks encourage students to share their strengths, a concept related to distributed expertise (Brown et al., 1993). Distributed expertise occurs when students with strengths in one area scaffold students who are not as strong in that area. In practice, students share necessary skills and knowledge to push the project forward. This is in contrast to what some students reported in this study, which was collaboration to ‘divide and conquer’ workload.

Inspiring discovery and relevance

Discovery is challenging to simulate at an introductory level. One previously identified reason is because introductory students are limited in their experiences and often lack the general knowledge and skills to confront novel situations. One well-used strategy to promote more discovery is connecting CUREs with existing research agendas. SEA-PHAGES (Jordan et al., 2014) is an example program that puts students in a position to make authentic, discovery-based contributions. Protocols and procedures are well-established, and students can readily see how others at the same academic level have contributed to the project in the past.

However, institutions do not necessarily have to go this route to facilitate discovery. They can, instead, make connections to authentic work going on at the institution and/or potentially link CUREs to local issues so students can become intimately familiar with the research and its applications (Staub et al., 2016). Furthermore, this arrangement makes communicating with the broader community more natural and provides a renewed perspective on the relevance of the work. This arrangement could also motivate students to follow protocols and procedures more carefully, as they know the direct context of their work and its potential implications.

Creating opportunities for iteration

Time pressure can be one of the biggest challenges for CUREs, and it can be debilitating to intrinsic motivation in education contexts (Niemiec & Ryan, 2009). Managing time in CUREs is particularly challenging as course schedules are often less flexible than research internships. In the current study, students and TAs felt the lack of iteration at times made the experience less than authentic.

When providing time for iteration is challenging, institutions could consider changing the modality of the CURE. Some CUREs are designed to span an entire semester, while others are embedded periodically around more traditional laboratory content (see Mader et al., 2017, for a categorization of CUREs based on timing and extent). In certain contexts, changing the modality of a CURE may provide more time for iteration. In other cases, instructors may decide to cut some material and instead, provide time for students to problem-solve and rethink previous methods and/or outcomes. The iteration sequence could include steps that require students to get conclusive results at certain points before progressing to the next steps. In general, these changes could facilitate a more authentic scientific context, as well as increase students’ competence and feelings of ownership (Staub et al., 2016).

Conclusion and limitations

This study applied SDT to students’ perceptions of an introductory CURE to determine how specific activities satisfied or frustrated motivational needs. Using these findings, researchers discussed specific strategies to increase motivational support in the most challenging areas. The strategies focused on promoting autonomy, bolstering competence, fostering stronger relatedness, and increasing authenticity to create more motivationally supportive CURE environments.

The authors recognize this study is limited by its use of volunteer participants and no control group. Furthermore, none of the findings should be interpreted as causal, and the suggested strategies do not warrant extension to every institution and context. In some cases, the strategies may need modifications. It is also important to note the strategies are not necessarily limited to introductory CUREs. To the contrary, variations of these strategies could be used to support motivation at multiple course levels. Finally, as past studies show CUREs have similar characteristics across many kinds of institutions (see DiBartolo et al., 2016; Mader et al., 2017; Staub et al., 2016), the authors hope this paper sparks the creativity of stakeholders at varied institutions to make their CUREs more motivationally supportive.

Acknowledgments

Special thanks to Dr. Vanessa Muilenburg, Assistant Professor of Biology Instruction at Hope College, for reviewing this manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Howard Hughes Medical Institute.