Introduction

The COVID pandemic forced most college courses online in 2020 which left many instructors and researchers wondering if this mode of instruction would be effective across the board. While there is a wide body of literature demonstrating the efficacy of distance learning (Allen et al., 2016; Chen et al., 2018; Choe et al., 2019; Yang, 2017), the rapid transition to online education in Spring 2020 presented several challenges and opportunities to STEM educators around the world. Courses featuring traditional wet-lab methods creatively met that challenge through the use of videos (e.g., Breslyn & Green, 2022), prepared kits (e.g., Njoki, 2020), or virtual simulations (e.g., Delgado et al., 2021, Alvarez, 2021). Courses featuring field experiences encouraged students to explore environments close to home (e.g., Gya & Bjune, 2021; Race et al., 2021), use citizen science applications (e.g., Lichti et al., 2021), and participate in virtual field trips (such as those from Teach the Earth and Arizona State University Virtual Field Trips). Recent studies demonstrated that students continued to make academic gains in online biology courses throughout the COVID transition and some studies even reported increases in students’ desire to learn biology (e.g., Gibson & Shelton, 2021; Sandrone et al., 2021; Supriya et al., 2021).

In Fall 2020, a novel course-based undergraduate research experience (CURE) was taught synchronously online at Pima Community College (PCC) in Tucson, Arizona, in partnership with the Sonoran Desert Museum and the University of Arizona (U of A). Like most classes at that time, the modality was constrained to be entirely remote. Thus, the class, which was originally intended to be in-person with a heavy wet-lab component, was online only. In the class, introductory biology students conducted bee DNA barcoding of samples collected by local museum volunteers. Students watched videos of their instructor performing the wet lab components of this research experience instead of doing any wet-lab work and then focused their engagement with the research on the bioinformatics needed to identify their bees. Students published the DNA sequences on the public Barcode of Life Database (BOLD) to be used by other researchers and reported their results to the community in an end-of-term poster session. This CURE provided the opportunity for PCC students to interact with community volunteers, students, and researchers at the U of A, and students within their class, all online. Our qualitative study (Creswell & Poth, 2018) seeks to understand the impact of online instruction in this laboratory-based CURE. We ask:

  1. 1.

    Did students form productive relationships with instructors, peer mentors, and peers in this online CURE course? (RQ1)

  2. 2.

    Did students develop science research self-efficacy in this online CURE course? (RQ2)

Online Laboratory Education and Student Engagement

Studies conducted prior to the pandemic explored conditions under which online laboratory learning is effective and which demographic groups of students choose online modalities. Online laboratory classes have positive outcomes (i.e., higher GPA, increased access, improved learning) when there is a practical component, when they are inquiry-based, and when they are part of an introductory biology course (e.g., Ardissone et al., 2019; Gya & Bjune, 2021; Reece & Butler, 2017). There are studies supporting the effectiveness of online laboratory learning when lecture and laboratory are taken concurrently and involve synchronous, small group activities (e.g., Majka et al., 2021; Rowe et al., 2018). Student attendance, participation, and comprehension can increase when given the flexibility of hybrid or online options (Riffel & Sibley, 2004; Lents & Cifuentes, 2009). Prior to 2020, online or hybrid biology classes were far more likely to be taught at two-year institutions to non-majors students (Varty, 2016). In a qualitative study from 2014, community college students who had the choice between online versus in-person coursework would opt for in-person when courses were perceived to be more challenging. Further, students would select in-person courses when seeking greater instructor interaction in challenging courses (Jaggars, 2014). This same study found that non-traditional students and those seeking greater flexibility and less interactions overall were choosing online education (Jaggars, 2014). Overall, prior work points to the fact that online laboratory learning is often effective when it is inquiry-based and part of an introductory course. Also, students may prefer online learning in circumstances in which they need flexibility, which may especially benefit students historically underrepresented in STEM. However, studies also found that online courses are not preferred when classes are considered more challenging and when students have a preference for greater in-class interactions.

Conditions were different during the COVID-19 pandemic as students were now mostly all online and there was rarely an option to select a modality. Wester et al. (2021) found that student engagement (class participation and interaction with instructor and peers) declined in the COVID online environment. Notably, engagement can be impacted by student affect (i.e., self-efficacy, science identity, sense of belonging) in the learning environment (Trujillo & Tanner, 2014) and students reported their education being severely negatively affected by the pandemic through increased stress and anxiety (Hu et al., 2022). Another recent study of college students from the USA, South Korea, and Colombia (non-STEM majors) found that self-efficacy “significantly influences students’ online learning engagement” (Zapata-Cuervo et al., 2021). Highly engaged students exhibit higher levels of self-efficacy and motivation in online classes, which results in further increases to self-efficacy and engagement. Less engaged students show lower levels of self-efficacy and motivation in these classes. However, Zapata-Cuervo et al. (2021) found that highly engaged students felt that online learning was less effective than in-person learning during the pandemic. In a study of STEM undergraduate students Camfield and colleagues found student perceptions of their academic self-efficacy dropped during the shift to online learning in Spring 2020 but could be improved after instructor intervention (Camfield et al., 2021). Thus, even for highly engaged students, pandemic online learning may have negatively affected self-efficacy. These results are important because self-efficacy not only impacts students’ course engagement but can also predict whether they will continue to engage with a task or discipline in the future.

Self-Efficacy and In-class Relationships

Developing science self-efficacy in students is an important outcome of introductory biology classes as it positively influences persistence in STEM (Chemers et al., 2011; Hurtado & Ruiz, 2012; Robnett et al., 2015). Self-efficacy is defined as one’s belief in their ability to function effectively in a given activity (Bandura, 1997). Self-efficacy is not fixed but rather “generative as an individual gains appropriate skills and is able to integrate those skills into effective execution of a task” (Bandura, 1997, p.37). According to Bandura (1986), self-efficacy in students is developed through four main pathways: mastery experiences (capable completion of a task), vicarious experiences (witnessing or hearing about a similar other complete a task), social persuasion (affirmation from an important other that a task can be successfully completed), and psychological states (the feelings associated with task completion). Mastery experiences, or successfully achieving a goal or completing a desired task, is often cited as the most impactful pillar of task self-efficacy formation (Bandura, 2008). While students often credit mastery experiences as being most important for feeling efficacious (Usher & Pajares, 2008), the other paths to developing self-efficacy also play a role and are closely related to a student’s in-class relationships.

Verbal persuasion and vicarious experiences can be influenced by students’ relationships with their instructor and peers (Beck & Blumer, 2021). High-quality relationships, in which students benefit from mentoring and community involvement, are predictors of the psychological mediators (i.e., self-efficacy and identity) leading to a commitment to STEM participation in the future (Syed et al., 2019). High-quality relationships contribute to self-efficacy paths because students are more likely to buy-in to verbal persuasion and see themselves as similar to others and therefore benefit from vicarious experiences. Thus, relationship building supports several pathways through which students develop science self-efficacy. Students with higher reported levels of science self-efficacy demonstrate increased persistence at a specific task especially when mediated by science identity (Hanauer et al., 2016). Therefore, instructor support and relationship building are important as they could help students continue to build self-efficacy and persist even when practical mastery experiences are not tenable. Yet, both mastery experience and instructor support may be at risk in online environments because courses may be asynchronous, rely on chat/discussion boards, or use simulations (Hodges & Murphy, 2009), decreasing the personal interaction element and replacing real-world mastery experiences with virtual approximations.

Supporting students in developing science self-efficacy is particularly important for populations typically underserved by online courses (e.g., Kahn et al., 2022; underserved students include those historically excluded or marginalized in certain fields including females, students of color, first generation students, and low-income students) and those more broadly underserved by STEM in general (e.g., Denson, 2017; Michel et al., 2021). These students make up the core population of community college attendees (CCRC, 2023). Community colleges teach over 44% of all students enrolled in post-secondary education in the USA (CCRC, 2023) with almost half of conferred STEM bachelor’s degrees awarded to students who have taken courses at community colleges (Tsapogas, 2004). These courses prepare students for a variety of majors including STEM, pre-health, and general education, and play a vital role in workforce development and in preparation for transfer to four-year institutions (Leonetti et al., 2023). Students at community colleges majoring in STEM and in workforce preparation fields benefit from engaging in research and evidence-based practices (Swede & Bouklas, 2018).

Our study looks at a particular CC context (large, multicampus Hispanic serving CC) in which the instructors engaged students in real research via a remote synchronous CURE and deliberately aimed to support the development of productive student-instructor and student–student relationships. The study serves the purpose of understanding how this context impacts students underserved in STEM and online education. The student population participating in the study consisted of many students who identified with underserved races or ethnicities in STEM (20 of 41 enrolled students), women (21), and non-traditional aged students (10) and students interested in pursuing a research-related science career (30). Notably, in 2015, 20.5% of all students enrolled at the PCC Northwest Campus were STEM majors and only 6.8% of Hispanic students were enrolled in a STEM major while 45% of the student population at the time identified as Hispanic. In our study, 3 out of 8 students (37.5%) identify as Hispanic, 6 out of 8 are female, and many identify with other underserved identities in STEM; thus, we consider it especially important to characterize these students’ experiences. Given the research-based focus of the course, we hypothesized that students would develop research self-efficacy and positive relationships. However, we also hypothesized that students would potentially face more challenges in developing relationships and self-efficacy due to the online course modality. The aim of this study is to present evidence to better understand how to support relationship formation and research self-efficacy in an online CC CURE setting.

Methods

This study is being conducted with approval from the Internal Review Board for Human Subjects at Pima Community College as exempt research, according to 45 CFR Part 690.101(b) (1) (2).

Positionality Statement

LAC is an educator and education researcher at an R1 institution. She is a former community college biology educator and held an administrative position as transfer student coordinator at a public R1 institution. In both positions, she worked closely with CC students to ensure that their educational experiences aligned with their academic and career goals, and she developed a passion for working specifically with the populations of students common to community colleges (e.g., students engaging with workforce training, returning students, more diverse student communities, military veterans). ADW is a non-traditional aged graduate student at an R1 institution focused on community engaged discipline-based education research. ADW began her post-secondary education at a commuter college and has spent many years working with students from CCs in bridge programs and with transfer students in field ecology programs. JBK has extensive teaching experience in CCs and has taught over 120 sections of introductory biology over 20 years. JBK is also the first woman in her family to finish college. WM is a biodiversity researcher, CURE educator, and curator of an active insect research collection at an R1 institution. LAC, ADW, JBK, and WM are passionate about understanding how undergraduate research in CUREs supports students’ persistence and development in STEM fields. These aspects inform their lens in this work: they strive to understand how contextual factors of CUREs at CCs influence student outcomes with the intention of improving undergraduate STEM education and access to research at CCs.

Course Context

Course-based undergraduate research experiences (CUREs) offer a way for many students to gain research experiences during traditional class time (Auchincloss et al., 2014) and conducting and contributing to research helps students to develop science self-efficacy (reviewed in Corwin et al., 2015; Rodenbusch et al., 2016). CUREs can take the form of multi-semester courses (e.g., Freshman Research Initiative, Beckham et al., 2015), semester-long courses (e.g., Bootleg Biology, DeHaven et al., 2022), or as a module embedded in a traditional course offering (e.g., Dune CURE, Stanfield et al., 2022). The CURE we investigate in this work, Bee the CURE is a module that took place in the final five weeks of an Introductory Biology course which is a core class for biology and pre-health majors at PCC. This class investigated bee diversity in Tucson, AZ and reported their results on a national database and to locals via museum archives and community poster presentations. As this course was synchronous online, the poster session was also synchronous online via Zoom and attended by students in the class, researchers from the University of Arizona, and Tucson community members (mostly docents from the Sonoran Desert Museum). Students made live poster presentations and were available to answer questions by attendees. The learning outcomes designed for this course align with state guidelines for transfer credit to other public institutions within the state and also reflect the goals of the instructor to engage students in real research. Specifically, 10 broad learning outcomes are targeted by the CURE module discussed here and include the following three outcomes: (1) Understand and apply basic techniques and concepts from molecular biology (i.e., extract, amplify, visualized, and sequence DNA), (2) Understand and apply basic techniques and concepts from bioinformatic analyses (i.e., taxonomically identify DNA sequences from unknown organisms and visualize data), and (3) Apply conventions displaying scientific data and research findings. Changes in student self-efficacy in wet lab and bioinformatics skills align with learning outcomes 1 and 2. Notably, if students are successfully achieving the required learning outcomes, we would expect that their self-efficacy in these respective areas would also increase.

Bee the CURE took place during the Fall 2020 semester (synchronous online) which started five months after the initial shut-down due to the COVID pandemic. While many schools opted for a hybrid format (some in-person, some online), PCC chose to operate in a fully remote modality. For students previously enrolled in Spring 2020 courses this was not new. However, unlike Spring 2020 courses, Fall 2020 courses were intended to be taught online; that is, instructors had planned for and designed courses with the online modality in mind. The instructor of Bee the CURE structured class time in a semi-flipped format. Students utilized readings and online “Lrnr” modules prior to every class meetings to familiarize themselves with the content. Class met online once a week for 2 h and 40 min synchronously via Zoom. The class met as a whole for a brief instructor-led presentation, then students completed discussion questions in small breakout rooms and then would return to the whole group for further discussion. Peer mentors (PCC students with previous class experience) were available online all semester during synchronous class time. Peer mentors guided breakout room discussions. They also worked individually with students to analyze and communicate their CURE-related data. Visiting peer mentors from U of A (students who had experience with DNA extraction, amplification, and BOLD database interpretation) were available to help with technical troubleshooting only during the bee DNA barcoding and bioinformatics synchronous course activities. Wet lab (i.e., DNA extraction and amplification) portions of the class that would have been conducted in-person were replaced with videos the instructor performing the procedures. Students watched three videos on the topics of (1) DNA extraction, (2) PCR, and (3) gel electrophoresis. Each video is 12–15 min long. The students saw videos of the instructor completing the wet-lab work. They also watched a video of the peer mentors introducing themselves and discussing their work and experience doing wet lab and bioinformatics research. Bioinformatics portions of the class, in which students interpreted the results of DNA amplification findings against the BOLD database, proceeded as they would have normally except that they were on Zoom. Thus, while students in this online version were not able to extract and amplify DNA from their bees, they were able to interpret the genetic code and compare it to the database to identify and publish their bee species, performing important steps in research and result communication. Additionally, the instructor met individually with each student via Zoom in the first few weeks of the semester to better understand student’s course loads, goals, and concerns for the semester.

Qualitative Study Design

Qualitative studies are often employed to understand a complex social phenomenon (Creswell & Poth, 2018). In this study of the CURE module participants, interviews are used to understand student experiences with the bee DNA barcoding course. Notably, while our results may provide insight into similar contexts (small, modular CUREs at CCs), our aim is to understand and report on the experiences of this particular population engaging in this online learning context mid-pandemic. While this scope is narrow, it can provide insight into particular elements of mid-pandemic online instruction and help inform choices made for CC populations in future online courses.

Participants and Recruitment

We recruited students from the course stated above at PCC. The demographics of the course we recruited from were as follows: 75% female, 44% Hispanic, and 44% first generation. These demographics are reflected in the demographics of the final participants in our study described in the table below. Our participants also reflect PCC demographics in their career aspirations. The most common jobs for individuals graduating from PCC are registered nurse, medical and health service managers, pharmacists, and nurse practitioners/midwives (DataUSA, 2023). Two of our eight interview participants had career aspirations in one of these allied health areas, five participants are STEM majors with research career aspirations, and one student is an education major. Recruitment of participants included emails and a class visit via Zoom by the researchers at the beginning of the semester for survey participation. Students then responded to a survey in which they were asked whether they were willing to participate in an interview. To avoid oversampling only highly motivated and engaged students in the interviews, a survey asking interested participants to rate their level of engagement in the course on a 3-point scale from not very engaged to very engaged was utilized. Students were also asked to report on their expected final grade for the course. Fifteen out of 41 students completed the survey with eight students electing to participate in interviews.

No students reported being not engaged. The eight students (Table 1) reported being either sort of engaged or very engaged in the class. These students reported expected grades of average (n = 1), good (n = 3), or excellent (n = 4). Interviews took place in the last 2 weeks of the semester. All interviews were conducted via Zoom, recorded, and transcribed. Interview participants included six female and two male students. This approximated the demographics of the class (51% female, 44% male, 5% prefer not to report).

Table 1 Interview participant demographic information
Full size table

Interview Design and Coding

All semi-structured interview questions (Table 2) were drafted by ADW in collaboration with LAC and JBK. The interview questions addressed relationships formed during the class (to instructors, peer mentors, and peers), self-efficacy, challenges encountered, career intentions, and student intentions for continued involvement. Questions were tested with one graduate student and one post-doctoral researcher for clarity and timing. Questions were adjusted after these interviews to improve question clarity. Interviews lasted between 18 and 38 min with a mean of 28.25 min and a median of 28.5 min.

Table 2 Interview protocol for the qualitative portion of the study. Questions and codes pertaining to challenges and coping were collected to inform research questions not addressed within this study
Full size table

Codebook development was an interactive and recursive process by LAC and ADW. Inductive coding (Saldana, 2016), a priori coding based on theoretical frameworks (CUREs and self-efficacy), and then thematic analysis (Braun & Clarke, 2021) was used to analyze interview transcripts. Initial coding included a priori codes based on interview questions and the theoretical framework described above (Bandura, 1997, Self Efficacy) and was supplemented with inductive coding upon first reviews of the transcripts. Initial codes were under the broad categories of challenges and coping, relationships, and self-efficacy. During multiple readings of transcripts, coding categories were expanded to include CURE features, career goals, motivations, and timing (of skill use). Using inductive coding and thematic analysis ADW read four interviews and added codes before re-reading and code checking (Saldana, 2016). LAC and ADW then read all interviews together, adjusted and expanded codes within categories, and came to consensus on codes and the codebook (Appendix). All transcripts were then reevaluated (re-read and coded) for all codes once the codebook was set. Codes were then summarized into broader themes that pertain to relationships and self-efficacy to address the research questions.

Results

These results report primarily on students’ responses to questions 1–12 from Table 1 as they were related to our research questions. In total, we coded 133 codes related to RQ1 (addressing relationships), and 84 codes related to RQ2 (addressing self-efficacy, see supplement for codebook) using our codes, we identified three themes related to relationship building and three themes related to research self-efficacy.

Relationships

Did Students Form Productive Relationships with Instructors, Peer Mentors, and Peers in an Online CURE Course?

  1. 1.

    Students were able to form positive relationships with their instructors and peer mentors.

    Overall, students had a highly positive relationship with their instructor and peer mentors, indicating that both were “accessible almost at any time” and that their interactions were valuable with regard to understanding the concepts and skills taught in class.

Instructor Relationships

All students reported having positive interactions with the instructor. In particular, students commented on how the instructor made learning accessible, “[the instructor] makes the knowledge, the information accessible to people who don’t come from a [science] background, but she is also there to provide details for people who do come from a science background, so she’s really good at balancing and she makes the classroom engaging for everybody.” Blanca mentioned, “I personally got more [interaction time] with her than I would have in person” in regards to the dedicated one-on-one Zoom meetings with the instructor. Further, Michelle experienced social-emotional support from the instructor:

“I've been going to school online since pretty much COVID happened since March is when I started again, and she's the only instructor I've had that did a one-on-one with everybody and wanted to get to know people and where they're at. What kind of classes they've taken in the past, what their track is right now. So, I think that helped a lot”.

Many students described working outside of school and living with extended family which added time constraints that 18–22-year-old students living in dorms at a four-year college are less likely to experience. Highlighting socio-emotional support provided by the instructor, Chelsea reported, “I was really stressed about trying to keep up with work, she pulled me aside and like one-on-one, tr[ied] to help me find a way to cope with that”. Students also felt comfortable and happy when interacting with the instructor. Carrie mentioned that the “ [instructor] is a pleasure to work with. She’s very intelligent and very smart. And also she listens great you know she’s, she’s very into what she does. And she’s passionate.”

Peer Mentor Relationships

All students reported positive interactions with near-peer mentors from both U of A and PCC. Students indicated that near-peer mentors were available and accessible with 7 students describing the relationship as collaborative and 6 students indicating that mentors provided resources. Allegra commented on how the peer mentors had helped provide a unique perspective to the bioinformatics lessons saying, “They [the peer mentors] were able to bring in slightly different perspectives and their own pursuits and describe things”. The near-peer mentors were able to share their educational experiences with the students during breakout room sessions which helped students think about their own educational directions. Carrie mentioned “this class was also to help me determine where I wanted to go and my next educational endeavor as one of [the peer mentors from UA] really helped me get the inside scoop of what you need to do for a PhD”.

  1. 2.

    Peer to peer relationships showed mixed results.

    Relationships with course peers were mixed particularly during breakout room sessions in Zoom. Veronica captured this when she described how she had a positive experience with one group member, but negative experiences with another. She explained how her active group member contributed, saying “she was one of the people that actually put in, you know, genuine effort into the discussions”. However, her other group member did not do this. “The other just, I could tell, Googled the answers because they were not things that I saw in [our readings]”. Ultimately, she felt taken advantage of by the other group member: “They plagiarized my answers, and I didn’t get credit for those answers, and I am still very bitter about it”. While other students did not describe as negative of experiences, many described this duality of interactions. Often one or two peers were interactive and collaborative, while others were absent or displayed “freeloading” behavior by not contributing and allowing others to shoulder the burden. Students generally blamed this behavior on the online format, stating that the online forum made it harder for students to interact with one another and harder to focus. Students reported making at least one connection to a peer in class while simultaneously feeling the lack of connection to the broader student population in class. For example, Allegra felt “we established a kind of [positive] group dynamic as time went on” yet she expressed “there's not really a lot of interaction that happens in the main classroom”.

  2. 3.

    Format of online class time made real-time relationship formation more challenging.

    Relationships with course peers were mixed particularly during breakout room sessions in Zoom. Veronica captured this when she described how she had a positive experience with one group member, but negative experiences with another. She explained how her active group member contributed, saying “she was one of the people that actually put in, you know, genuine effort into the discussions”. However, her other group member did not do this. “The other just, I could tell, Googled the answers because they were not things that I saw in [our readings]”. Ultimately, she felt taken advantage of by the other group member: “They plagiarized my answers, and I didn’t get credit for those answers, and I am still very bitter about it”. While other students did not describe as negative of experiences, many described this duality of interactions. Often one or two peers were interactive and collaborative, while others were absent or displayed “freeloading” behavior by not contributing and allowing others to shoulder the burden. Students generally blamed this behavior on the online format, stating that the online forum made it harder for students to interact with one another and harder to focus. Students reported making at least one connection to a peer in class while simultaneously feeling the lack of connection to the broader student population in class. For example, Allegra felt “we established a kind of [positive] group dynamic as time went on” yet she expressed “there's not really a lot of interaction that happens in the main classroom”.

    Despite overall positive relationships with peer mentors and instructors, several students (5) commented on how the online format hindered learning during their interactions with their instructors. Several expressed the sentiment that it was “not my choice of format.” Allegra added that this was her first virtual class and felt that online interactivity was “pretty impersonal”, and that online learning made it “hard to actually investigate more interesting questions with a large audience. But that has nothing to do with an instructor”. Michelle shared “I don't know what [other students] involvement was in the class, they just kind of seemed a little disconnected from it” describing that some members of the student breakout room would not engage in small group discussion. Students commented that it was “difficult to find the time to [ask questions]” online and “it’s not very interactive”. They expressed that this made it more difficult to interact overall and sometimes prevented them from engaging in ways that would advance their learning. When describing her interaction with others in the class, Michelle mentioned:

    “The online format…doesn't really work for me. It's not my choice of learning. I just learn by doing, and I get a little distracted, you know, at my own house. And then I got, you know, at the mercy of the internet, whether it's working or not. So just kind of, like, technological challenges and just not being like the way that I'm used to learning or that I know works for me”.

    This distraction by items or events in their immediate surroundings took away from potential interactions with others in the class. These distractions may not have been present in an in-person course modality.

Self-efficacy

Did Students Develop Research Self-efficacy in an Online CURE Course?

  1. 1.

    Students reported developing self-efficacy in bioinformatic skills.

    Most students in the study expressed that engaging with bioinformatic practices was a new skill that they learned during the semester. Allegra described how “being able to work through the systematics of that programming and achieve and get some answer at the end” was rewarding. Manuel described how their confidence grew as they returned to the website several times in order to work through the process, “but, you know, after going back on the website many times to try to figure it out. It is a lot easier to use”. Veronica recounted “we completely started a project from scratch, we, you know, when we built the project we submitted the project BOLD. So that's definitely something new and something that's kind of, it's really cool to watch and see the process there”. Students in the class primarily reported their growing research self-efficacy in bioinformatics as stemming from mastery experiences that they had had using the website.

  2. 2.

    Online lab participation disrupted student’s gaining “hands-on experiences” and seldom led to development of science self-efficacy in wet lab skills

    We also asked students about whether watching videos on the wet lab procedures had allowed them to develop understanding of and confidence in performing wet lab experiments. We were curious about this since COVID had prevented the students from engaging in in-person wet lab procedures. Results were mixed, but mostly indicated a lack of self-efficacy development. Students without prior experience did not find watching the videos to be helpful in developing wet bench self-efficacy. Those students expressed that the lack of a “hands-on experience” was a detriment to learning. Manuel, who did not have previous wet lab experience, explained “I don't think I could [perform wet lab procedures] just because it was like, I'm a very visual person, like I have to be there to actually understand”. Elijah, who also did not have previous wet lab experience, felt the videos would encourage him to try gel electrophoresis in the future and expressed “I would be interested in getting good at this because the more [you do] these kind of projects, the more experience, I will get to be better at this”. Notably, Elijah did not feel that they had developed expertise from watching the videos. Allegra, who had some experience in the distant past, shared that she could potentially perform the DNA extractions “maybe if the video could play right in front of me, but otherwise holding an object versus watching somebody hold the object, it's a very different situation”.

    Conversely, a student who had had prior wet lab experience found the videos helpful and explained how the videos could be a guide for what to do in the lab. When asked if she felt she would be able to do DNA extractions and run a gel based on watching the videos, Veronica responded,

    “I think it helps with me having [prior experience in a wet lab] that I kind of know how labs work anyway. Yeah, so kind of compounds my knowledge of how labs operate. But yeah, I think if I just watched [the videos] real quick again, and went into the lab. Absolutely”.

    In this instance the student already had a degree of wet lab self-efficacy and it improved slightly after seeing the videos. However, another student with previous experience felt they missed the opportunity to improve their skills. They felt stagnant in their skill development “like in the static feeling and kind of wish I was in the classroom to recreate everything. So, I remember the last time I did it. I did not stick my DNA in the proper pore in the gel, and it looks really complicated and, you know, practice makes perfect”.

    It was apparent that some areas where students may have made more gains were compromised by the online course format; videos did not replace what students referred to as “hands-on” experience.

Conclusions

This work has several important conclusions. First, it is notable that positive relationships can be formed online during an introductory course including a modular CURE. The CC students in this study were grateful for the approachability of the instructor and peer mentors and that they were available to answer questions consistently, which was not their experience in other classes. Similarly, students appreciated the socioemotional support provided by the instructor and peer mentors. Students in this study expressed that challenges such as course-load and working other jobs outside the classroom (common experiences for CC students, Gándara & Cuellar, 2016; CCRC, 2023) were among those that instructors addressed with socioemotional support. This may be especially relevant for CC students.

Instructors’ intentionality when supporting students facing challenges is a best-practice in online courses and higher levels of interpersonal interaction may result in better student performance in online courses (CCRC, 2023). Instructor behaviors that are flexible, non-judgmental, responsive, and competent also support student success in an online format (Kara, 2021), and these in turn can support positive relationship formation. We see this reflected in our research when students describe the support received and troubleshooting they engaged in with the instructor of the course.

Second, despite the overall positive experiences learning in the class, the online format often hindered peer to peer interactions and sometimes strained students’ ability to interact with instructors. Students specifically identified that having a large number of people in an online room hindered productive course questions and that peer participation was highly variable in breakout rooms where monitoring by instructors was limited. Similar to our findings, Wester et al. (2021) found that students’ behavioral engagement (participation in class discussion and studying together outside of class) decreased during remote learning. Our results, and those of other studies, highlight that we need to better understand how to build peer to peer relationships in online formats. Redmond et al. (2018) online engagement framework suggests five types of engagement that students could exhibit in an online course: social, cognitive, behavioral, collaborative, and emotional. These authors encourage instructors to be reflexive regarding which types of student engagement will enhance student learning (Redmond et al., 2018). While we did investigate student perceptions of engagement, we did not characterize or study how instructors or peer mentors were trained for online instruction or their reflexive practices. This is something future studies could address. One study specifically focused on CC emergency online instruction during the pandemic found that mentored professional development regarding course design and delivery can improve the student experience in an online course (Achen & Rutledge, 2022). Work by Thacker et al. (2022) found that online instruction that fosters interpersonal connections positively impacted student engagement particularly for underrepresented students. Clearly there are productive lines of research focusing on the potential for professional development to increase student engagement which could help to enhance online CC instruction and relationship formation.

A third conclusion of this work is that students reported developing research self-efficacy in bioinformatics as a result of engaging with this process in the online environment. This points to the importance of students being able to actually do the procedures associated with research skills they aim to gain. While this is not surprising, given that a primary source of self-efficacy is mastery experiences (Usher & Pajares, 2008), it has important implications for online learning. The online format and COVID pandemic prevented many students from accessing the wet lab, which in turn prevented them from gaining wet lab experiences that may have allowed them to develop wet lab self-efficacy. Observing instructors perform tasks and encourage them (social persuasion), and hearing peer mentors describe their experiences with tasks (vicarious experiences) and provide encouragement (social persuasion), was predicted to be an important source of self-efficacy development. However, students did not report self-efficacy development from watching videos alone. Thus, social persuasion and vicarious experiences may not be enough for students to develop self-efficacy, unless students have some prior research experience (and some mastery experiences). This result, while not entirely surprising, is important since it indicates that in this case, the outcomes of the course may not have been fully achieved. Students who expected to gain hands-on experience and wet-bench skills did not report an increase in self efficacy which may translate to them not having the transferable skills they desire for their pre-health or STEM careers. Also, it is the expectation of transfer institutions that students come into their programs with specific skills (including wet bench skills), and these may not have been realized. Considerations of how to maintain important outcome achievement are paramount when courses switch modality.

The implication of this finding is that, for students early in their undergraduate career, it may be important to engage them in the practices of the discipline to begin self-efficacy formation. Teaching them about and having them view practices, even with encouragement from instructors, may not be sufficient. This is supported by numerous studies. For example, Aikens and Kulacki (2023) found that collaborative group work experiences that resulted in mastery supported the formation of self-efficacy around quantitative skills in introductory biology students. Specifically, being able to get help from their peers in problem solving contributed to the development of self-efficacy. Research self-efficacy development through independent work, though with support from peers and instructors, was found in a study of CURE participation by Gin et al. (2018). The referenced examples along with the results of this study imply that developing research self-efficacy, while often supported by vicarious experience and social persuasion, begins with mastery experiences.

Implications

As we progressed and overcame instructional challenges related to the pandemic, we saw a proliferation of publications reporting positive outcomes of online and remote education (e.g., Race et al., 2021; Sandrone et al., 2021) and more specifically triumphs of laboratory learning done at home or in remote or online settings (e.g., Gya & Bjune, 2021; Lichti et al., 2021). While it is immensely encouraging to view such positive learning outcomes across different contexts, it is also important to recognize and report on the limitations of online environments. It is even more important to do this research within contexts that include historically underserved students who may be more likely to leave STEM fields, such as CCs. Our study addressed two important components that contribute to self-efficacy formation and persistence in STEM and that we were concerned might be negatively impacted by an online environment: relationship formation and self-efficacy development. In contrast to many other studies that highlight only successes and positive results from online learning, our results are mixed. The implications of this are that online learning can indeed pose a challenge and perhaps even a threat to self-efficacy development, especially when considering skill development and self-efficacy in “hands-on” areas, such as wet-bench research skills. Yet optimistically, our study highlights the potential of online CC learning environments to provide mastery experiences in online science contexts (e.g., bioinformatics) and opportunities for relationship building. While the results of this study are not generalizable to all online education environments, we believe that this research adds to the body of evidence supporting the need to provide professional development opportunities to instructors and peer mentors who will be providing online laboratory courses. We also caution administrators and educators against decisions that might remove opportunities for students to engage with direct mastery experiences in field and wet-bench science as these may be important for early self-efficacy formation.

Limitations

A limitation of this study is that it investigates a small sample population for one particular institution. While this limits the generalizability of our results, this enables us to characterize the environment of an online CC CURE delivered during pandemic instruction, a unique opportunity. Furthermore, this study examines the experience mainly of students who reported they were at least somewhat engaged. Thus, our research does not represent students who would consider themselves not to be engaged in the course. While we cannot understand experiences of students who did not engage, this is a benefit to our study since students with some level of engagement may be the most capable of accurately reporting the efficacy of online educational experiences. This study represents a demographically limited sample that focuses on white and Hispanic students. However, this is an HSI and the Hispanic student population is one of the most rapidly growing demographics in the USA in higher education. Thus, it is highly important to better understand experiences of this population. Finally, some of the interviews with students occurred prior to the poster presentation which could have been a mastery skill obtained during the course and could have skewed self-efficacy results for those particular students. However, we consider this unlikely since the format of the poster session did not allow much direct interaction with participants (students only interacted with attendees if the attendees asked them questions, which unfortunately did not happen often). Upon examination of our data different themes were not apparent between students who had interviews prior to and after the poster presentation.