The reimagining of science education emphasizes scientific literacy and critical thinking (American Association for the Advancement of Science, 2009; National Research Council, 2003; Michaels et al., 2008). Professional organizations—including the American Society for Biochemistry and Molecular Biology, the American Chemical Society, and the American Association of Physics Teachers—recommend that graduating students have science communication and comprehension skills, and the MCAT (Medical College Admissions Test) focuses on critical analysis (Kirch et al., 2013; White et al., 2013). To build STEM (science, technology, engineering, and mathematics) critical thinking, undergraduates should learn to analyze data and written materials, synthesize concepts independently, build and interpret models, and communicate information (Alberts, 2009; American Association for the Advancement of Science, 2009; Gormally et al., 2012; Trujillo et al., 2016a; Zagallo et al., 2016). This analytical skill set is imperative for undergraduates entering graduate or professional schools, where problem-based learning and journal clubs (regular group meetings to discuss current scientific and medical literature) are prevalent (Afifi et al., 2006; Alguire, 1998; Bauer, 2015; Dahiya & Dahiya, 2015; Edwards et al., 2001; Kellum et al., 2000; Kitazono, 2010; Kozeracki et al., 2006; Lachance, 2014; Linzer et al., 1988). Development of such abilities as an undergraduate can improve student confidence and ability, help alumni secure admission to advanced degree programs, and provide advantages over peers; however, past studies focused on voluntary, cocurricular programs targeted to students performing laboratory research, and thus a limited student population is reached (Kozeracki et al., 2006; Wiegant et al., 2011). Furthermore, the use of journal clubs in undergraduate curricula is still viewed as rare (Hall & Wolfson, 2000; Kitazono, 2010).

Three main roadblocks to using journal clubs with undergraduates emerge from existing literature. The first roadblock involves helping students transition from reading scientific literature to understanding what they read. Robertson’s workshop method orients students to critical reading and comprehension; however, this requires the time-intensive task of selecting appropriate articles for undergraduate learning (Kitazono, 2010; Muench, 2000; Robertson, 2012; Sandefur & Gordy, 2016; Woodham et al., 2016). The second concern is finding a balance between faculty and student leadership and learning how to transition between the groups; many existing models place a large weight on the faculty member in guiding student preparation for journal club and in article selection (Clark et al., 2014; DebBurman, 2002; Kitazono, 2010; Muench, 2000; Roberts, 2009). The final hurdle involves developing the norm of broad student engagement during class discussions, as students are often reticent to share ideas and interpretations (Kozeracki et al., 2006).

In counterpoint to the journal club method, several groups developed classroom strategies that address similar skill sets, combating some of these issues. The CREATE method (Consider, Read, Elucidate hypotheses, Analyze and interpret data, Think of the next Experiment) develops students’ critical-thinking skills by using a series of articles from the same research group to follow a path of scientific discovery, which, when maintained throughout an entire semester, creates a positive shift in students’ skills and confidence (Hoskins et al., 2011). Similarly, DebBurman’s work (2002) incorporates scientific literacy and data analysis into a coordinated set of assignments including journal club, medical news journalism, review articles, seminar presentations, and writing. Additionally, applying the Process Oriented Guided Inquiry Learning (POGIL) method to primary literature, by incorporating between one and three scientific literature–focused sessions in a POGIL-based course, increased undergraduate students’ confidence with approaching literature (Murray, 2014).

Other instructional activities, including Round and Campbell’s Figure Facts technique, focus on data rather than the paper as a whole, meaning that students develop analytical and interpretation skills by studying a single figure (Kitchen et al., 2003; Round & Campbell, 2013; Sato et al., 2014; Spiegelberg, 2014; Woodham et al., 2016). Finally, to build students’ modeling skills, the TRIM (Teaching Real Data Interpretation with Models) method teaches students to learn the interconnection between data and modeling by working through paired sets of information; the MACH (Methods, Analogy, Context, How) method is a metacognitive tool to help students explain complex scientific processes and interpret new information (Modell, 2000; Trujillo et al., 2016a, ; ).

In this article, I describe an advanced course intended to take the strongest aspects of existing pedagogies and develop a model in which skill development is the driving goal, one that could be overlaid on course content in any STEM discipline. The flexibility of content and the focus on literature help emphasize the recommendation from the National Research Council’s Bio2010 report that courses highlight cutting-edge developments in science (National Research Council, 2003). Driving factors in course development were to (i) build on the initial growth in skill development students achieve in mid-level cell biology courses, which adds skill development assignments to a traditional course; and (ii) prepare students for postgraduate degree programs and the scientific workforce through development of critical scientific thinking skills.

Course design for Bio404: Nuclear Structure and Function

Bio404 is an advanced cell biology course taken by undergraduates who had previously completed Bio201: Foundations in Cell and Molecular Biology and Bio302: Cell and Molecular Biology at a small undergraduate institution. The course averages 10 students, who represent majors in molecular biology, biology, and neuroscience (Table 1). While course content focuses on architectural aspects of nuclear cell biology and molecular understanding of genome structure, the major objective is to use primary and secondary literature to broaden students’ scientific skill set in content comprehension, data analysis, productive scientific discussion, oral presentation, and scientific writing. The course schedule is divided into four units (Nuclear Periphery, Nuclear Pore, Chromatin, and Nuclear Bodies), each composed of multiple 1-week modules that consist of three 1-hour course sessions (see Figure 1). Figure 1 describes module design, with a content-specific example in Figure 2 (Jung et al., 2012; Krishnan et al., 2011; ; ). Each module introduces students to a new concept, then builds on this knowledge by developing models from data before concluding with a presentation and discussion of current research.

Session 1: Lecture

In this session, secondary articles are paired with a traditional lecture to introduce students to content, ensuring everyone in the class has a similar knowledge base. Students receive (via the course management system) a recent review article on the week’s topic to read prior to class. The article can be changed yearly or reused. The course is taught in the instructor’s area of expertise, so the identification of current articles provides a mechanism for staying engaged in recent research while also preparing for teaching. The class meeting is a traditional PowerPoint lecture that introduces the week’s theme. Students receive guided notes with the PowerPoint images, models, and diagrams from the sources used, but without text, and they are encouraged to annotate slides.

Session 2: Work It Out (WIO) Wednesday

In the second session, students engage in data analysis on content from the literature without textual context to develop analytical and model-building skills while furthering content knowledge. Students receive a packet of data derived from one or more primary research articles, accompanied by probing questions about the methodology, results, and interpretation. Sources used are typically ones cited in the lecture review article, allowing for more in-depth learning about that content area. Initial development of the packets takes time, but these are not typically changed yearly. In planning for a 1-hour session, packets typically include five pages of data, plus a model-building activity, and preparation time is comparable to planning a lecture. This organization is based on the CREATE and POGIL methods (Hoskins et al., 2011; Murray, 2014). The students work in groups of three or four to analyze the data using the provided questions, then develop models to explain the data and its links back to lecture concepts. This session occurs in the college library’s “Hub,” a collaborative space that includes workstations where students can share their computer screen with their group (Figure 1). Students are encouraged to research methods and information that are not familiar to them. The packets are not graded, and the instructor and an undergraduate teaching assistant (TA) circulate between groups to offer guidance and discuss outcomes and models with students. If time allows, groups share and discuss models with other groups. Students are responsible for the content learned from the WIO sessions on course exams.

Session 3: Journal Club

The third session is a journal club, which is modified from existing designs to address the issues of faculty efficiency and student engagement. One student is responsible for leading the journal club discussion on an assigned primary article that correlates with the information learned in the previous two sessions. Because the topic has been introduced and analyzed throughout the week, the instructor can select a more intensive article than would typically be used for an undergraduate course and thus is not constrained by students’ level of understanding, a time-intensive aspect of selecting an article for instructional use (Kitazono, 2010; Muench, 2000; Roberts, 2009; Robertson, 2012; Sandefur & Gordy, 2016). The same articles can be used yearly or updated as the instructor sees fit; once again, this facilitates the opportunity for the instructor to simultaneously stay abreast of current research while engaging in course responsibilities. The instructor or undergraduate TA leads the first session to model the expected style and depth of presentation or discussion. Following the session, students complete a 10-question quiz. The student presenter writes the quiz and answer key and is graded on the quality of both. Students are instructed to write a short-answer quiz with approximately five questions that focus on data interpretations and conclusions. The instructor, using the student-provided answer key, grades all other students on their correct completion of the quiz. Shifting the balance of quiz and answer key creation to the students reduces a large amount of the instructor’s burden so the instructor can focus on selecting articles and observing discussion. The post-session quiz and the fact that journal club content is present on exams act as incentives for students to participate in the discussion and ask questions, overcoming the issue of students’ reticence to participate. Furthermore, students are graded on their participation, receiving credit for speaking up, whether they are asking questions or providing insight.

Course assessments

The course includes an additional 3-hour weekly laboratory session, which involves multiweek research projects and is used for exams. The lab is a curricular requirement of the college’s biology courses, and the module format can be taught without one. Assessments tied to the modules include journal club quizzes, participation, presentation, and four unit exams; the exams address content from the lecture, WIO Wednesday, and journal club sessions from three or four modules within a single unit and contain short-answer, data analysis, and model development and interpretation questions. The cumulative final exam requires students to make a poster where they draw a model of the nucleus, incorporating as many concepts and structures from the course as possible (Figure 3). The goal is for students to interconnect the models they have built during the entire semester into a single nuclear structure and function picture. Each student then receives a 15-minute testing window to explain their poster to the instructor and answer questions.

Growth in self-reported ability and confidence in critical-thinking skills

An initial small-scale study was undertaken to begin addressing whether this instructional strategy can improve students’ ability and confidence in scientific critical thinking. Self-reported confidence and ability-ranking data on tasks related to comprehending, analyzing, and presenting scientific research were collected from the 2018 cohort (N = 9) and the alumni of the 2016 course (N = 9). The study’s experimental design was approved by the college Institutional Review Board. All students and alumni completed an informed consent form and answered demographic questions. Students completed entry, middle-of-term, and exit surveys containing Likert-scale questions via SurveyMonkey regarding their confidence and abilities with data analysis, scientific literacy, and critical-thinking skills, as well as their perceived benefits of the course, which were compared to a single alumni survey. Questions were adapted from surveys used in the CREATE studies, Murray’s POGIL study, and Kozeracki’s journal club program (Hoskins et al., 2011; Kozeracki et al., 2006; ). The scores for each question were compared across pre-, mid-, and post-surveys using a Friedman test, and current students and alumni were analyzed using a Kruskal-Wallis test. Composite scores (of all ability questions or all confidence questions) were similarly compared.

Self-reported abilities in reading, discussing, and presenting scientific articles, as well as analyzing and comprehending data, were tracked. Alumni reported strong abilities (< 4 on a scale of 1 to 5) for nearly every category, a ranking higher than the Bio 404 students at the beginning of the semester but comparable to students at the end of the semester (Table 2). The exceptions included identifying patterns in data and developing models, where the alumni reported only “some ability” (~3 on the same scale). Interestingly, these were the skills where current students showed consistent growth. Large gains were self-reported in all abilities except reading scientific articles, as students already had experience with reading articles from other classes, but this was their first attempt at intense analysis and presentation. The highest gains were in identifying patterns in data and using data to develop models, with higher self-reported ability than the alumni. A mid-semester analysis showed stepwise progression in most areas except for the ability to present a scientific article, which showed that students had lower ability than they did initially, likely due to the fact that several students had not presented yet and were wary of their first experience (Table 2). As such, the factor of confidence was also taken into consideration.

Students ranked their confidence in the same set of critical-thinking skills on a scale of 1 to 5, with 1 representing no confidence and 5 representing extremely confident, as used by Hoskins and colleagues (2011) to distinguish a student’s skill and their perceived ease and comfort with performing said skill. Students displayed the largest gains in areas where they grew in self-reported ability, with the most growth in identifying data patterns and analyzing articles (Table 2). Students showed incremental growth in areas that started low in confidence, and upon completion of the course, students showed significant growth for all skills and were at a level comparable to alumni. As seen with the ability scores, the students’ confidence in presenting dropped mid-semester; however, confidence increased dramatically by the end of the semester.

Closing the gap between confidence and ability

Because of the differences observed in student gains for self-reported ability versus confidence, the relationship between these two factors was addressed. At the beginning of the semester, students displayed a statistically significant difference in overall confidence versus ability, with higher ability rankings (Table 3). This gap was reduced but still existed at the midpoint of the semester, but by the end of the semester, there was no significant difference between the two, a characteristic also seen in the course alumni (Table 3). This result supports the idea that continuous immersion in and repetition of skills built students’ confidence in the abilities they gained from this literature- and data-heavy instructional design.

To ascertain if the student gains in ability and confidence were specifically linked to modular course design, the rankings were compared to those for students in the prerequisite Bio302 Cell and Molecular Biology course. While the Bio302 class touches on many of the skills emphasized in Bio404, it does so to a lesser extent, with less repetition and fewer in-depth experiences, focusing primarily on reading papers, which is only done twice during the semester. While Bio302 is a prerequisite for Bio404, the same students were not in both cohorts used for this study (Bio302, fall 2018; N = 21). The courses have similar gender and class standing demographics; the major difference was that there was more variation in student majors (i.e., there were biochemistry majors in Bio302 but not in Bio404) and a higher percentage of students in Bio302 were interested in health careers (86% vs. 66%). Bio404 students’ ability rankings were higher than those of Bio 302 students (4.2 vs. 3.8 at the end of the semester). Furthermore, the overall gains in ability and confidence were higher for Bio404 students than for Bio302 students, with increases in ability and confidence of 0.7 and 1.2, respectively, for Bio404 students, compared to 0.3 and 0.5 for Bio302 students. In the context of a small undergraduate institution, these results support the enhanced benefits of the modular course design in building scientific critical-thinking skills versus overlaying literature-based assignments in a lecture course.

Postgraduation benefits

A productive course impacts student understanding of the value of achieving mastery in critical thinking and its implications after graduation. The ability to present a scientific article had the lowest perceived benefit for alumni (rating of 4 on a scale of 1 to 5, where 5 represents “strongly agree”), whereas Bio404 students uniformly strongly agreed at course completion about the benefits of all critical-thinking skills (rating = 5; Table 4). Alumni agreed or strongly agreed that the ability to read and interpret data positively benefited them in their careers, despite many indicating low levels of direct use of these skills in their positions; additionally, alumni agreement was very high (4.9) for the statement that Bio404 positively impacted their success in their current position. These rankings suggest that the broader scientific critical-thinking skills students gained currently benefit them postgraduation, regardless of the specific tasks they are performing in their jobs or postgraduate programs (Table 1 and Table 4).

The positive impact of Bio 404 is further reflected in the students’ and alumni’s open-ended responses to the prompts “Please reflect on how, if at all, you believe the course Bio404 will impact (has thus far impacted) your ability and confidence in reading scientific journal articles and analyzing scientific data” and “Please reflect on how, if at all, the course Bio404 will assist (assisted) you in your career postgraduation.” Three major themes developed out of the responses (Table 5):

1.Students and alumni progress to describing confidence in tasks beyond the original course objectives.

2.There is a transition from reading articles to the processes of understanding and analyzing.

3.At the end of the semester, students can highlight specific advantageous skills, and the alumni repeatedly indicate preparation for postgraduate studies and distinct skill advantages over their peers, a benefit on par with other programs, including a more selective Howard Hughes Medical Institute–linked program (Kozeracki et al., 2006).

By developing an entire package of scientific critical-thinking skills, this course prepares a broadly trained student who is capable of success in health professions, graduate school, and other careers. Students’ understanding of their new skills and the changes in their attitudes and beliefs are in and of themselves benefits of the curriculum.

Repetition and variation of tasks for developing a broad base and depth of analytical skills

The modular design of Bio404 was intended to go beyond journal club to build students’ analytical skills and do so in a manner that was productive and beneficial for both students and the instructor. Through the use of WIOs and journal club, students were provided authentic tasks with relevant connections to their development as scientists, giving students the important explicit, intentional stimulation of multiple skills (Abrami et al., 2008; Tsui, 1999; Wiegant et al., 2011). Variation in the mechanisms used for skill development is thought to provide a distinct advantage, as studies that focus on one method of training display gains limited to a subset of skills (Kozeracki et al., 2006; Murray, 2014; Roberts, 2009; Round & Campbell, 2013; Sandefur & Gordy, 2016). While variation increases the students’ skill set, repetition is also a key factor in developing more robust student skills in any given area (Murray, 2014). This initial small-scale study demonstrated increased student gains in ability by mid-semester, and further repetition of the module format led to additional gains by the end of the semester. The variation of activities without repetition led to high initial gains but drastic decreases in students’ self-reported abilities within the next year (DebBurman, 2002). Conversely, Bio404 alumni (2 years after course completion) reported gains comparable to students who had just finished the course, emphasizing the benefit of repetition. In sum, it appears that in this context a more comprehensive program, which includes both repetition and multiple activities, shows promise for development of broader gains across many specific skills along with deeper gains in scientific critical-thinking abilities. The next steps to take to generalize these conclusions should include more rigorous, tailored research methods and analysis, larger sample sizes, and the application of the model to a broader set of contexts.

Transferability of the module method to other courses

Each existing method for developing scientific critical-thinking skills positively affects students; thus, when implementing a new design, instructors need to consider what type of impact they want to focus on (which abilities, level of confidence, types of benefits) and whether they want a short- or long-term intervention. This preliminary study suggests that the module style (lecture, WIO Wednesday, journal club) allows an instructor to help students develop multiple critical-thinking skills and scientific abilities in a manner that provides a sustained intense impact, with several instructor and instructional benefits. The course is a low-cost alternative to labs for the development of inquiry-based skills and can be applied to nearly any scientific content. Understandably, course redesign is time intensive, but there is the added benefit of staying on top of the literature in the instructor’s field of interest, especially for faculty who like to blend their research and teaching. Several papers were selected from presentations attended at national scientific conferences or by reading recent review articles used for lecture content. Furthermore, if reading literature for lecture is an added time constraint, lecture topics can be derived from a textbook already in use, with the literature cited as a source of paper ideas and problem books as basis for creating WIOs. Finally, content goals need not differ from an existing course; the variation here is in the method, which allows existing lectures to be transitioned using data to WIO or journal club, simultaneously delivering content and skills.

In addition to presenting content in a novel manner, this course design can help instructors overcome many of the issues that come with simply adding on a journal club session to a class or program. The use of lecture and WIO sessions builds student knowledge so that article selection becomes less difficult and time consuming as the instructor is aware of students’ background knowledge. The repetition of reading papers, performing data analysis, and participating in a journal club creates students’ ownership of their knowledge development and builds their independence in critically analyzing the literature so they are prepared to work with a broader variety of more difficult articles. Finally, the graded aspects of the journal club presentation, post-session quizzes, and exams provide higher stakes that encourage student participation, especially as discussion is graded on level of involvement rather than knowledge, where a good question is just as worthy as a strong insight into data interpretation.

The easiest, most expedient transition into this course model could occur at similar small undergraduate-focused institutions, where faculty have more control over curriculum and small courses are more prevalent. While the model is designed for classes of 10 to 15 students meeting three times per week, aspects of it can be modified for medium-size courses, large lectures, and various schedules. In a class that meets twice per week, for example, the instructor could combine lecture and WIO sessions and expand journal club or split time with lecture or WIO. With medium-size courses, journal club presenters can work in pairs or small groups so every student still has an opportunity to present.

In larger courses, if recitation or review group sessions are available, smaller sections of the class could simultaneously hold the same journal club. Furthermore, the WIO work is not graded; therefore, in large lectures using instructional space that allows for collaboration, as many groups as are needed could be developed, with teaching assistants and instructors circulating for guidance. In this case, the final model developed in the WIO should be discussed as a class, especially if instructors did not meet with every group. While such transition to this style at larger institutions may seem daunting and would likely require coordination of several individuals, other models, most notably the CREATE method, have been successfully scaled to community colleges as well as master’s and doctoral universities (Kenyon, Onorato et al., 2016; Kenyon, Cosentino et al., 2019). Additionally, while students appear to benefit from the weekly repetition involved in this method, if the implementation at a given institution is not conducive to the use of the system during an entire semester, parts of the course rather than the whole semester could be taught using the module style, or one aspect (such as the journal club or WIO) could be incorporated on its own. This block method of implementation would also reduce instructor preparation in terms of developing the WIOs and selection of journal club articles, and it has been used to scale and implement the CREATE method at various institution types (Stevens & Hoskins, 2014).

In summary, the module design was developed to teach and enhance students’ scientific critical-thinking skills. Early assessment of this strategy at a small undergraduate institution suggests that the modular variation of methods, lecture, data analysis, and journal club can lead to enhancement of students’ abilities across a broad range of skills and that these gains are sustained postgraduation. Students gained confidence in their skill set, closing the gap between confidence and ability present in students from other courses. In addition, alumni of the program perceived benefits and advantages from the course style regardless of their career path. These outcomes suggest that by combining aspects of existing structures, this course style can overcomes various instructor and student barriers. Future research should aim to determine if applying this strategy to a broader range of content objectives and course settings can similarly build these key scientific skills.

Karen Resendes (resendkk@westminster.edu) is a professor in the Department of Biology and director of Undergraduate Research at Westminster College in New Wilmington, Pennsylvania.