Examining the Impact of Project-Based Learning on Students’ Self-Reported and Actual Learning Outcomes

Sparked by John Dewey’s constructivist theory of knowledge, active learning has found increasing research support as an effective method of teaching. Defined as a “student-centered concept denoting a participative process of engagement in classes and materials where students are involved in constructing their own learning,” (Prasad et al., 2022, p. 2) active learning is often considered the opposite of passive learning in which instructors relay information to their students (e.g., lecturing) (Chickering & Zelda, 1987; Fields et al., 2021; Powner & Allendoerfer, 2008). It engages students in the process of their learning through activities, emphasizes higher-order thinking, and often includes group work (Freeman et al., 2014).

The constructivist approach emphasizes that knowledge is constructed by the actions of the learner and encourages an activity-based framework (Jonassen, 1991; Linton et al., 2014). Importantly, when compared to traditional lecturing methods, active (or constructive) methods have demonstrated enhanced long-term retention of information (Gijbels et al., 2005). As active learning engages students in the learning process, collaboration is often utilized as a method of facilitating and enhancing student engagement. Collaborative team-based learning has been demonstrated to increase levels of engagement, motivation, learning, enjoyment, and to enhance critical thinking in a multitude of fields, although the literature is limited for public health (Eukel & Morrell, 2021; Freeman et al., 2014; Laal & Laal, 2012; Michael, 2006; Prince, 2004; Shin et al., 2015). In this setting, teachers act as facilitators while students work together in groups to answer questions and find solutions while working (Smith & MacGregor, 1992). As described by Smith, collaborative learning can also support students’ social and emotional needs, as well as academic literacies (Smith & MacGregor, 1992).

Combining both active and collaborative learning, project-based learning (PBL) is an instructional approach which involves students working on projects to explore and solve authentic real-world problems. This method focuses on providing students with an opportunity to acquire deep knowledge through investigation of an authentic problem. Typically, students learn the core curriculum and apply their knowledge by asking questions, designing plans, collecting and analyzing data, drawing conclusions, and creating artifacts in order to develop solutions to their driving questions (Blumenfeld et al., 1991). Not only does PBL help students develop the skills described previously with active and collaborative learning but it also improves important life skills, such as communication, time management, and has been demonstrated to narrow achievement gaps for underrepresented students in a variety of academic fields as well as increase overall enjoyment and learning (Theobald et al., 2020). Overall, PBL can enhance student engagement and motivation, leading to deeper learning and improved academic performance (Almulla, 2020; Blumenfeld et al., 1991; Guo et al., 2020; Helle et al., 2006; Maida, 2011; Ngereja et al., 2020; Tadesse et al., 2022).

Although PBL has been demonstrated as an effective teaching method, there is still much to be understood about its true impact on student learning. While there has been some research on the difference between self-reported learning and actual learning in traditional educational settings, there is little research on this distinction in project-based learning contexts. Self-reported learning (or perceived learning) refers to knowledge or skills individuals believe they gained through a learning experience. This form of assessment is often used in educational research but may not always reflect actual learning. Actual learning refers to the actual knowledge or skills gained, as measured by objective assessments or performance evaluations.

This is an important area for further investigation, as PBL is becoming increasingly prevalent in educational settings and health promotion is a critical societal need. Health promotion is a growing field due to the increasing importance of disease prevention and health promotion initiatives, especially after the COVID-19 pandemic. In tandem with the rise of chronic diseases and the aging population, there is greater need for professionals trained to address these multifaceted challenges. PBL provides students with hands-on experience and practical skills which are essential for identifying and addressing health disparities and creating effective health promotion strategies (Trevena, 2007). As such, it is essential to optimize learning interventions in this field.

This study aims to investigate how an active and collaborative project-based health promotion assignment affects both actual and self-reported learning.

Descriptive statistics for the demographics of the sample (N = 33) are shown in Table 1. The sample was majority female (76%), Hispanic (64%), upperclassmen (79%), first-generation college students (79%), with an average age of 20.2 years (SD = 0.95).

There was a statistically significant difference in pre- (M = 0.57, SD = 0.11) and post-actual learning scores (M = 0.83, SD = 0.09); t(32) = −11.85, p < .001, demonstrating an average 25% increase of actual content mastery. Cohen’s effect size value (d = 2.06, 95% [−2.66, −1.45]) suggested a large effect and high practical significance.

There was a statistically significant difference in pre- (M = 0.47, SD = 0.06) and post-self-reported learning scores (M = 0.76, SD = 0.09); t(32) = −17.91, p < .001. These results suggest, on average, post-SR learning scores were 29% higher than pre-SR learning scores. Cohen’s effect size value (d = 3.11, 95% [−3.95, −2.28]) suggested a large effect and high practical significance.

Additionally, there were statistically significant differences among the SR learning scores and actual learning scores for both the pre-assessments (t(32) = −6.40, p < .001; with an average discrepancy of 11%) and the post-assessments (t(32) = −3.12, p = .003; with an average discrepancy of 7%). Cohen’s d also suggested large to medium effect sizes for both comparisons, reporting (d = 1.10, 95% [−1.53, −0.66]) and (d = 0.55, 95% [−0.92, −0.18]), respectively.

Due to the small sample size, the independent samples t-tests conducted to determine differences between demographic groups were underpowered and few statistically significant differences were detected (Table 2).

Only 11 students completed the post-semester Evaluation of the Health Promotion Project Survey out of the original sample of 33 students (see Table 3). This high dropout rate was most likely due to the survey being distributed after the semester was over when students were less likely to check and respond to their emails. The demographics of this subset of students reflected those of the larger sample. This subset was similarly majority female (73%), Hispanic (82%), upperclassmen (100%), first-generation college students (73%), with an identical average age of 20.2 years. In order to evaluate other potential differences in the subgroup of 11 students, the statistical analyses were re-run on their SR and actual learning scores. The results were identical to those of the larger sample.

Results from the Evaluation of the Health Promotion Project Survey were overwhelmingly positive and suggested the project was a valuable learning tool. All students either agreed or strongly agreed the project helped them gain an in-depth understanding of, and an opportunity to relate better to, health promotion concepts learned in class. Similarly, all students reported the project both helped them see the steps required to plan a health promotion program and feel confident in their ability to design programs better in the future because of the experiences gained from working on the project assignment. The majority of students reported the project gave them an authentic health promotion experience (82%) and they enjoyed working with their fellow group members very much (82%). Ninety-one percent of students reported they would rate their group efforts as outstanding (i.e., collaboration, communication, and knowledge sharing within the team).

Although PBL is widely accepted and touted as an effective pedagogical style, it is vital to measure its efficacy on actual learning. This study built on previous recommendations (Bacon, 2016; Deslauriers et al., 2019; Trevena, 2007) and primarily attempted to investigate the differences between self-reported and actual learning before and after a PBL health promotion course.

Assessment results suggest both SR and actual student learning increased by an average of 26% over the course of the semester. However, on both the pre- and post-learning assessments, students consistently underestimated their knowledge of core course concepts by approximately 9% (11% and 7%, respectively). As expected, this aligns with previous literature which would expect students’ self-reported learning scores to be more accurate at the end of the semester after they had been exposed to course content (and therefore, would have increased recognition of their deficits) (Surdilovic et al., 2022).

Historically, educational research has tended to focus on mainstream populations, excluding those from diverse backgrounds, including ethnic and racial minorities. This lack of inclusivity perpetuates existing inequalities. Several studies have demonstrated that PBL is beneficial to students of all backgrounds (Qondias et al., 2022; Rahat, 2022; Singaram et al., 2008). As such, PBL is touted by many educators as a best teaching strategy for achieving educational equity as it is rigorous, helps students build skills which are useful outside of the classroom, and increases connection to course content. Given the nature of our predominately Latino and first-generation sample, our findings support the claim PBL may be beneficial for students of underrepresented backgrounds.

As mentioned, previous literature has emphasized the importance of distinguishing actual from perceived learning in educational settings. PBL studies often measure perceived learning instead of actual learning. This can lead to problematic interpretations of the effects of PBL on learning as the two are distinct phenomenon. This is well demonstrated by Donald R. Bacon’s work. In 2011, Bacon replicated a previous study by Revere et al. which demonstrated a positive learning effect when students completed a quiz individually and then in groups (i.e., team testing) (Revere et al., 2008). Importantly, Bacon noted their measure of learning was self-reported, and therefore only measured perceived learning. When the study was replicated by Bacon, he found team testing did have a positive effect on perceived learning but had no significant effect on actual learning (Bacon, 2011).

As such, the current study offers a unique insight into student perception of learning as well as an objective measurement of their actual learning. The close alignment between SR learning and actual learning outcomes suggests students had a relatively accurate perception of their own learning progress. This awareness can be beneficial as it allows students to reflect on their strengths and areas for improvement, which in turn fosters a sense of confidence and motivation. Moreover, the alignment between students’ perceptions of their learning and measurable increases in their knowledge validate the effectiveness of collaborative, PBL approaches, emphasizing the importance of providing students with opportunities for meaningful and impactful learning experiences in health promotion.

Despite this strength, this study is not without limitations. The sample size is relatively small, neither the actual nor SR learning assessments have been validated, and the follow-up survey evaluating the health promotion project was completed by a small subset of students. The Evaluation of the Health Promotion Project Survey was distributed a few weeks after the semester was completed, which most likely decreased the number of students who were aware of, or interested in completing the survey. Ideally, the survey should have been distributed along with the post-learning assessment at the end of the semester, not after its termination. Additionally, we could not compare the results from the survey to the learning scores due to the discrepancy in response rate. These limitations should be kept in mind when considering the implications of this work. Although we believe the results are still valuable and insightful, future work should prioritize higher participant recruitment and improve the timing of assessments to achieve an adequate response rate.

A notable importance of this study is its focus on conducting educational research on an underrepresented population. Including underserved and underrepresented students in these investigations ensures educational practices, policies, and interventions are equitable and inclusive. It allows researchers to identify and address disparities, biases, and any barriers which exist for these students. Educational instructors must acknowledge students’ diverse cultural backgrounds, languages, abilities, socioeconomic statuses, and learning styles. Students’ cultural identities and backgrounds influence their educational experiences and outcomes. Therefore, we advocate for inclusive educational research on diverse populations in order to inform evidence-based practices.

In conclusion, project-based learning (PBL) shows promise in diverse educational settings. However, future research must differentiate between actual and perceived learning, prioritizing the measurement of actual learning in educational interventions. This crucial distinction is often overlooked beyond a limited group of engaged faculty, impacting our understanding of PBL’s true impact and its equitable application. Integrating self-reported learning assessments, while acknowledging their limitations, complements traditional assessment tools (Pekrun, 2020). Employing a balanced approach encompassing both objective measures and self-perceptions holds vast implications for education, including health education. To achieve this, educators can utilize various assessment tools, fostering equity, inclusivity, and student agency for a richer learning experience. Additionally, exploring students’ insights through qualitative measures like focus groups can provide valuable perspectives. However, it’s crucial to supplement self-reported assessments with objective measures for comprehensive learning evaluation. Understanding the differences between actual and perceived learning aids educators in tailoring teaching methods, especially crucial in addressing equity and diversity in classrooms (Payán, 2021). Employing diverse teaching and assessment methods on varied populations allows triangulation of data, leading to a more inclusive evaluation of students’ learning in contexts like health promotion.