First-Year STEM College Students’ Study Strategies: Perceived Effectiveness and Use

Abstract

Effective studying is important to learn better and increase academic achievement in postsecondary education, which also holds true for the challenging content of science, technology, engineering, and mathematics (STEM). Informed by previous research, this study mainly aimed to investigate first-year STEM college students’ study habits and perceptions of the effectiveness of different study strategies, and the frequency of use of these strategies. To this end, this study employed a cross-sectional survey using the Prolific platform. The results revealed that participants use various study strategies, including more and less effective ones, generally do not study in a planned way nor believe that learning takes hard work, and also prioritize approaching deadlines. The results also showed that the participants (a) frequently use the study strategies that they think are effective, suggesting that perceived effectiveness can have an important role in students’ strategy choice, and (b) mostly use study strategies for studying only or for both studying and while learning for fun. However, the frequency of the use of strategies partially aligned with the perceived effectiveness of the strategies. Overall, these results suggest the need to further investigate the conditions under which college students find study strategies effective, which can affect their choices.

1. Introduction

Science, technology, engineering, and mathematics (STEM) fields are typically challenging for students due to the complexities involved with the content. The first year of college is crucial since it functions like a springboard for student success for the rest of college education (e.g., Hurtado et al., 2007). Therefore, it is important for STEM college students to feel successful during their early undergraduate years, which can be boosted by employing more effective study strategies. Research on effective studying has been occurring for over one hundred years (Carpenter, 2023; Morehead et al., 2016) and has resulted in numerous effective strategies for different learner groups in different learning contexts. Despite this strong research base, researchers have found that students often use less effective or suboptimal strategies when studying for course assessments. Thus, given the vulnerability of the first college year and quite challenging content of STEM college courses, it is important for first-year STEM college students to have insights into effective studying, about which we know little. Namely, questions remain as to what study strategies first-year STEM college students use, how they use them, and why they do so while dealing with the challenging content they need to learn. Gaining such insights can help in the design and development of effective curricula, learning experiences and tasks, and interventions that can encourage STEM college students to adopt and use more effective study strategies to increase their motivation and academic achievement. As a result, the current study examined first-year STEM college students’ study habits and study strategies with a focus on perceived effectiveness and frequency of use.

1.1. Effective Studying and STEM

There is an increasing level of concern about the quality of undergraduate education, especially in STEM fields (Hora & Oleson, 2017). Many concerns revolve around lower persistence rates of underrepresented groups in STEM (e.g., Bernacki et al., 2020; Chang et al., 2014). Students leaving STEM fields often report that they are lacking the skills needed to be successful in STEM courses and careers (Bernacki et al., 2020). Of particular importance to student success and motivation is how students perform in gateway STEM courses (Perez et al., 2014). In this sense, early college years involve important facilitators and barriers to student success (e.g., Hurtado et al., 2007). One way of supporting students in achieving a better evaluation of their capabilities in early STEM courses and addressing issues such as poor persistence is to help them engage in effective studying since the study strategies students use are important for their learning in STEM education (Rodriguez et al., 2018b). Many of these strategies are retrieval-based or generative as informed by desirable difficulties (e.g., Bjork & Bjork, 2011, 2023; Bjork et al., 2013) and generative learning (e.g., Brod, 2020; Mayer, 2021, 2022, 2024), thus guiding the present study. In contrast with practice that seems to ease and increase the rate of the learning process, desirable difficulties consist of practice that provides learners with difficulties, thus slowing the rate of learning (Bjork & Bjork, 2011, 2023). However, the decreased learning rate and increased difficulty pay off by resulting in better learning (Bjork et al., 2013; Kirk-Johnson et al., 2019). Desirable difficulties and generative learning strategies also relate to self-regulated learning (e.g., Usher & Schunk, 2018; Zimmerman, 2008) especially in terms of learners’ performing or using learning strategies within the scope of this study.

The act of retrieving information from memory can be a powerful learning tool (Karpicke, 2012). Practice involving the retrieval of knowledge has been demonstrated to be useful in many contexts and across various age groups; for example, it has been effective for children involved in pictorial learning (e.g., Ma et al., 2020) and for college students in analogical problem solving (e.g., Wong et al., 2019). Retrieval practice in the form of testing can have a forward testing effect (i.e., enhancing the learning of upcoming information) (e.g., Pastötter & Frings, 2019) or a regular testing effect (i.e., enhancing the learning of previously studied information) (e.g., Cho & Powers, 2019). Interestingly, previous findings suggest that testing can also enhance later free recall of non-tested content (e.g., Rowland & DeLosh, 2014). Importantly, the use of self-testing has also been found to relate to equity gaps. For example, Rodriguez et al. (2018b) found that although underrepresented students use self-testing less, when they use it, their performance becomes comparable to their non-underrepresented peers. This finding is crucial since it suggests that the use of more effective strategies like testing could help to bridge student success gaps existing between different student populations.

Furthermore, retrieval in the form of spaced or distributed practice, which refers to learning sessions that are spread out across time, has been found to be useful for learning (e.g., Benjamin & Tullis, 2010; Butowska-Buczyńska et al., 2024; Kornell, 2009). Researchers have found benefits of spaced practice for numerous content areas such as STEM. For example, Butler et al. (2014) found that spacing homework problems in addition to increased corrective feedback and retrieval practice improved students’ performance in a signals and systems course for electrical and computer engineering students. In an introductory calculus course for engineering students, Hopkins et al. (2016) employed spaced retrieval that targeted different learning objectives; the study indicated that spaced retrieval enhanced students’ short and long-term learning of mathematics. Other researchers have found spacing and self-testing to predict student learning and/or grades in biology courses (e.g., Rodriguez et al., 2018b, 2021). Overall, distributed retrieval and testing have been found to promote learning for various groups of learners using a myriad of materials, tasks, and contexts (see Dunlosky et al., 2013).

There are other more effective strategies that can also be used in STEM courses. For instance, generative learning can help students learn foundational content by helping them select important information, organize this information in working memory, and integrate new and prior knowledge in long-term memory (Fiorella & Mayer, 2015, 2016; Mayer, 2021, 2022, 2024). In other words, these strategies can be effective by helping learners make sense of the targeted content through mental reorganization and integration with prior knowledge and can also encourage the transfer of knowledge (Fiorella & Mayer, 2015, 2016). Generative learning includes the following strategies: summarizing, drawing, imagining, self-testing, mapping, self-explaining, and enacting (Fiorella & Mayer, 2015, 2016). Additionally, using concrete examples (e.g., Rawson et al., 2014), dual coding or presenting information through text and corresponding visuals (e.g., Schweppe et al., 2015), interleaving or mixing hard-to-distinguish topics in one study session (e.g., Rohrer et al., 2019), and elaboration of the targeted content (e.g., Karpicke & Smith, 2012) can also enhance STEM learning. Interestingly, combining these strategies such as spacing and interleaving (e.g., Rohrer, 2009) or generative and retrieval tasks (e.g., Obergassel et al., 2025) can also be beneficial for learning. After all, Ewell et al. (2023) found that using active or more effective study strategies (e.g., self-quizzing) is related to better exam performance in biology undergraduate courses.

1.2. The Use of Study Strategies

Despite the effectiveness of many different types of study strategies, students tend to use less effective strategies as highlighted by previous research (Hora & Oleson, 2017). For example, Kornell and Bjork (2007) and Hartwig and Dunlosky (2012) found that college students enrolled in an introductory psychology course do not always use beneficial strategies and can underappreciate more effective strategies such as testing. Karpicke et al. (2009) found that college students most frequently reported engaging in re-reading as compared to self-testing. Further, in line with previous research, Morehead et al. (2016) found that college students frequently reported engaging in study methods that were not ideal for learning. Finally, Yan et al. (2014) found that learners mostly older than 24 years of age use strategies like those of younger college students, wherein their strategy use is often motivated by deadlines and often does not include more effective strategies such as self-testing. Regarding STEM fields, students often use a mixture of effective and less effective strategies that can impact their learning (e.g., Lopez et al., 2013; Rodriguez et al., 2018b). Unfortunately, there have been fewer studies on college students’ use of study strategies informed by cognitive science and psychology in STEM, and these studies tend to focus on a limited number of STEM fields and study strategies (e.g., Chouvalova et al., 2024; Ewell et al., 2023; Rodriguez et al., 2021; Walck-Shannon et al., 2021).

One reason for the use of less effective strategies may be due to time constraints (e.g., Hartwig & Dunlosky, 2012; Kornell & Bjork, 2007) wherein students may engage in cramming that is driven by approaching deadlines (e.g., Yan et al., 2014). Rea et al. (2022) found that the most reported barriers included the time the strategies took to implement/complete, the effort the strategies took, and the concern that the strategies could not be used effectively. Ewell et al. (2023) highlighted that even though biology undergraduates may be able to fully identify effective and ineffective strategies, they may not be able to explain the effectiveness or ineffectiveness of these strategies. Other reasons could be that students are unaware of more effective strategies (e.g., McCabe, 2011) or that they may believe the strategies they use are effective for their learning and performance. In other words, the perception of effectiveness combined with less time or efficiency could further boost students’ use of less effective study strategies. Yüksel et al.’s (2024) correlational findings including positive and statistically significant relationships between perceived effectiveness and frequency of use of most strategies support this possibility.

To this end, we need further insights into STEM college students’ use of study strategies since most research has been conducted outside of STEM fields. In other words, further research is necessary to see whether STEM college students use study strategies depending on their perception of effectiveness and efficiency, and what study habits accompany them. Moreover, given that difficulties with effective studying can correlate negatively with GPA (e.g., Sauve et al., 2016), understanding why students tend to use less effective strategies and how to change this practice is crucial for improving STEM fatigue and attrition. An important first step in this endeavor is understanding which strategies students use and why. Hora and Oleson (2017) pointed out that less is known about the study habits of STEM students in postsecondary education and there is a need for such descriptive research. Rodriguez et al. (2018a) also indicated that the benefits of more effective strategies would change for college students at different levels of study. As such, the current exploratory study aimed to provide comprehensive insights into study habits and strategies used by first-year STEM students.

Hora and Oleson (2017) also critiqued previous research on study habits, skills, and strategies for not paying attention to the contexts in which they are used by learners. Specifically, whether STEM college students use more effective strategies for studying purposes only or for learning for fun, or both, is not clear. To this end, the present study focused on different contexts by examining studying as it relates to college education and to learning for fun to determine whether students transfer their study strategies to informal contexts.

Given current gaps in the literature, the four primary research questions addressed by this study included the following:

• What are the common study habits of first-year STEM college students?
• Do first-year STEM college students use study strategies while studying, learning for fun, or both?
• Is there a relationship between perceived effectiveness and frequency of the use of study strategies?
• Are there perceived effectiveness and frequency of use differences among empirically supported effectiveness levels?

2. Methods

2.1. Research Design and Participants

This study was an exploratory cross-sectional survey study (Creswell, 2014; Fowler, 2009) focused on the study strategies and habits of first-year STEM college students in the [country]. Due to the nature of the research design, the participants constituted a cross-sectional (i.e., freshman college students from STEM fields only) convenience sample (Creswell, 2014). Research data were collected from first-year college students from STEM programs across universities in the [country] who were members of the Prolific online research platform. Specifically, first-year STEM college students were asked to participate in this study within Prolific, and a filtering question asked the participants to indicate whether they were a first-year STEM college student. To this end, we also collected department and college information and double-checked them to see whether a participant was a from a STEM field.

There were 130 initial participants; however, 32 incomplete (no answers) surveys were eliminated, and 98 surveys were included in the final analyses. The incomplete surveys had various characteristics. For instance, some participants disagreed on the consent form, while some others did not answer as to whether they were from a STEM field. The students ranged from 18 to 57 years of age (M = 23, SD = 8). Thirty-nine (40%) participants reported their gender as female, fifty-five (56%) as male, two (2%) as non-binary, and one (1%) as a trans-man, whereas one (1%) participant did not indicate their gender. Participants came from various STEM fields: (a) computer science and computing (n = 13, 13%); (b) engineering (n = 22, 22%); (c) science (n = 24, 25%); (d) technology (n = 16, 16%); and (e) mathematics (n = 23, 24%).

The participants reported that they started college in 2022 (n = 1, 1%), 2023 (n = 32, 33%), or 2024 (n = 63, 64%), and two (2%) participants did not report their college start date. However, these participants answered yes to the question of whether they were a first-year STEM college student. Of those who started college in 2024, 17 (17%) did so in fall 2024. As such, it is very likely that these 17 participants reported their high school GPAs since data collection ended by October 2024. Of these participants, 1 (1%) participant indicated that their GPA was 4.00 out of 5.00, and GPA scores for the other 16 (16%) participants ranged from 3.00 to 4.00 (M = 3.75; SD = 0.32). For the remaining 81 participants, 73 (74%) had GPA scores that ranged from 2.60 to 4.00 (M = 3.22, SD = 1.20) and 8 (8%) participants did not indicate their GPA. As for AP courses taken in high school, (a) 17 (17%) participants indicated that they did not take any AP courses; (b) 15 (15%) did not answer the question; and (c) 66 (67%) indicated taking various AP courses that may or may not have been related to their college major.

2.2. Instruments

Generative artificial intelligence tools were not used in the present study. The following instrument was used for data collection purposes.

Survey

A survey (Appendix A) starting with a filtering question asking participants whether they were first-year STEM college students (those who answered no were asked to not continue) and consent form, continuing with demographic, study habits/beliefs, and study and notetaking strategy questions, and questions about Cornell notetaking was the main data collection instrument of the present study. The demographic questions asked for general demographic information including gender, age, department or college, college start date, current GPA, and previous AP course experience in high school. The study habits/beliefs component included multiple-choice and open-ended questions about how they study and take notes in relation to coursework and learning for fun. Some of the study habit/belief questions included in the survey were adapted or adopted from previous research (e.g., Kornell & Bjork, 2007; Morehead et al., 2016).

The study strategies section, adapted from Yüksel et al. (2024), asked participants about the effectiveness (using a scale ranging from 0 (not at all) to 10 (effective)) and the frequency of use (using a scale ranging from 0 (never) to 10 (always)), and we added the purpose of use (i.e., studying for exams/quizzes/tests, learning for fun, both, or no use). Our perceived effectiveness and frequency of use questions asked for general perceptions (i.e., how effective do you think it is? how often do you use it?), and strategies were worded in more detail. Previous research on more effective strategies informed this part of the survey (e.g., Dirkx et al., 2019; Dunlosky et al., 2013; Fiorella & Mayer, 2016; McCabe, 2011; Miyatsu et al., 2018). Study strategy utility (low, moderate, and high) as described by Dunlosky et al. (2013) for the different survey items was determined based on a synthesis of strategy effectiveness across several reviews (i.e., Donoghue & Hattie, 2021; Dunlosky et al., 2013; Fiorella & Mayer, 2015, 2016; Yan et al., 2024). For those strategies that involved a combination of low- and moderate-utility components (e.g., rereading with intervals in-between study sessions), they were rated as moderate–low.

We also created seven notetaking strategies presented together with the study strategies questions above, and participants were again asked about their effectiveness (using a scale ranging from 0 (not at all) to 10 (effective), and the frequency of use (using a scale ranging from 0 (never) to 10 (always), and we added the purpose of use (i.e., studying for exams/quizzes/tests, learning for fun, both, or no use). Some of these strategies included more effective aspects such as using visuals and corresponding text. Of note, after data collection, we realized that the wording of the four notetaking strategies misaligned with the purpose of use question since they were explicitly to referring to studying or learning for fun (see

Table 1. Notetaking strategies of the first-year STEM college students in our sample.

Notetaking Strategies	Used for				Effectiveness		Frequency
	Studying n (%)	Fun n (%)	Both n (%)	No Use n (%)	M 2 (SD 3)	Md 4	M (SD)	Md
I take verbatim notes in classes.	42 (43%)	-	-	35 (36%)	4.83 (3.33)	5.00	4.46 (3.80)	4.00
I take notes on important points in classes.	67 (68%)	-	-	4 (4%)	6.20 (3.01)	8.00	5.41 (3.60)	8.00
I take verbatim notes while learning for fun.	- 1	17 (17%)	-	61 (62%)	7.12 (2.52)	4.00	6.62 (2.84)	0.00
I take notes on important information while learning for fun.	-	37 (38%)	-	41 (42%)	4.90 (3.30)	7.00	4.31 (3.80)	3.00
I use visuals and corresponding verbal information in my notes.	37 (38%)	4 (4%)	27 (28%)	30 (31%)	6.30 (2.92)	7.00	5.60 (3.40)	5.00
I take notes in a way to express critical points in fewer words.	44 (45%)	5 (5%)	35 (36%)	14 (14%)	5.80 (3.35)	8.00	5.02 (3.54)	8.00
My notetaking includes summarization, paraphrasing or using my own words.	45 (46%)	3 (3%)	38 (39%)	12 (12%)	7.62 (2.54)	8.00	6.50 (3.00)	7.00

¹ refers to those questions that had a wording misalignment with the purpose of use question. ² M = mean; ³ SD = standard deviation; ⁴ Md = median.

for all items). For instance, the item “I take verbatim notes while learning for fun” already referred to its use as while learning for fun. Accordingly, we reported the results pertaining to notetaking strategies descriptively as they relate to their purpose of use as indicated in the items only, and did not use them for main analyses.

Finally, following Morehead et al. (2019), participants were asked whether they were familiar with the Cornell notetaking method and, if so, how frequently they used it on a scale ranging from 0 (never) to 10 (always). They were also asked how effective they thought Cornell notetaking was on a scale ranging from 0 (not at all) to 10 (effective).

2.3. Procedures

2.3.1. Data Collection

The online survey was created in Qualtrics and shared through the Prolific platform on 20 March 2024, and it ended on 30 September 2024.

2.3.2. Data Preparation

All data were first checked for any incomplete submissions. In total, 32 incomplete responses were removed from the data, resulting in 98 complete surveys. Incomplete submissions often occurred due to the filtering question at the beginning of the survey asking participants whether they were a first-year STEM college student. We also double-checked whether participants were in STEM fields based on department and/or college information provided through the survey and checked for any missing points or unanswered scale questions. Less than 5% (i.e., 3%) of the data were missing, which allowed for the application of any techniques to replace missing points (Tabachnick & Fidell, 2013). As such, these missing data were replaced with the mean of each question following Gerolimatos et al. (2012).

There were no outliers or extreme scores, and the frequency of use and perceived effectiveness data did not violate the normality assumption based on the Kolmogorov–Smirnov test (p > 0.05) except for high-utility (effectiveness and frequency) and moderate–low (effectiveness) strategies. Finally, since there were different numbers of strategies pertaining to each empirically supported effectiveness level, effectiveness and frequency ratings were standardized.

2.3.3. Data Analysis

We analyzed the data based on the four primary research questions. The first two research questions were answered descriptively by presenting participants’ answer choices and percentages following previous similar research (e.g., Hartwig & Dunlosky, 2012; Kornell & Bjork, 2007; Morehead et al., 2016; Yan et al., 2014). The third and fourth questions were answered through correlational and one-way analysis of variance (ANOVA) tests.

3. Results

3.1. Participants’ Study Habits/Beliefs

Table 2. Study habits/beliefs of the first-year STEM college students in our sample.

Study Habits/Beliefs	Kornell and Bjork (2007)	Hartwig and Dunlosky (2012)	Yan et al. (2014)	Morehead et al. (2016)	Present Study
Would you say you study the way you do because a teacher (or teachers) taught you to study that way?
Yes	20%	36%	40%	36%	30%
No	80%	64%	60%	64%	68%
What time of day do you most often study?
Morning	-	<1%	-	4%	12%
Afternoon	-	11%	-	20%	16%
Evening	-	69%	-	57%	41%
Late night	-	20%	-	18%	31%
How do you decide what to study next?
Whatever’s due soonest/overdue	59%	56%	75%	63%	63%
Whatever I feel I am doing the worst in	22%	24%	12%	9%	18%
Whatever I find interesting	4%	5%	3%	4%	4%
Whatever I haven’t studied for the longest time	4%	2%	3%	3%	0%
I plan my study schedule ahead of time and I study whatever I’ve scheduled	11%	13%	7%	21%	14%
All other things being equal, what do you study more for?
Projects	-	-	-	-	8%
Multiple choice exams	22%	22%	18%	22%	24%
Essay/short answer exams	29%	20%	35%	27%	40%
About the same	49%	58%	47%	51%	28%
Learning takes time, effort, and discipline to a certain extent.
Agreed	-	-	-	-	1%
Disagreed	-	-	-	-	99%
During what time of day do you believe your studying is (or would be) most effective?
Morning	-	15%	-	17%	24%
Afternoon	-	27%	-	36%	32%
Evening	-	50%	-	40%	27%
Late night	-	9%	-	6%	17%
Do you believe you have a specific learning style (e.g., visual, auditory etc.)?
Yes	-	-	-	58%	43%
No	-	-	-	14%	14%
No, I learn best through multiple methods.	-	-	-	28%	43%
Learning must be fun or enjoyable all the time.
Agreed	-	-	-	-	24%
Neutral	-	-	-	-	55%
Disagreed	-	-	-	-	21%
Are you familiar with the Cornell Notetaking System? 1
Yes	-	-	-	-	61%
No	-	-	-	-	39%
If your answer was “yes” to the question above, how frequently do you use it?
Always	-	-	-	-	1%
Never	-	-	-	-	39%
If your answer was “yes” to the first question above, how effective do you think it is?
Effective	-	-	-	-	1%
Not at all	-	-	-	-	20%

¹ Cornell notetaking questions were added to Table 2 to save space. Please note that there are percentage differences among Cornell notetaking questions, which indicates that some participants who answered the familiarity question did not answer the other two questions.

presents the participating first-year STEM college students’ study habits (values were rounded to whole numbers). Interestingly, the majority (68%) of the first-year STEM participants indicated that the way they studied had not been instructed by a teacher(s), whereas the other participants indicated that it had been instructed by a teacher(s).

When asked what time of day participants most often study, the highest percentage reported evening (41%) followed by late night (31%). These results differed from what they indicated they believed was the most effective time to study, with the largest percentage of responses for the afternoon (32%) followed by the evening (27%), morning (24%), and late night (17%). When studying, most participants indicated that when choosing what to study next, they chose based on what was due the soonest or was overdue (63%). Additionally, the highest percentage of participants reported studying more for essay or short-answer exams (40%) compared to projects (8%) and multiple-choice exams (24%). Just over one quarter of participants (28%) indicated that they studied as much for exams as they did for projects.

To gather information on whether participants believed that learning takes time, effort, and discipline, they were asked if they agreed or disagreed with that statement; 99% of respondents disagreed indicating that they believed learning does not take time, effort, and discipline. Interestingly, when asked if they thought learning should be fun or enjoyable all the time, almost a quarter (24%) agreed, and over half (55%) were neutral. This discrepancy may indicate that the reality can be in between for most first-year STEM college students in our sample. Thus, the wording of the items, including their focus on the present tense (e.g., all the time), needed to be changed in a way that would capture what would happen more naturally; sometimes, learning may take time, effort, and discipline, and sometimes learning may be enjoyable, or both.

To gauge how prevalent the learning styles myth was amongst the participants as informed by Morehead et al. (2016), they were asked if they believed they had a specific learning style. The most common answer was either “yes” (43%) or “no, I learn best through multiple methods” (43%). Only 14% of participants said they did not believe they had a specific learning style. Finally, in contrast with Morehead et al. (2019), most participants reported that they were familiar with the Cornell notetaking method (61%), even though only 1 participant (1%) reported always using it while 38 (39%) participants reported not using it at all. Lastly, only 1 participant (1%) reported that Cornell notetaking was effective, while 20 (20%) participants indicated that it was not effective at all.

Even though participants’ answers to open-ended study habit/beliefs questions were generally short and not very informative, they revealed some interesting insights. For instance, the answers suggested that the students used a wide variety of study and notetaking strategies that can range from less effective to more effective. Regarding studying for course assessments, one participant stated, “I read it over and over again and try to picture it in my mind” while another participant stated, “I take practice exams, whenever I don’t know something, I go back and watch videos on it to learn it”. As for notetaking in class, one participant stated, “I copy what the professor says”, and another one stated, “I handwrite them and I only write the important information”. There were also students who indicated that they did not take notes in class: “I never take notes”, and “I don’t really take notes”. Likewise, generally, the participants tended not to take and use notes while learning for fun: “I have never taken notes while learning for fun”, “I don’t take notes for fun”, and “I generally do not take notes while learning for fun”.

The participants’ answers to the question of how they use their notes also included more and less effective strategies or a combination of them: “Reread and test my knowledge”, “I read over them to refresh them in my head”, and “Compile into a study guide”. As for those who may take notes while learning for fun, they use them in various ways too: “I just read them over. That’s probably the only thing I can do”, and “I find key words and study what they mean exactly”.

One common way in which the participants learned for fun turned out to be online videos that may be supported by books, social media, and other Internet resources: “I learn for it by watching YouTube videos, asking people questions online, and looking at any available books”, and “Search on the internet for information”. While learning for fun, the participants mainly thought that they did not study: “I learn for fun by mostly watching videos on YouTube. I don’t truly study it though”, and “I do not study for something I am interested in, I just read about it, and it usually sticks with me since I find it engaging”. Interestingly, despite the reality that artificial intelligence is all around, only three participants (3%) indicated that they used it: “I use ChatGPT and Google Gemini to break down the assignment into time segments and I set aside blocks of time to study using the size of the blocks and the time I have available” (when studying), “All of my classes are available online. I highlight each segment of the lecture that I am having trouble with and copy and paste it into a Google doc. I then use Google Gemini to help me find explanations to further help with my understanding of the concept” (when studying), and “Notion AI note taking app” (when taking notes).

3.2. Participants’ Study Strategies

The participants also indicated whether they used each strategy to study, to learn for fun, both, or whether they studied with them at all (see

Table 3. Study strategies of the first-year STEM college students in our sample.

Study Strategies (Empirical Effectiveness)	Used for				Effectiveness		Frequency
	Studying n (%)	Fun n (%)	Both n (%)	No Use n (%)	M 1 (SD 2)	Md 3	M (SD)	Md
I take practice tests over the target content. (high)	66 (67%)	1 (1%)	22 (23%)	9 (9%)	8.50 (2.20)	9.50	7.42 (2.70)	8.00
I test myself or use self-testing over the target content. (high)	50 (51%)	3 (3%)	34 (35%)	7 (7%)	8.11 (2.20)	8.50	7.14 (2.85)	8.00
I practice retrieving information from each class NOT immediately but after some time, and then I retrieve the same information several times in the future again. (high)	48 (49%)	3 (3%)	27 (28%)	19 (19%)	7.10 (2.80%)	7.00	5.40 (3.20)	6.00
After retrieving information from the most recent class, I also practice retrieving earlier important information. (high)	48 (49%)	4 (4%)	23 (23%)	23 (23%)	6.90 (3.06)	8.00	4.70 (3.30)	5.00
In a single study session, I work on different topics, problems, or materials especially when they are hard to discriminate but still are related. (high)	52 (53%)	3 (3%)	22 (22%)	20 (20%)	6.50 (2.85)	7.00	5.50 (3.15)	6.00
I explain visuals in my own words and compare it with the target content. (moderate)	33 (34%)	8 (8%)	31 (32%)	25 (26%)	6.31 (3.21)	7.00	5.00 (3.65)	5.00
I draw various simple to complex visuals (e.g., timelines, graphics, infographics) that correspond to the target content. (moderate)	31 (32%)	6 (6%)	24 (24%)	36 (37%)	6.30 (3.30)	7.00	4.24 (3.65)	4.00
I attempt to create mental images of the target content I read or listen to. (moderate)	23 (23%)	8 (8%)	50 (51%)	16 (16%)	6.70 (2.83)	7.00	6.07 (3.33)	6.50
I write summaries of the target content. (moderate)	40 (41%)	4 (4%)	23 (23%)	27 (27%)	6.53 (2.90)	7.00	4.80 (3.43)	4.50
I explain how things are related to each other or how I solve a problem to myself. (moderate)	36 (37%)	6 (6%)	44 (45%)	11 (11%)	7.62 (2.54)	8.00	6.50 (3.00)	7.00
I create explanations for things that are true or not. (moderate)	29 (30%)	7 (7%)	30 (31%)	30 (31%)	5.80 (3.35)	6.00	5.02 (3.55)	5.00
I create concrete examples as they relate to the target content. (moderate)	41 (42%)	8 (8%)	29 (30%)	19 (19%)	6.90 (2.75)	8.00	5.34 (3.30)	6.00
I analyze worked examples. (moderate)	44 (45%)	4 (4%)	37 (38%)	13 (13%)	4.83 (3.35)	5.00	4.50 (3.81)	4.00
I create concept maps related to the target content. (moderate)	18 (18%)	2 (2%)	12 (12%)	66 (67%)	4.03 (3.54)	4.00	2.82 (3.64)	0.00
I reread/rewatch/relisten to the target content several times with time intervals in between. (moderate–low)	50 (51%)	10 (10%)	33 (34%)	5 (5%)	7.31 (2.23)	8.00	6.63 (3.00)	7.00
I try to remember what I have read/ watched/listened to before rereading/ rewatching/relistening. (moderate–low)	33 (34%)	8 (8%)	52 (53%)	4 (4%)	7.12 (2.52)	7.00	6.62 (2.84)	7.00
I highlight important parts of sentences, paragraphs, or phrases after first reading. (moderate–low)	43 (44%)	5 (5%)	31 (32%)	19 (19%)	6.30 (3.00)	7.00	5.60 (3.40)	6.00
I underline important parts of sentences or paragraphs while reading. (moderate–low)	35 (36%)	4 (4%)	26 (27%)	31 (32%)	5.90 (3.32	7.00	4.70 (3.60)	5.00
I revisit my notes taken several times with time intervals in between later. (moderate–low)	45 (46%)	3 (3%)	29 (30%)	21 (21%)	5.80 (3.43)	7.00	4.12 (3.90)	6.00
I consecutively reread/rewatch/relisten to the target content several times. (low)	38 (39%)	4 (4%)	31 (32%)	24 (24%)	6.20 (3.01)	7.00	5.41 (3.60)	6.00
I highlight the whole sentences, paragraphs, and phrases while reading for the first time. (low)	34 (35%)	2 (2%)	24 (24%)	38 (39%)	4.90 (3.30)	5.00	4.31 (3.80)	4.00
I underline whole sentences and paragraphs while reading. (low)	26 (27%)	5 (5%)	18 (18%)	49 (50%)	4.31 (3.53)	3.50	3.63 (3.72	3.00
I use some memory aids (e.g., acronym, keywords, songs) to learn the target content. (low)	33 (34%)	3 (3%)	39 (40%)	23 (23%)	6.80 (3.00)	6.00	5.30 (3.74)	5.00
I revisit highlighted or underlined text parts later. (low)	38 (39%)	4 (4%)	32 (33%)	22 (22%)	6.03 (3.00)	8.00	5.00 (3.41)	5.00
I generally study for exams/quizzes/tests in the last few days. (low)	63 (64%)	4 (4%)	25 (26%)	5 (5%)	7.50 (2.42)	8.00	7.03 (3.01)	7.50
I review what I highlight and/or underline later. (low)	40 (41%)	3 (3%)	32 (33%)	22 (22%)	7.34 (2.50	8.00	6.40 (2.90)	7.00

¹ M = mean; ² SD = standard deviation; ³ Md = median.

). Table 3 also presents the utility categorization of the survey items about different study strategies (values were rounded to whole numbers). For the five high-utility strategy items, participants typically reported using the strategy for studying purposes (ranging from 49% to 67%) and rarely reported using them when learning for fun (ranging from 1% to 4%). These findings tended to be consistent across the other strategy items and utility ratings, with additional reports of using many strategies for both studying and while learning for fun. Interestingly, when looking across strategies and utility categories, some strategies were highly reported as not being used at all, such as creating concept maps (67%) and underlying whole sentences and paragraphs while reading (50%). When asked about the strategy of cramming (i.e., studying for exams/quizzes/tests in the last few days), most participants (64%) indicated they used it for studying purposes. Overall, participants tended to use many different study strategies for studying or for both studying and while learning for fun. The Cronbach’s alpha was 0.90 for the effectiveness questions, and 0.88 for the frequency questions.

3.3. Participants’ Notetaking Strategies

The participants also indicated whether they used notetaking strategies to study, to learn for fun, both, or whether they studied with them at all (see Table 1: values were rounded to whole numbers). Taking verbatim notes in class and taking notes while learning for fun were strategies used by fewer participants (43% and 17%, respectively) than their more effective counterparts, including taking notes on important points (68% and 38%, respectively). There were also more participants (36% for in-class notes and 62% for notes taken while learning for fun) who reported not using verbatim notetaking at all compared to those (4% for taking notes in class and 42% for while learning for fun) who reported not taking notes focusing on important points. As for the other more effective three notetaking strategies, more participants reported using them for studying purposes (ranging from 38% to 46%) and rarely reported using them while learning for fun (ranging from 3% to 4%). A significant number of participants also reported using these more effective three notetaking strategies for both studying and while learning for fun (ranging from 28% to 39%). Lastly, there were fewer participants (ranging from 12% to 31%) who reported not using the more effective three notetaking strategies. However, more participants (31%) indicated not using visuals and corresponding text compared to focusing on critical points (14%) and using their own words in the form of summarization or paraphrasing (12%). Overall, participants seemed to use various notetaking strategies for studying or for both studying and while learning for fun. The Cronbach’s alpha was 0.83 for the effectiveness questions and 0.79 for the frequency questions.

3.4. The Relationship Between Perceived Effectiveness and Frequency of the Use of Study Strategies

Several Pearson (r) correlational analyses were run separately for the whole dataset and each utility category (see

Table 4. Correlations between perceived effectiveness and frequency use across utility levels.

	Whole Data		High-Utility		Moderate-Utility		Moderate–Low-Utility		Low-Utility
	1	2	1	2	1	2	1	2	1	2
Effectiveness (1)	-		-		-		-		-
Frequency (2)	0.64 *	-	0.59 *	-	0.61 *	-	0.61 *		0.76 *	-

Note. * p < 0.001 (2-tailed). High-utility, moderate-utility, moderate–low-utility, and low-utility refer to subsets of the whole data corresponding to each utility level of learning strategies.

) to provide insight into whether participants frequently used the strategies they thought were effective. All correlations below are statistically significant and large based on Cohen’s guideline of being larger than 0.50 (Pallant, 2007, p. 132) especially for low-utility strategies.

3.5. The Perceived Effectiveness and Frequency of Use Differences Among Empirically Supported Effectiveness Levels

Two repeated-measures ANOVAs were run to test whether the effectiveness and frequency of strategy use differed across high-utility, moderate-utility, moderate–low utility, and low-utility categories. Regarding effectiveness, Mauchly’s Test of Sphericity revealed that sphericity assumption was violated, χ²(5) = 102.47, p < 0.001, with a Greenhouse–Geisser epsilon (ε) value of 0.63, and Huynh–Feldt epsilon (ε) value of 0.64. Even though multivariate statistics do not depend on sphericity (Pallant, 2007), we reported both multivariate statistics and corrected statistics here; we checked both the Greenhouse–Geisser and Huynh–Feldt corrections while interpreting the results, and the F, p, and η_p² values (23, <0.001, and 0.19, respectively) were the same for both corrections and the sphericity assumed option. The observed power for both multivariate statistics and corrections was 1. Therefore, there was a statistically significant main difference across utility categories, Wilks’ Lambda = 0.60, F (3, 95) = 22, p < 0.001, with a large effect size, multivariate η_p² = 0.40, based on Cohen’s guideline (Pallant, 2007, p. 255). The corrections resulted in a slightly bigger F value (23) but smaller practical significance (η_p² = 0.19) even though it can still be considered large based on Cohen’s guideline (Pallant, 2007, p. 255).

As for the frequency of use differences across the utility levels, Mauchly’s Test of Sphericity indicated that sphericity assumption was violated, χ²(5) = 61.08, p < 0.001, with a Greenhouse–Geisser epsilon (ε) value of 0.76 and Huynh–Feldt epsilon (ε) value of 0.78. Again, we checked and reported both multivariate statistics and corrected values here; F, p, and η_p² values (11.14, <0.001, and 0.10, respectively) were the same for both corrections and the sphericity assumed option. The observed power for both multivariate statistics and corrections was 1. Accordingly, there was a statistically significant main difference across utility categories, Wilks’ Lambda = 0.72, F (3, 95) = 12.17, p < 0.001, with a large effect size, multivariate η_p² = 0.28, as informed by Cohen’s guideline (Pallant, 2007, p. 255). The corrections resulted in a slightly smaller F value (11.14) but smaller practical significance (η_p² = 0.10) thereby resulting in a moderate or medium value based on Cohen’s guideline (Pallant, 2007, p. 255).

Pairwise comparisons with Bonferroni adjustment showed that high-utility strategies were rated more effective than moderate-, moderate–low-, and low-utility strategies (p’s < 0.001). Likewise, moderate (p = 0.007) and moderate–low strategies (p = 0.020) were rated more effective than low-utility ones. However, moderate-utility strategies were rated as effective as moderate–low ones (p > 0.05). The pairwise comparisons run in relation to frequency of use showed similar findings. Participants reported using high-utility strategies more frequently than moderate- (p < 0.001), moderate–low- (p < 0.001), and low-utility (p = 0.003) strategies. However, they reported that they use moderate-, moderate–low-, and low-utility strategies as frequently as each other (p > 0.05).

4. Discussion

This study mainly investigated the study habits/beliefs and study strategies employed by first-year undergraduate students in STEM fields in the [country]. The current results align with previous findings in interesting ways. For instance, participants often indicated that the way in which they studied had not been taught by their teachers, which is also highlighted by previous research (e.g., Hartwig & Dunlosky, 2012; Kornell & Bjork, 2007; Morehead et al., 2016; Yan et al., 2014). Like findings in other studies (e.g., Hartwig & Dunlosky, 2012; Kornell & Bjork, 2007; Morehead et al., 2016; Yan et al., 2014), participants frequently reported that they prioritized what was due soonest when it came to deciding on what to study next rather than preparing a study schedule in advance. Interestingly, almost all the participants disagreed that learning takes time, effort, and discipline to a certain extent, although only one quarter agreed that learning should be fun or enjoyable all the time. Given that factors such as perseverance (e.g., Gurung et al., 2022; Xu et al., 2021), self-discipline (e.g., Ma & She, 2023; Shi, 2024), self-regulation (e.g., Ergen & Kanadlı, 2017; Xu et al., 2021; Zhu et al., 2016), and self-directed learning (e.g., Li et al., 2022; Rashid & Asghar, 2016) are related to or can affect academic success or learning, the perception that learning does not take time, effort, and discipline is concerning.

Despite previous research findings that college students study almost equally for multiple-choice exams as they do for essay/short-answer exams (e.g., Hartwig & Dunlosky, 2012; Morehead et al., 2016), a higher percentage of participants in the present study reported studying more for essay or short-answer exams as compared to multiple-choice exams. Given that the content of STEM college courses can be quite challenging for students, students may feel that more studying time is needed to do well on their essay or short-answer exams. Furthermore, in contrast with previous research where participants chose the evening as the most effective period to study (e.g., Hartwig & Dunlosky, 2012; Morehead et al., 2016), the largest percentage of participants in the present study indicated that the afternoon is the most effective time to study. However, in line with the same previous research, participants also indicated that they studied most during the evenings. One reason for the discrepancy between perceptions of the most effective time to study and when participants studied could be that classes and other possible commitments (e.g., social events, part-time employment, on-campus employment) prevented them from studying during the day. Interestingly, such commitments take time and would also make time important for college students, thereby affecting their study strategy choices especially in terms of efficiency.

Learning styles have been questionable since 2008 (e.g., Pashler et al., 2008), and more recent work suggests that they do not exist (e.g., Kirschner, 2017). Some researchers have expressed that learning styles can be detrimental when instructional resources, including time and effort, are spent on matching teaching approaches to different learning styles (e.g., Nancekivell et al., 2019) and when they lead to erroneous assumptions about academic capacity or potential (e.g., Sun et al., 2023). Fortunately, in the present study, the results indicated that a large percentage of first-year STEM college students believed they learned best through multiple methods, and some reported that they did not have a specific learning style. This finding contrasts with Morehead et al. (2016)’s finding that most college students believed they had a specific learning style. As such, most of the first-year STEM college students in our sample seemed to be cognizant that they may not have had a specific learning style.

When examining notetaking behaviors, most of the participants reported being familiar with the Cornell notetaking tool in contrast with Morehead et al. (2019) but either don’t use it or don’t think it is effective. One reason would be that Cornell notetaking can encourage effective studying by including retrieval, which may increase perceived mental effort but decrease perceived learning (e.g., Hui et al., 2022), and summarization, which may increase mental effort (e.g., Olive & Barbier, 2017). So, a sense of increased mental effort possibly together with a sense of decreased learning would be something most participants in our sample would avoid. Another reason the participants may not have used this type of notetaking is because if they believed as many reported that learning does not take time, effort, and discipline to a certain extent, they may not have believed this method was needed to be successful.

As for the study strategies based on their utility, we found that participants frequently used strategies across the different utility levels, which was also the case in their answers to open-ended study habits/beliefs and notetaking questions. Likewise, the participants indicated that they also used various strategies for notetaking purposes. Namely, these answers showed that the participants used a mix of more and less effective strategies while studying and notetaking. They also tended to use the various study and notetaking strategies for studying or for both studying and learning for fun more as compared to learning for fun. This finding overall suggests that the first-year STEM college students who participated in the present study did not tend to transfer their study and notetaking strategies to informal learning contexts where they would have fun. Interestingly, most participants reported that they did not use concept maps, which aligns with their perception that they were of lower effectiveness. This finding is not surprising, given that creating concept maps can be demanding (Brod, 2020). However, previous research also suggests that concept maps can enhance learning (e.g., Nesbit & Adesope, 2006; Schroeder et al., 2018), especially for university students (e.g., Brod, 2020), and concept maps can be useful for both STEM and other fields (e.g., Schroeder et al., 2018). Thus, we need future research to focus on why first-year STEM college students may not find them effective nor use them frequently.

Likewise, most students indicated that they did not draw various visuals, which is in contrast with the benefits of quality drawing (e.g., Fiorella & Mayer, 2016; Fiorella & Zhang, 2018). Fiorella and Zhang (2018) noted that drawing can be quite tedious, and students may need help to create quality drawings. According to Schmidgall et al. (2019), drawing can be demanding since it entails both creating images and externalizing them. This insight also aligns with and can explain the finding that more participants chose not to use visuals and corresponding text while taking notes at all compared to other more effective strategies in this study. The possibly challenging nature of drawing can also relate to Ewell et al.’s (2023) finding that fewer biology undergraduate students preferred to make diagrams compared to other more effective and less effective strategies except for reading textbooks. Given that a large percentage of participants indicated that they did not receive instruction on how to learn better, they might need instruction or pretraining on these additional strategies (e.g., Fiorella & Mayer, 2016). Participants also might not have used concept maps or drawn because the use of creating mental images of the target content might have served as a proxy for drawing. This is called the imagining strategy (e.g., Leopold & Mayer, 2015), and it encourages learners to create their own mental images of the target content and can be as effective as drawing (Schmidgall et al., 2019). Still, given that visually representing content is important in STEM, our participants’ tendency not to use it and the corresponding reasons should be examined with larger samples. Finally, another possible reason why certain strategies were reported as less used by the participants is that these strategies may not be as suitable for STEM content as compared to non-STEM content.

For the low-utility strategies of underlining and highlighting whole sentences, paragraphs, and phrases, a large percentage of participants indicated that they did not use these strategies, suggesting that they might have been aware of the diminished effectiveness (having rated these strategies as less effective) despite the low mental effort these strategies may require. Furthermore, in line with previous research (e.g., Hartwig & Dunlosky, 2012; Morehead et al., 2016), cramming is used very commonly, and only 5 out of the 98 participants indicated not using it. The participants all perceived cramming to be effective, which is not surprising given that cramming can work in the short-term, leading to higher scores achieved on exams (Hartwig & Dunlosky, 2012).

Regarding the possible relationship between the perceived effectiveness and frequency of the use of different strategies, the correlation analyses revealed that the two are significantly related to one another. Large correlations between these two variables existed for all types of strategies including high-, moderate-, moderate–low-, and low-utility ones. For instance, the correlations between perceived effectiveness and frequency of use are even larger for low-utility strategies. Given that these are correlations and causal conclusions cannot be drawn, it is important to determine whether perceptions of effectiveness lead students to use certain strategies or whether the successful use of certain strategies leads to increased perceptions of their effectiveness. Such information would help in developing interventions to support students in their adoption of more effective strategies.

Interestingly, participants’ perception that high-utility strategies were more effective than moderate-, moderate–low-, and low-utility strategies suggests that the participants were aware of the beneficial effects of testing, retrieval, and interleaved practice. Likewise, participants rating moderate and moderate–low strategies as more effective than low-utility strategies aligns with research-based insights. The results that participants reported moderate- and moderate–low-utility strategies as equally effective might have been due to the addition of more effective aspects (e.g., time intervals) to less effective strategies (e.g., rereading). These results suggest that the first-year STEM college students in our sample were aware of many more effective strategies that could be used to increase successful studying, and that more and less effective aspects can be combined while studying in real life. We need more research on this combination of more and less effective studying strategies using different survey questions or other question types including scenarios with different samples in different learning and teaching contexts since they may better represent real life.

Importantly, however, the extent to which participants used those strategies they deemed more effective is questionable. Participants’ frequency of strategy use partially aligned with their perceived effectiveness of the strategies. For instance, the participants reported using moderate- and moderate–low strategies as frequently as low-utility strategies even though the former two were perceived to be more effective than the latter. However, moderate and moderate–low strategies were perceived to be equally effective, and the participants reported using moderate- and moderate–low-utility strategies as frequently as each other. These results suggest a need to further investigate (a) what makes low-utility strategies attractive to use despite their lower levels of perceived effectiveness especially compared to moderate- and moderate–low-utility ones; (b) whether moderate- versus moderate–low-utility distinction would hold true using different wording and samples; and (c) how to improve students’ perceptions and use of moderate- and moderate–low-utility strategies, as well as decrease those of low-utility strategies.

The results above should be approached carefully due to some limitations. First, this is a survey study and due to the nature of data collection through an online research platform, interview data could not be collected to dive deeper into the participants’ study habits/beliefs and strategies and the contexts in which they are used. Future research should consider employing additional data sources to gather more information about learners’ use of various study strategies. Second, the correlational insights gained between perceived effectiveness and use of study strategies are relational and do not refer to cause-and-effect relationships. As such, future studies may be designed to experimentally investigate causes of participants’ strategy use and/or perceptions of these strategies. Third, this survey study relied on self-reporting, and the question of whether participants used the strategies they indicated they used or whether they used them on a frequent basis is difficult to say with absolute certainty. As such, the researchers can provide the participants with a list of strategies to use during a study period and observe which were chosen. Fourth, this study did not cover collaborative and/or cooperative study strategies that are common in STEM education, which entails future research. Interestingly, examining the use of more effective strategies, including retrieval in a collaborative or cooperative way, would provide rich data-driven insights. Fifth, it seems that some participants indicated using multiple strategies that may look mutually exclusive at first glance (e.g., taking verbatim notes vs. taking notes on important points in class). This is an interesting overlap, and we need further research focusing on it to investigate the reasons for using such strategies and the conditions under which they are used. Sixth, this study checked the participants’ transferring study strategies from formal to informal contexts where they would learn for fun. Likewise, it was also possible that students transferred the strategies they used to learn for fun to formal contexts, which needs further research. Lastly, the sample size was relatively small even though they came from different STEM fields and thus increased the risk of generalizability. Therefore, the sample cannot represent all first-year STEM college students, even though the participants’ answers indicated significant and interesting directions for future research.

To conclude, given that the first and/or second years of college are important in students’ STEM trajectories (e.g., Meyer & Marx, 2014; Tsui, 2007; Ulriksen et al., 2017) and that more effective study strategies can contribute to first-year STEM college students’ academic performance (e.g., Hopkins et al., 2016; Voice & Stirton, 2020), it is important to ensure that students are aware of and use more effective strategies. An important step in this endeavor is understanding students’ current perceptions and use of different strategies within STEM. The findings of the present study provided insight into the perceptions and uses of different study strategies as reported by first-year STEM college students in the [country]. This insight can help us design and develop more effective curricula, learning experiences and tasks, and interventions that can lead STEM college students to adopt and use more effective study strategies, thereby enhancing their motivation and academic achievement.