Using self-efficacy beliefs to understand how students in a general chemistry course approach the exam process

Abstract

Retention of students who major in STEM continues to be a major concern for universities. Many students cite poor teaching and disappointing grades as reasons for dropping out of STEM courses. Current college chemistry courses often assess what a student has learned through summative exams. To understand students' experiences of the exam process, including how students prepare for an exam, take an exam, and respond to feedback from an exam, data was collected through interviews with students in an introductory college chemistry course. The interview data was analyzed using emergent coding to describe students' experiences of the exam process using phenomenography. Data indicated that students' experiences with the exam process could be categorized based on their reported exam performance. Overall, differences could be seen between these students' self-efficacy beliefs and metacognitive skills based on the grade each student reported receiving on the exam. The students who performed highest on the exam had self-efficacy beliefs primarily from their mastery experience, and middle performing students' self-efficacy beliefs came from vicarious experience. The lowest performing student had low self-efficacy beliefs. Students who received the highest grades on the exam viewed learning as making meaningful connections between topics, and students who received lower grades viewed learning as memorization. By further understanding students study habits, their views on the exam process, and the development of their self-efficacy beliefs, instructors may be better able to assist low and middle performing students in our general chemistry courses. The findings from this study suggest several ways instructors could facilitate more effective studying and promote higher self-efficacy beliefs, including promoting group work, talking with students about study skills, and encouraging attendance at office hours to review exam responses.

---

Introduction

Recruitment and retention of students in science, technology, engineering, and mathematics (STEM) fields continues to be a focus for many colleges and universities, with substantial resources being used to help students succeed in these fields. While there are many diverse factors that contribute to students' decisions to leave STEM majors, many students cite poor teaching and disappointing grades in introductory science courses, such as general chemistry, as a reason for dropping out of STEM programs (Seymour and Hewitt, 1997). Many college chemistry courses assess what a student has learned through summative exams. The pedagogical ecology of college chemistry courses, particularly at the introductory level, is such that summative exams tend to be the primary way to determine a student's grade and are also designed to be exceptionally difficult (Grunert and Bodner, 2013). Students are often unaware of how low the class average is and equate their exam score with a failing grade when that is often not the case. The problematic assumption that exams are accurately measuring their chemistry knowledge can lead to plummeting self-efficacy beliefs and decisions to leave STEM majors for fields where students believe they will be more successful based solely on first-year experiences.

Many studies approach the idea of first-year college student performance using measures of self-efficacy (e.g., Chemers et al., 2001; Hull-Blanks et al., 2005; Dalgety and Coll, 2006). According to Chemers et al. (2001), students with high self-efficacy saw the college experience as an exciting challenge rather than as a threat. Therefore, students with higher self-efficacy worked harder and persisted longer in their studies than students with lower self-efficacy. Dalgety and Coll (2006) looked more specifically at self-efficacy with regards to first-year science students in chemistry and found that students were generally confident about the tasks that were required in a first-year chemistry course. This study found an increase in students' confidence from the first semester to the second semester, which was attributed to less confident students dropping out rather than an overall increase in confidence. Students' self-efficacy beliefs were found to be related to prior learning experiences, such as prior achievement in high school. Other studies (e.g.Vuong et al., 2010; Wright et al., 2012) found that course self-efficacy beliefs were an indicator of future persistence and academic success for college students.

Numerous studies have demonstrated the relationship between self-efficacy beliefs and performance on a specific assessment (i.e., Zajacova et al., 2005; Galyon et al., 2011; Grunert and Bodner, 2011). Galyon et al. (2011) explored the relationship between self-efficacy beliefs and variables like exam score, course grades, and GPA. From this study, students with low and medium self-efficacy beliefs were significantly more likely to be in the low exam performance group than in the high performance group. In contrast, students in the high self-efficacy group were more likely to be in the high, rather than low, exam performance group. However, this study showed no significant difference between levels of self-efficacy beliefs and GPA levels. The authors showed that there may be a weak correlation between self-efficacy beliefs and product variables (i.e., exam scores), and only when self-efficacy is high is it likely to impact academic performance. This implies that interventions to improve self-efficacy beliefs may be most effective for the low and medium self-efficacy groups.

Most literature on the exam process, including how students prepare for an exam, take an exam, and respond to feedback from an exam, focused on how student variables, such as self-efficacy beliefs, motivation, and score prediction, affected the outcome of exam performance using quantitative research methods. There are very few studies that examined the exam process qualitatively rather than quantitatively (e.g.Scheja, 2006; Brown and Wang, 2013). Although it appeared Scheja (2006) had the goal of documenting the exam process for students, the data analysis focused primarily on students who were falling out of phase with their studies. Brown and Wang (2013) asked focus groups of university Hong Kong students to draw pictures of assessment to understand students' perceptions of the assessment's purpose. This study found that university students demonstrated many negative emotions about exams. Although both of these studies are important contributions to understanding students' perceptions of the exam process, it also demonstrates that there is a lack of literature on students' experiences with the exam process, including how they navigate preparing for an exam, taking an exam, and responding to feedback from an exam in an introductory course.

Given the complex role of exams in both assessment of learning, students' self-efficacy beliefs', and retention of students in STEM fields, we decided to explore students' perceptions of the exam process. There is a need for these types of studies not only in introductory chemistry courses, but STEM courses broadly and chemistry courses specifically. However, we have a unique opportunity in introductory chemistry courses to instruct students at the beginning of a variety of STEM programs. The goal of this exploratory study was to develop an understanding of how college students approach the exam process in a general chemistry course. We are defining the exam process as how the students experience the entire exam, including how they prepare for an exam, take an exam, and respond to feedback from an exam. Because literature demonstrates that many factors have effects on exam performance, they also will have an effect on the exam process. Therefore, this exploratory study focused on the students' perceptions and experiences with the exam process guided by the following research questions.

(1) How does a student approach the exam process (i.e., preparing for an exam, taking an exam, and responding to feedback from an exam)?

(2) How do students' self-efficacy beliefs affect their experience with the exam process?

Through the re-iterative analysis process, we allowed themes to emerge from the student stories. As we worked with the data, self-efficacy beliefs emerged as one important theme during the coding process. This led us to choose self-efficacy beliefs as the guiding framework to explore students' exam processes as presented here. Bandura (1997) cites four sources that contribute to self-efficacy beliefs: mastery experience, vicarious experience, social persuasion, and psychological and affective state. Mastery experience is how students' previous successes or failures contribute to their self-beliefs. Vicarious experience is how students' experiences with their peers may affect students' perceptions of their abilities. Social persuasion is how the positive or negative comments or feedback from others can affect students' perception of their ability. Psychological and affective state deals with how students' perceive signals from their bodies (i.e., if they are nervous this might cause them to perspire) to assess their ability to do a certain task. Self-efficacy beliefs are exhibited in many different ways across a variety of tasks. In this study, we are looking at how self-efficacy beliefs relate to students' exam process. The effect of these sources of self-efficacy on students' exam process will be discussed in the results section.

Methods

We recruited and subsequently collected data from a convenience sample of students enrolled in the first semester of a two-semester general chemistry course at a large, Midwestern university in the USA. Students were recruited from two sections with different instructors. One course was an Honors section of general chemistry, and the other was non-honors section of the same course. Although the instructors and sections were different, the two courses covered the same topics and were assessed using similar multiple-choice summative exams.

The co-authors created the interview questions to encompass all aspects of the exam process. We also received feedback from other chemistry education researchers. These questions were used to pilot the interview protocol and then the questions were adjusted to finalize the protocol. Each of the two interviews with Group A lasted approximately 30 minutes and the interview with participants in Group B lasted approximately 45 minutes to one hour.

Students who agreed to participate in the study were asked to fill out an online demographic survey. This survey was used to narrow the participants to 10 students to interview. We arranged the participants into two interview groups while trying to balance the demographics in each group for age, race/ethnicity, and major. We also considered each student's self-reported course grade and the section of the course they were enrolled in (honors or regular). For example, we interviewed one female first-year chemistry major who had self-reported an expected course grade of a C in her chemistry class in Group A, and likewise we interviewed a female first-year chemistry major with an expected course grade of a C in Group B. We ultimately interviewed eight students between the two interview groups (Table 1) due to some students being unresponsive during the interview scheduling process. We found saturation in the interview data for the middle and highest performing groups because we heard similar responses from students in each unique group. We did not achieve saturation in the lowest performing group because we only had one participant from this group. We believe that we did not have more failing students in the sample because interviews were conducted in November when many low performing students may have already dropped the course.

Table 1 Summary of demographics (study ID, gender, class year, major, and predicted course grade) for the two interview groups

Group A					Group B
A1	Female	First-year	Chemistry	B	B1	Female	First-year	Chemistry	B
A2	Male	First-year	Engineering	BA	B2	Male	First-year	Engineering	BA
A3	Female	First-year	Biochemistry	A	B3	Male	First-year	Engineering	B
					B4	Male	First-year	Physics	A
					B5	Male	Third-year	Biology	A

---

The two groups only differed in the number of interviews that they were asked to participate in (Table 2). Group A was interviewed twice, once before an exam to ask about their preparation before the exam and once after an exam to ask questions about reflecting on taking the exam and receiving feedback from the exam. Group B was interviewed only after the exam to ask about the entire exam process from start to finish. Students received a $20 gift card in exchange for participating in the interviews.

Table 2 Interview schedule for the two interview groups

	Before exam	After exam
Group A	Pre-exam interview about exam preparation	Post-exam interview about reflecting on taking exam and exam feedback
Group B		Post-exam interview about the entire exam process

---

The reason we separated participants into two groups was to check for researcher impact on the group that was interviewed before and after the exam. As qualitative researchers, we acknowledge that the researcher plays a role in constructing an interview and, therefore, the data, with the participant (Creswell, 2013). The rationale behind constructing two interview groups was to see how students' answers differed between the group who was interviewed both before and after the exam and the group who only had one interview after the exam. We understand that talking to students before the exam could have changed how they prepared before the exam. There was the possibility that talking to students only after the exam could have impacted the amount and type of information they were able to recall about their exam process. Dividing the interview group was our attempt to find a balance between these concerns. Overall, we did not find any observable differences between the two interview groups.

The first author conducted all interviews to maintain consistency throughout data collection. Every interview was conducted using a semi-structured interview protocol. One interview protocol focused on the pre-exam questions and a second interview protocol focused on the post-exam questions. These two protocols were used for the Group A interviews. The third interview protocol was a combination of these two protocols designed for a single interview that was used for the Group B interviews. The interview protocols are included in the appendices.

Because our focus was on the students' different experiences of the exam process, we chose to use phenomenographic research methods. After collecting interview data, each interview was transcribed and the original recordings were deleted. We read each transcript through several times to remember the context of each interview and familiarize ourselves with the content of the interview. Then, we made individual notes about the different ways students viewed exam preparation and feedback from the exam. Each unique comment was given a code. Emergent codes were reviewed to make sure they sufficiently described the students' views presented in the interviews. Once the codes were no longer changing, these codes were applied to each interview transcript using qualitative analysis software. For example, we had coded the following quotation from Participant A2 as “Previous Experience with College”.

So, I had a really hard class that forced me to re-prioritize. I'm sure if I hadn't had that course and this was me being in college for the first time, then my experience would have been totally different.

After reviewing the initial codebook, we realized that this and examples like it could be coded as Mastery Experience, one of Bandura's (1997) sources of self-efficacy beliefs. At this point, we reviewed other codes and found the other sources of self-efficacy beliefs, including Vicarious Experience, Social Persuasion, and Physiological and Affective States.

Two external auditors repeated the coding process using the final codebook. Each auditor coded one of the eight transcripts and this was used to calculate a percent agreement as described by Campbell et al. (2013). The authors describe calculating percent agreement by “dividing the total number of agreements for all codes by the total number of agreements and disagreements for all codes combined” (Campbell et al., 2013, p. 309). The percent agreement between the authors and the two external auditors was 82%.

Limitations

It is important to recognize that phenomenological traditions make no claim to uncovering truth. Rather, the primary assumption is that how the phenomenon is experienced is due to an interaction between individuals and their experiences with the external world (Marton, 1986; Marton and Booth, 1997). Therefore, phenomenography can have the “tendency to equate participants’ experiences with their accounts of those experiences” (Orgill, 2007, p. 136). This criticism suggests there is a disconnect between what a researcher would observe the participant's experience to be and how a participant perceives and describes his or her experiences. Therefore, it is important to make the distinction between students' experiences and the unique account each provides of their individual experiences.

The students interviewed in this study discussed their perceived experiences, but this does not mean that we can know definitively how they experienced the exam process. As a phenomenographic study, we relied on what each student's recounting of his or her individual experiences. We are limited to these eight students' accounts of their experiences when we discuss the phenomenon of the exam process. Therefore, one limitation is the sample size of eight participants. These participants were recruited from the entire class, but participation was based on volunteers. Therefore, these participants may not be representative of the entire population. Despite these limitations, there is value in understanding students' experiences with the exam process because exams are often used to determine students' final grades.

Results and discussion

The goal of phenomenography is to categorize different experiences of the same phenomena, the exam process. There were no observed differences seen between students' perceptions of the exam processes and their major or gender. There was not enough variability in our participants to see differences based on race/ethnicity or class year. Most of the students in this study were Caucasian/White and were in their first-year of university. The initial demographic survey that was used to place students in interview groups asked students for their mother's and/or father's highest level of education, which could indicate a student's level of social capital and comfort with the college setting (Bourdieu, 1986; Kruse et al., 2015). Participant B5, a student in the highest performing group, indicated that his parents both had advanced graduate degrees and Participant B1, from the middle performing group, indicated that her parents had high school diplomas and had taken some college courses. This difference in level of parental education could have an effect on student performance, but there was not enough information to say for certain. Overall, there was not enough data to see a pattern between parents' education level and student performance, which may be an important area for further study. However, we did observe clear correlations between students' self-efficacy beliefs and reported exam performance.

The most significant differences between students' exam processes were seen when using their reported exam performance to categorize participants. We divided reported exam performance into three performance categories: high, middle, and low. The university where this study occurred uses a modified letter grading system (i.e. A, BA, B, CB, C, DC, E, with E as a failing grade). The breakdown of participants’ performances on the exam, based on their self-reported exam grades, can be seen in Table 3. According to the data, each group had unique sources of self-efficacy beliefs for chemistry exams, which can also be seen in the table below and are explained further in the results.

Table 3 Categorizing participants by exam performance and main source of self-efficacy beliefs

Level of performance on exam	Participant	Exam grade reported	Main source of self-efficacy beliefs
High performance	A2	A	Mastery experience
	A3	A
	B5	A
Middle performance	B1	C	Vicarious experience
	B2	B
	B3	C
	B4	C
Low performance	A1	E (failing)	Lower self-efficacy beliefs than other two groups

---

Mastery experience

Bandura (1997) describes mastery experience as previous experiences students have that contribute to their perceptions of their abilities. Analysis of the student interviews revealed that there were multiple events that potentially contributed to their mastery experience. These included previous experiences as a student in high school, as a student in college, and with chemistry content.

The students in the highest performing category (A2, A3, B5) each cited previous experiences with rigorous college courses as helping them with their success in this chemistry course. Even though Participant A2 was a first semester student at this university, he had taken a course at a community college over the summer. Participant A2 said,

So, I had a really hard class that forced me to re-prioritize. I'm sure if I hadn't had that course and this was me being in college for the first time, then my experience would have been totally different.

Students with previous college experience, like Participant A2, seemed to know what to expect out of a college course. They had realistic expectations of how difficult the course would be and knew how much time to dedicate to their studies in part due to this previous college experience. Participant A2 also said,

[My first semester of college] would have been so much worse because I would have been trying to figure out how to shift everything during this semester.

This student saw that having a previous college course contributed to his students' self-efficacy beliefs in a positive way, and these students typically had high beliefs in their ability to be successful. Participant B5 said,

I think you get very good at picking up … and maybe that just comes with going to school as long as I have and having lots of experience in the classroom. You can tell when an instructor is trying to emphasize something and when they are not.

Participant B5, who had a lot of college experience, had high self-efficacy beliefs. He felt confident that he had the ability to determine what content was important due to his experience in other courses. The highest performing students agreed that it was easier to focus on what the instructor was placing importance on because of their previous experience. Participant A2 said, “For me and people who have already done chemistry, [you can say] this is the most essential part of it”.

In general, the students in the middle-performing group felt initially confident in their abilities in this chemistry course because of their high school experience. For example, Participant B2 said,

I was an honors student in high school and it was always really easy and I never had to study and here at the university, it's like … well, it's my first year, I might have to study a little bit, but it shouldn’t be too big of a deal and then I was completely wrong.

This student thought that because of her previous experience as a successful high school student that this should automatically translate to being a successful college student. In reality, she found that this experience should not necessarily have contributed as strongly to her college chemistry self-efficacy beliefs. Similarly, Participant A2 said,

I took [chemistry] back in my sophomore year of high school, so it was still from two years ago that I have to recall. It just made it so I was less stressed about studying, so I knew exactly what to study.

This quotation demonstrates that this student felt more confident in his ability to prepare for an exam because of his previous high school experience. This made him more confident in his abilities in college, for example, Participant A2 said, “I learned that a lot of the college course is what I did in high school.” This allowed his studying to also be more focused on the concepts he felt the least confident in, so he felt more confident in the material before an exam.

Overall, students discussed a variety of experiences that contributed to their high self-efficacy beliefs in this first-year general chemistry course. The self-efficacy beliefs of students in the middle-performing group were found to be too high, meaning that students were over-confident in their abilities at the beginning of the course and found that their expectations did not match their actual experiences. Other students, specifically those with previous experiences in college had high-self efficacy beliefs that were validated during their experiences in their general chemistry course. Students with high self-efficacy beliefs as a result of previous college experiences received higher exam grades and discussed feeling more confident throughout the exam process.

Vicarious experience

For this study with college chemistry students, vicarious experience meant that students were using their peers’ exam performance to judge their own perceived ability. From the interview data, vicarious experience appeared to have had an effect on students' experiences of the exam process, particularly for middle-performing students. Several students compared themselves to others in the class. The question is, did this have a positive or negative impact on their self-efficacy? For Participant A1, the lowest performing student, comparing herself with peers negatively impacted her belief in herself if she felt her peers understood more or were better prepared than she was. She said,

I don't like studying with other people because if I am studying with other people and they are understanding it more than I am or I understand it more than they do, it's not really helping me … if they understand it more than I do, then I don’t feel comfortable. In that situation I’d rather ask the teacher to help me.

Participant A1 received the lowest score on the exam. She had high confidence before the exam and felt that she would do well on the exam; probably in part due to the amount of time she spent in office hours with the instructor and tutoring sessions. This did not translate to high self-efficacy beliefs. This meant that vicarious experiences with her peers during her exam process did not have a positive effect on her self-efficacy beliefs. She did not have any strong connections with her classmates in the class. For example, Participant A1 said, “I don’t really talk to people in the class. I just go in there and watch the lecture and leave.” Her experience with peers may have had a negative impact on her belief in herself. She did not want to compare herself to her peers and felt that her peers could not add anything to her exam process to make her more successful.

However, other students, including the middle and highest performing students, saw this process of peer comparisons as having a positive impact on their own self-efficacy beliefs. For example, Participant A2 said,

Well if someone does a lot better than you than it gives you … it makes you think about ‘where did I go wrong?’ there's also that fallback to ‘I didn’t do so well and can I double check what happened?’ Generally, if I know that I did really well … then I can say ‘hey, I did really good, is it just me or did you do well too?’

In this example, Participant A2 compared his score with his engineering peers’ scores. He saw these peers as his equals, i.e., that they have similar ability to him, which meant that he used his peers’ scores as a way to judge his own performance and success. If he did better than the peers he studied with he felt better about his own performance. This had a positive effect on his self-efficacy beliefs.

Participant A3, another of the highest performing students, said, “we (my peers) have all come to the agreement that it is really helpful if you teach someone else something you have learned.” This quotation demonstrates the higher level of learning with which the highest performing students approached the exam process. These highest performing students used vicarious experience to learn the material more deeply and to give the consent more meaning during their exam process.

Participant B2, one of the middle performing students, had a similar positive experience with her peers.

Sometimes we talk in this group about how we did on the exams. For all of us, our grades kept going up through the course of the semester as we learned how to study. We were all first-year students, so it helped all of us see everyone else in our group getting better together, which helped our morale.

Overall, we saw substantial differences in the ways students relied on vicarious experiences to inform their self-efficacy beliefs. The highest-performing and middle-performing students both had positive experiences with peer comparison, which translated to higher self-efficacy beliefs. For the lowest-performing student peer comparison had a negative effect on her self-efficacy beliefs during the exam process. This was also reflected in how these students formed and used study groups, which will be explored further in the next section.

Study groups

Another difference was seen between the students' use of study groups. The middle performing and highest performing students were both using study groups during the exam process and the lowest performing student was not. Middle performing students developed study groups later in the semester after receiving unsatisfactory scores on the first exam. These students typically viewed studying as memorization, which was reflected in how they used study groups. Participant B1 described her experience with working with her study group when she said, “We went through all of the Chapter 10 problems and figured out what we understood and which things we need to remember and worked on the equations we needed to have memorized.” This is demonstrative of how the middle performing students viewed their study group experience. They would mainly focus on helping each other memorize content.

The highest performing students made their study groups right at the beginning of the semester with other students from their major or students with whom they had a lot of other classes. These students viewed studying as more than memorization; they saw the importance of making connections between pieces of knowledge. For example, Participant A3, discussed her views in the following quotation, “I think it's important, especially for retention, to make sure that you understand the view and the overall concepts and how they all inter-link because they do”. She achieved this by working with her study group to teach each other concepts. Participant A3 also said,

I feel like I’m a teacher to a point. I’m usually with a couple to four friends who all do fairly well in the class, so it is really a give and take with that because anything I don’t understand … So I can bring my own strengths to the group and use other members’ strengths when I have a weakness.

Overall, middle performing students had a different view of studying than the highest performing students and this was reflected in how they used study groups. Middle performing students viewed studying as memorizing information and therefore used their study groups to memorize content. The highest performing students viewed studying as making connections between pieces of content and used their study groups in a way that allowed them to understand and make connections between concepts.

Social persuasion

The interview data demonstrated that mastery and vicarious experiences were the most common sources of self-efficacy that were present during the students' exam process. However, there were few instances of social persuasion, which Bandura (1997) described as how outside parties can influence a student's perceived ability using words of affirmation. Conversely, negative words can have a negative impact on students' self-efficacy beliefs.

Exam feedback can be a form of social persuasion in the context of the exam process (Rittmayer and Beier, 2009). According to Rittmayer and Beier (2009), feedback that can have an impact on social persuasion should be genuine, positive and realistic. Students received different feedback depending on the section of the course they were in. The honors students received an answer key and were allowed to keep the questions from the exam. There is no social interaction associated with the answer key, so this is not by itself a form of social persuasion. However, the students in the honors section received more feedback, including an answer key, and they were able to keep their original exam questions. Better feedback may have included encouraging students to meet with the instructor, giving more detailed answer sheets with mathematical work shown, or giving exams where the instructor is able to give more feedback. Students in the non-honors section of the course did not receive an answer key and were not allowed to keep the questions that were on the exam. The students in the non-honors course, especially the middle performing students, were the most likely to use their memory to review their responses on the exam, in the absence of instructor feedback. For example, Participant B3 said,

I basically go back to the chapter notes and figure out … thinking back to the problems, but you can’t remember a lot of those problems by the time you are trying to figure out the grade that you got, unfortunately, but trying to look back and figure out what concepts or what problems I probably messed up on.

Students, like Participant B3, who received middle or average grades on the exam, were also least likely to seek out additional feedback on an exam. These students described trying to simply remember what was on the exam and looking through their notes to find what the right answer was. They often cited not knowing where their instructor's office was or when office hours were as reasons to not seek out additional feedback, thereby missing out on social persuasion that could have impacted their self-efficacy beliefs.

Participant A1, who was the lowest performing student, discussed an email exchange with the course professor after receiving a poor grade on an exam. She said,

When I got that score I emailed my instructor and asked, “how bad does the score affect my overall grade in the class?” And so far I am not failing, I still have a passing grade for the class and if I do really good on my final I can get an even better grade, so our instructor said that when I am going to take my finals, don’t go in saying that “I can’t do it”, go in thinking that I can do it.

These positive words added to the student's self-efficacy beliefs as she prepared to take the final. She had obviously not lost hope that she could receive a good grade on the final that would give her a better grade in the course overall.

The lowest performing student and the highest performing students sought out feedback from their instructors, but for different reasons. This feedback, sought out during office hours, was a possible form of social persuasion that could have an effect on the exam process. The lowest performing student (A1) discussed looking at questions because they might appear on the final. This did not affect her study strategies, which she did not adapt during the semester. She instead cited simply needing “more time” to study the same way before the next exam. The highest performing students sought out feedback to learn how to better prepare for the next exam and were not looking only towards the final. The strategy the highest-performing students used to seek out feedback seemed to have a positive impact on their self-efficacy beliefs. For example, they were better able to adjust and understand what types of questions an instructor would ask and use that knowledge to their advantage when preparing for future exams and therefore felt more confident in their abilities.

Psychological and affective states

According to Bandura (1997), students also judge their ability based on what their bodies and emotions tell them. In general, psychological and affective states seemed to have the least effect on students' exam processes. This component of self-efficacy beliefs may show itself more when students discuss taking the exam, which was the hardest component of the exam process to ascertain in this study. In general, students described trying to stay calm during exams and not to be thrown by a single item on an exam. Participant B3 said,

It's kind of like … personally I try not to freak out when I don’t know what something means, but sometimes I can’t help but think “geez, if this is what this problem looks like what is the rest of the exam going to look like?” Most of the time, if I come across a problem like that [that I’m struggling with], I’ll just skip that problem and I’ll do the ones that I’m really sure about instead and then I can think more clearly and focus on the ones I’m not sure about.

Some students mitigated this worry by keeping a set routine on exams. Participant B5 said,

I have a very, very set routine with every single exam that I take. The first thing I do is … well at first the instructor doesn’t want you to look at it, then when I finally can, I flip it over and put my name and form letter at the top and then I stop and close my eyes and take about three deep breaths and I read the first page.

Overall, the highest performing students had the most to say about taking the exam because they had a very proscribed way of taking the exam. There were not many other instances in the data of students talking about taking the exam in detail. This could be, in part, because students had taken the exam about a week before in some cases and were no longer thinking about that at the time of the interview.

Conclusions

Some general differences in self-efficacy beliefs could be seen between the students interviewed that were correlated to their performance. Students in the highest performing group demonstrated characteristics that suggested high self-efficacy beliefs, primarily based on previous mastery experiences. These high performing students' beliefs in their abilities most closely matched with reality when they received their exam grades. High performing students were also most likely to adapt their study strategies based on learning more about what types of questions or what type of content the instructor focused on and to utilize an exam-taking routine to help manage stress and anxiety. These students occasionally sought out additional feedback from their instructor. The middle performing students received low grades on the first exam of the semester, which pushed them to re-evaluate their exam processes. They began to work in study groups following poor performances on the first exam, which led to vicarious experiences that informed their self-efficacy beliefs. However, these students were least likely to seek out feedback from the instructor outside of class, limiting possibilities for social persuasion. The lowest performing student who was interviewed in this study sought out additional feedback from the instructor, but used this feedback differently than the highest-performing students. The lowest performing student used feedback as a way to prepare for the final, but did not use this feedback to adapt her study strategies over the course of the semester. Feedback, a possible form of social persuasion, also did not seem to have an impact on her exam process. For this student, a possible misalignment between her self-efficacy beliefs and performance may have kept her from making necessary changes to be successful in the course.

(1) How does a student approach the exam process (i.e., preparing for an exam, taking an exam, and responding to feedback from an exam)?

Students' exam processes were categorized by performance. Students who received the highest grades on the exam viewed studying as making meaningful connections whereas middle performing students and the lowest performing student viewed studying as memorization. Students also differed in how they used study groups based on their exam performance. The highest performing students made study groups to teach each other the material. The middle performing students made study groups with other middle performing students after receiving an unsatisfactory grade on an earlier course exam. They used these study groups to memorize content and work through difficult problems together. The lowest performing student did not work with her peers during the exam process. Finally, students received different feedback based on what course they were in. The honors section was able to keep the exam questions and received the correct answers with their score and the regular section only received a score. This had an effect on how students used feedback on the exams. Differences were also seen in which students sought out instructor feedback, with both the highest and lowest performing students meeting with their instructors. Middle performing students in the regular course who did not receive feedback were not likely to seek out additional feedback from the instructor.

(2) How do students' self-efficacy beliefs affect their experience with the exam process?

The highest performing students main source of self-efficacy beliefs was mastery experience. These students had all taken a rigorous college course before this chemistry course, and they used these previous experiences to inform their self-efficacy beliefs and be successful in this course from the beginning of the semester. The middle performing students had vicarious experience as their main source of self-efficacy beliefs at the time they were interviewed. These students received a lower grade than they expected on the first exam of the semester, which caused them to re-evaluate their self-efficacy beliefs and their exam process. These students made study groups with other middle performing students, which reportedly improved the grades of everyone in their group and helped inform each other's self-efficacy beliefs. The lowest performing student was overconfident in her exam ability, despite lack of mastery experience or vicarious experience. She had self-efficacy beliefs that did not match her course performance. She was not using her previous experiences to adapt her exam process and did not work with her peers on course material. Feedback from the instructor was the primary factor contributing to her self-efficacy beliefs.

Implications and future studies

This study is potentially valuable for current college chemistry instructors. These instructors have a unique opportunity when working with first-year college students who are potential science majors. These courses tend to have large dropout rates and it is important to understand these courses to improve retention in STEM courses (Watkins and Mazur, 2013). By further understanding these students and their views of the exam process (i.e., preparing for an exam, taking an exam, and responding to feedback from an exam), we may be better able to assist low and middle performing students in our general chemistry courses. One main difference between the lowest performing student and the middle and highest performing students in this class was the use of study groups. For middle performing students, study groups contributed to their self-efficacy beliefs through the effect of vicarious experiences. The highest performing students created these groups early in the semester based on others they had met from their intended major. The middle performing students created these groups out of a sense of necessity after receiving unsatisfactory grades on their early exams. It is our recommendation that instructors help students get in the habit of forming groups by incorporating small group activities into their courses to help bolster students' self-efficacy beliefs. With the limited instances about social persuasion and physiological and affective states as sources of self-efficacy beliefs during the exam process, this may be a relevant area for future study.

Helping students develop self-efficacy beliefs that are in-line with their abilities is important for their success in chemistry (Pazicni and Bauer, 2014). Recognizing that there is a mismatch between exam performance and self-efficacy beliefs led the middle-performing students in this study to change their study habits. The lowest performing student in this study did not recognize there was a mismatch between her self-efficacy beliefs and her exam performance, so she continued to study in the same ineffective ways. Utilizing exams to help students more accurately assess their learning and inform their self-efficacy beliefs, in addition to other interventions, might help students to be more successful.

We also suggest that instructors consider how summative exam grades are perceived by students, especially given that grades are often equated with knowledge and self-worth. Discussing the role of exams with students and the instructor's views of the exam process and grades could help alleviate some students' decision to leave STEM fields based on their experiences in introductory chemistry. It is also helpful for instructors to recognize that some students have access to more social capital, through previous college experiences they or their family members have had, which can influence their experience with summative exams and college courses. Being explicit about the role of exams and performance expectations can help students navigate a new and unknown environment.

There are several ways this work could be continued in the future. This exploratory study was designed to characterize the current state of how students view the exam process. This study looked specifically at an introductory chemistry course. This study could be repeated in a longitudinal fashion in a similar chemistry course to look more authentically at how students change their exam process in real-time over the course of a semester. This would potentially mitigate one weakness of this study, which is that a student was asked to talk about their exam process after they had experienced the phenomena. Additionally, data could be collected using open-ended surveys to capture a larger range of experiences by sampling more students. This sample could potentially be more representative of the class as a whole. Also, the data was collected late in the semester, and we suspect that the lowest performing students may have withdrawn from the course prior to study recruitment. Starting data collection earlier in the semester might capture a wider range of student experiences by including more students at risk of failing. Future research could also focus on looking at the exam process in other introductory or higher-level science courses. Is the exam process the same in an introductory biology course as it is in a general chemistry course? Do students who are chemistry majors and are enrolled in an organic chemistry course view the exam process differently than these general chemistry students?

In summary, this exploratory study attempted to characterize the current state of the exam process in a course that relies heavily on summative exams. The exam process has typically not been viewed through a qualitative research lens. However, value can be added to current existing quantitative literature with these qualitative data highlighting students' perceptions and experiences with the summative exam process.

Conflicts of interest

There are no conflicts to declare.

Appendix A: before exam interview protocol for group A

Guiding question:
How does a student approach taking an exam?
Possible questions:
1. What thoughts come to mind when you think about taking an exam? 2. How do you prepare for taking an exam in general? 3. How have you prepared for taking this exam in your general chemistry course? 4. How often have you studied? 5. How long did you study? 6. Who have you prepared for this exam with? 7. How successful do you think you are going to be on this exam?	In a student's answers, I will be looking for the following topics in responses: • Amount of time preparing for exam • Recurrence of exam preparation (i.e. how often) • Types of materials reviewed to prepare • Use of materials to prepare • Solution when a question or problem arises during preparation • Environment where preparation occurs

Appendix B: after exam interview protocol for group A

Guiding question:
How do students respond to the results of their exam?
Possible interview questions:
1. Think back to the day of the exam, what did you do before the exam? 2. What thoughts/emotions/feelings did you have as you were taking this exam? (i.e. What were your thoughts as you read the first question on the exam?) 3. What are your thoughts on the results of your exam? 4. What does your score mean to you? 5. What feedback did you receive? If you didn’t receive feedback, will you seek it out? 6. How do you determine how well you did on the exam? 7. Do you compare your performance to the performance of your peers? 8. How did you respond to the results of your exam?	In a student's answers, I will look for the following topics in responses: • Reflecting on teacher feedback and self-feedback • Reflecting on connection to preparation • Understanding individual feedback from instructor • Applying individual feedback from instructor

Appendix C: after exam interview protocol for group B

Guiding questions:
How does a student approach taking an exam?
How do students respond to the results of their exam?
Possible interview questions:
1. What thoughts come to mind when you think about taking an exam? 2. How do you prepare for taking an exam in general? 3. How have you prepared for taking this exam in your General Chemistry course? 4. How often did you study? 5. How long did you study? 6. Who did you prepared for this exam with? 7. How successful did you think you were going to be on this exam? 8. Think back to the day of the exam, what did you do before the exam? 9. What thoughts/emotions/feelings did you have as you were taking this exam? (i.e. What were your thoughts as you read the first question on the exam?) 10. What are your thoughts on the results of your exam? 11. What does your score mean to you? 12. What feedback did you receive? If you didn’t receive feedback, will you seek it out? 13. How do you determine how well you did on the exam? 14. Do you compare your performance to the performance of your peers? 15. How will you respond to the results of your exam?	In a student's answers, we looked for the following topics in responses: • Amount of time preparing for exam • Recurrence of exam preparation (i.e. how often) • Types of materials reviewed to prepare • Use of materials to prepare • Solution when a question or problem arises during preparation • Environment where preparation occurs • Reflecting on teacher feedback and self-feedback • Reflecting on connection to preparation • Understanding individual feedback from instructor • Applying individual feedback from instructor

Acknowledgements

The authors wish to thank all of the participants for their involvement in this study. The Western Michigan University Department of Chemistry and the Mallinson Institute for Science Education supported this project, in addition to the support of Dr James Kiddle and Dr Elke Schoffers. Finally, thank you to all of our research group members for their support throughout this study.

References

1. Bandura A., (1997), Self-efficacy: the exercise of control, New York: W.H. Freeman.
2. Bourdieu P., (1986), The forms of capital, in Richardson J. (ed.), Handbook of theory and research for the sociology of education, New York, NY: Greenwood, pp. 241–258.
3. Brown G. T. L. and Wang Z., (2013), Illustrating assessment: how Hong Kong university students conceive of the purposes of assessment, Stud. High. Educ., 38(7), 1037–1057.
4. Campbell J. L., et al., (2013), Coding indepth semistructured interviews: problems of unitization and intercoder reliability and agreement, Sociological Methods Research, 42(3), 294–320.
5. Chemers M. M., Hu L.-t. and Garcia B. F., (2001), Academic self-efficacy and first-year college student performance and adjustment, J. Educ. Psychol., 93(1), 55–64.
6. Creswell J., (2013), Qualitative inquiry and research design: choosing among five approaches, Los Angeles: SAGE Publications.
7. Dalgety J. and Coll R. K., (2006), Exploring first-year students' chemistry self-efficacy, Int. J. Sci. Math. Educ., 4, 97–116.
8. Galyon C. E., Blondin C. A., Yaw J. S., Nalls M. L. and Williams R. L., (2011), The relationship of academic self-efficacy to class participation and exam performance, Soc. Psychol. Educ., 15, 233–249.
9. Grunert M. L. and Bodner G. M., (2011), Finding fulfillment: understanding women's self-efficacy beliefs and career choices in chemistry, Chem. Educ. Res. Pract., 12(4), 420–426.
10. Grunert M. L. and Bodner G. M., (2013), Pedagogical environments in chemistry: effects on women's self-efficacy beliefs, 2013 ASQ Advancing the STEM Agenda Conference Proceedings, http://asq.org/edu/2013/04/change-management/pedagogical-environments-in-chemistry-effects-on-womens-self-efficacy-beliefs.html?shl=112196.
11. Hull-Blanks E., Kurpius S. E. R., Befort C., Sollenberger S., Nicpon M. F. and Huser L., (2005), Career goals and retention-related factors among college freshmen, J. Career Dev., 32(1), 16–30.
12. Kruse T., Starobin S. S., Chen Y. A., Baul T. and Santos Laanan F., (2015), Impacts of Intersection Between Social Capital and Finances on Community College Students Pursuit of STEM Degrees, Comm. Coll. Res. Pract., 39, 324–343.
13. Marton F., (1986), Phenomenography – a research approach to investigating different understandings of reality, J. Thought, 21(3), 28–49.
14. Marton F. and Booth S., (1997), Learning and Awareness, Mahwah, NJ: Lawrence Erlbaum.
15. Orgill M., (2007), Phenomenography, in Bodner G. and Orgill M. (ed.), Theoretical frameworks for research in chemistry and science education, Upper Saddle, NJ: Pearson Prentice Hall, pp. 132–151.
16. Pazicni S. and Bauer C. F., (2014), Characterizing illusions of competence in introductory chemistry students, Chem. Educ. Res. Pract., 15, 24–34.
17. Rittmayer M. A. and Beier M. E., (2009), Self-Efficacy in STEM, in Bogue B. and Cady E. (ed.), Applying Research to Practice (ARP) Resources, retrieved from http://www.engr.psu.edu/AWE/ARPresources.aspx.
18. Scheja M., (2006), Delayed understanding and staying in phase: students' perceptions of their study situation, High. Educ., 52, 421–445.
19. Seymour E. and Hewitt N., (1997), Talking about leaving: why undergraduates leave the sciences, Boulder, CO: Westview Press.
20. Vuong M., Brown-Welty S. and Tracz S., (2010), The effects of self-efficacy on academic success of first-generation college sophomore students, J. Coll. Stud. Dev., 51(1), 50–64.
21. Watkins J. and Mazur E., (2013), Retaining Students in Science, Technology, Engineering, and Mathematics (STEM) Majors, J. Coll. Sci. Teach., 42(5), 36–41, retrieved from http://www.jstor.org/stable/43631580.
22. Wright S. L., Jenkins-Guarnieri M. A. and Murdock J. L., (2012), Career development among first-year college students: college self- efficacy, student persistence, and academic success, J. Career Dev., 292–310.
23. Zajacova A., Lynch S. M. and Espenshade T. J., (2005), Self-Efficacy, Stress, and Academic Success in College, Res. High. Educ., 46, 677–706.

---