A. A.
Flaherty
School of Education, University of Limerick, Limerick, Ireland. E-mail: Aishling.flaherty@ul.ie
First published on 6th April 2020
In the past twenty years there has been a surge of research on chemistry students’ attitudes, self-efficacy, self-concept, expectations, values, interest, motivation, effort beliefs and achievement emotions. This research has sought to understand how students feel when learning chemistry and how this may be influencing how they perform. However the wealth of this research has yet to be reviewed as a whole to identify its major themes and findings. This article reports on a review of 91 affective chemistry education research studies published since the year 2000. A focus of this review is to survey the methodological approaches used throughout research. The main finding of this review is that quantitative research regimes overwhelmingly dominant the landscape of affective chemistry education research. Of the studies reviewed, 85% (n = 77) are quantitative, 10% (n = 9) are mixed-methods while just 5% (n = 5) are qualitative research studies. Five overarching themes of affective chemistry education research are revealed. These themes manifest as the purposes behind these research studies which include; (i) to measure and compare affective states across various student demographics and contexts (32%, n = 29), (ii) to assess the influence of a learning intervention on student affect (30%, n = 28), (iii) to correlate measured affective states to performance in exams (24%, n = 22), (iv) to develop and validate scales for chemistry education research (10%, n = 9) and (v) to quantitatively model affective theoretical frameworks (3%, n = 3). The dominance of quantitative research regimes to investigate student affect may be challenged given the highly subjective and unstable nature of measured affective states. The findings of this review offer a series of implications for affective chemistry education which will be later discussed with a view to indicating potential directions for future affective chemistry education research.
Immordino-Yang and Damasio (2007, p. 5)
Emotion plays a powerful role in education. How people think and feel are intricately connected. Emotion cannot be divorced from cognition, even from a neurological standpoint. Evidence for this can be found in a study carried out by Damasio et al. (1990) on individuals who sustained injury to the frontal cortex of their brain. The aim of this study was to determine why these individuals could no longer make decisions that would benefit them despite having no impairment to their knowledge or knowledge access. The study involved measuring the emotional arousal of brain injured and non-brain injured individuals to various stimuli. Emotional arousal was measured using a skin conductance response technique that measures the number of sweat glands activated on their hands. The stimuli ranged from the participants being shown pictures, sometimes having to verbally respond to these pictures, taking a deep breath and hearing a hand clap. It was found that the brain injured individuals could not evoke emotion when stimulated. They could not activate affective states linked to punishment and reward which inherently affected their ability to make a beneficial decision. Immordino-Yang and Damasio (2007) argue that when this neurology research is considered in the context of education, it reveals the significance of emotional processes hidden inside the frontal cortex of the brain which ultimately determine how people learn and make decisions. It provides a glimpse into how intricately connected emotion and cognition are, and seeking to understand how emotion and cognition are linked should be a priority for those who want to support students as they learn and make decisions.
The study by Galloway et al. (2015) is one of the foremost studies in chemistry education research to explore the relationship between thinking and feeling in the chemistry lab. Following a series of interviews with students, it was found that their descriptions of how they felt in the lab influenced how they thought and behaved. Not feeling in control of what was happening was considered to reduce the extent to which students were thinking about what they were seeing and doing. This study highlights the precedence of feeling, as well as thinking in a learning environment.
Our research community has published widely on students’ attitudes, self-efficacy, self-concept, expectations, values, interest, motivation, effort beliefs and achievement emotions. We have invested considerable effort into understanding how students feel when learning chemistry and how this influences their performance. This article reports on a review of 91 affective chemistry education research studies published since the year 2000. A focus of this review is to survey the methodological approaches used throughout research. The findings of this review offer a series of implications for affective chemistry education which will be later discussed with a view to indicating potential directions for future affective chemistry education research.
The search criteria specified the inclusion of affective research on chemistry students at all levels of education. Affective research on pre- or in-service teachers, or students from science disciplines other than chemistry were not included for review. This search yielded 85 articles; 55 from CERP, 18 from JCE, 6 from JRST, 5 from IJSE and 1 from SE. It was necessary to also include 6 other articles published in 4 other journals (Research in Science Education, Assessment & Evaluation in Higher Education, Educational and Psychological Measurement Journal of Science and Mathematics Education and the Journal of Environmental Psychology) as they were found to be relevant to affective chemistry education research. Therefore, 91 articles in total were included in this review.
Once the studies that met the inclusion criteria were identified and collected, I conducted an inductive qualitative analysis under four headings. These headings included (i) the affect construct being investigated, (ii) the research approach as quantitative or qualitative, (iii) the data collection instrument used and (iv) the general findings of the research. Given the prevalence of measurement scales which measure several affective constructs simultaneously, it was necessary to designate an additional section to discuss the research which have used these multi-construct tools.
Quantitative (n) | Qualitative (n) | Mixed methods (n) | |
---|---|---|---|
Number of studies | 77 | 5 | 9 |
Table 2 details the methodologies employed for each affective construct. The sum of each methodology in Table 2 will not equate to the number of studies that are quoted to each methodology in Table 1 as some studies investigated more than one affective construct simultaneously.
Quantitative (n) | Qualitative (n) | Mixed methods (n) | |
---|---|---|---|
Attitudes | 37 | 0 | 5 |
Self-efficacy | 8 | 3 | 3 |
Self-concept | 12 | 0 | 1 |
Expectations & values | 8 | 2 | 1 |
Interest | 6 | 0 | 2 |
Motivation | 7 | 0 | 0 |
Effort beliefs | 1 | 0 | 0 |
Achievement emotions | 1 | 0 | 0 |
Multi-construct | 12 | 0 | 0 |
This review reveals five overarching themes of affective chemistry education research. As indicated in Table 3, these themes manifest as the purposes behind the research studies which include; (i) to measure and compare affective states across various student demographics and contexts (32%, n = 29), (ii) to assess the influence of a learning intervention on student affect (30%, n = 28), (iii) to correlate measured affective states to performance in exams (24%, n = 22), (iv) to develop and validate scales for chemistry education research (10%, n = 9) and (v) to quantitatively model affective theoretical frameworks (3%, n = 3).
D & C (n) | LI (n) | P (n) | S (n) | T (n) | |
---|---|---|---|---|---|
Number of studies | 29 | 28 | 22 | 9 | 3 |
Table 4 details the number of studies within each theme with respect to the various affective constructs. The sum of each theme in Table 3 will not equate to the number of studies that are quoted to each theme in Table 4 as some studies investigated more than one affective construct simultaneously.
D & C (n) | LI (n) | P (n) | S (n) | T (n) | |
---|---|---|---|---|---|
Attitudes | 8 | 19 | 9 | 3 | 3 |
Self-efficacy | 4 | 4 | 3 | 3 | 0 |
Self-concept | 4 | 3 | 4 | 2 | 0 |
Expectations & values | 10 | 0 | 1 | 0 | 0 |
Interest | 2 | 2 | 4 | 0 | 0 |
Motivation | 2 | 2 | 3 | 0 | 0 |
Effort beliefs | 0 | 0 | 1 | 0 | 0 |
Achievement emotions | 0 | 0 | 0 | 1 | 0 |
Multi-construct | 3 | 4 | 5 | 0 | 0 |
The first theme includes studies which have measured students’ affective states in respect to their demographic backgrounds and contexts. This research seeks to consider the influence of gender, experience, learning context or geographical area in which the research took place on student affect. Studies on students’ expectations and values (n = 10) and attitudes (n = 8) were common affect studies associated with this theme.
The second theme includes studies that have evaluated the impact of a learning intervention on students’ affective states. While the priority of these learning interventions is to enhance learning or knowledge acquisition outcomes, the impact on student affect as a bi-product was also sought. Examples of these learning interventions include inquiry-based, flipped and problem-based learning environments, simulations, case-study instructional approaches and real-life learning contexts. The impact of these interventions have mostly been evaluated in terms of their influence on students’ attitudes (n = 19).
The third theme includes studies that have ran correlation tests between students’ affective states and their performance in exams. Attitude was most prevalent affective construct studied throughout these correlation studies (n = 9).
The fourth theme includes nine articles that report solely on the development and validation of measurement scales specifically for implementation in chemistry education contexts. These articles describe the development and validation of the Attitude toward the Subject of Chemistry Inventory version 1 (Bauer, 2008) and its refined second version (Xu and Lewis, 2011). While a third version of this scale has recently been published, the associated study reporting on its validation included a performance correlation analysis which will be discussed later in this review (Rocabado et al., 2019). The Chemistry Attitudes and Experiences Questionnaire is designed to measure both students attitudes and self-efficacy beliefs (Dalgety et al., 2003). Another self-efficacy scale is the College Chemistry Self-Efficacy Scale (Uzuntiryaki and Aydın, 2009) which has been modified and validated to use with high school students (Çapa-Aydın and Uzuntiryaki, 2009). The Chemistry Self-Concept (CSCI) Inventory (Bauer, 2005) has also been modified and validated to use with high school students (Nielsen and Yezierski, 2015). Finally, Raker et al. (2019) recently validated the Achievement Emotions Questionnaire for use in organic chemistry (AEQ-OChem).
The fifth theme includes three studies which modelled affective theoretical frameworks using various structural equation modelling analyses. Brandriet et al. (2013) modelled Ausubel and Novak's theory of meaningful learning (Ausubel, 1968; Novak, 2010) using the refined second version of the ASCI (Xu and Lewis, 2011) and students’ exam performance. This theory of learning stipulates that for learning to be meaningful, integrating how students think, feel and perform is necessary for them to make connections between what they already know and the new knowledge they are digesting. The study by Brandriet et al. (2013) provides statistical evidence for the need to integrate these three domains if the desired instructional goal is to facilitate learning that is meaningful. Focusing on improving how students think and preform without addressing how they feel is ultimately insufficient.
The significance of how students feel is further illuminated by studies that have ran structural equation modelling analyses on students’ attitudes as interpreted by the ASCI instrument (Ross et al., 2018; Ross et al., 2019). Ross et al. (2018) investiagted the relationship between the cognitive, affective and behavoural components of the attitude construct that underpins the ASCI. This study reveals that while the behavioural component of attitude was the most positive component of attitude with students, the affective component of attitude most predicted the cognitive and behavioural components of attitude (Ross et al., 2018). In a follow up study, Ross et al. (2019) show how the influence of low involvement (as a reflection of how invested students are about the attitudinal object) on chemistry achievement was predominantly mediated by the affective subcomponent of students’ attitude.
The next part of this review will discuss the research associated with each affective construct. Given that the focus of this review is to survey the methodological approaches used throughout affective chemistry education research, emphasis will be placed on identifying and discussing the tools which research has relied on for collecting and analysing data.
The development of the Attitude toward the Subject of Chemistry Inventory (ASCI) (Bauer, 2008) was a landmark development in affective chemistry education research. Appearing in 17 articles in this review, the ASCI and its subsequent refined versions, is the most used scale in affective chemistry education research. The ASCIv1 (ASCI version 1) consists of a 20 item semantic differential instrument set out to measure students’ interest and utility of chemistry, their anxiety learning chemistry, their perceptions of its intellectual accessibility, their fear and emotional satisfaction. Brown et al. (2014) used the ASCIv1 to indicate the similarly positive attitudes of both foundation and first year undergraduate students in an institution in the South Pacific. Chan and Bauer (2014) included the ASCIv1 in conjunction with a number of other affective scales throughout various studies to account for various high, medium and low affective groupings of first year general chemistry students. Here, the high affective group performed significantly better in exams than the low affective group. The ASCIv1 has been used to show how neither a Process-Oriented, Guided-Inquiry Learning (POGIL) approach or a Peer-Led Teaching and Learning approach had a significant influence on students’ attitudes (Chase et al., 2013; Chan and Bauer, 2015).
The ASCI was refined as the ASCIv2 (ASCI version 2) by Xu and Lewis (2011). In this ASCIv2, the number of items was reduced from 20 items to just 8 items and included two subscales, “intellectual accessibility” and “emotional satisfaction”. The validity of the ASCIv2 has been further confirmed by other studies (Brandriet et al., 2011; Brandriet et al., 2013; Kahveci, 2015; Montes et al., 2018). The ASCIv2 has been used in several studies which correlate attitude and performance. When controlling for math ability, attitudes and prior conceptual knowledge contributed a significant unique portion to the prediction of chemistry achievement in general chemistry (Xu et al., 2013). Brandriet et al. (2011) found a positive relationship between students’ attitudes and general chemistry success and that females can have less favourable attitude towards the subject of chemistry than males. Brandriet et al. (2013) demonstrated the statistical evidence for Ausubel and Novak's construct of meaningful learning (Bretz, 2001; Novak, 2010) which underpins meaningful learning as the integration of thinking, feeling, and performance. Higher achieving high school students had more positive attitudes towards chemistry in both Turkey (Kahveci, 2015) and Chile (Montes et al., 2018). No significant difference was found between the influence of online or face-to-face learning environments on students’ attitudes or performance (Nennig et al., 2020). However, the ASCIv2 has indicated the efficacy of POGIL classes in significantly improving students’ attitudes (Vishnumolakala et al., 2017). Chan and Bauer (2016) included the ASCIv2 in conjunction with a number of other affective scales to show how students who had stronger positive affective characteristics were more likely to be autonomous learners, relying less on tutors when preparing for exams and understanding the notes they took in class.
More recently, two of the eight items that compose the ASCIv2 were switched to form the ASCIv3 (ASCI version 3) (Rocabado et al., 2019). This switch did not affect the validity of the internal structure of the scale. The ASCIv3 found that over the course of a flipped class course, Black female students had lower attitudes both before and after the course and also performed worse on their exams compared to their peers.
While the ASCI measures students’ attitude towards the subject of chemistry, the Chemistry Attitudes and Experiences Questionnaire (CAEQ) probes for students’ attitudes from a societal perspective also (Dalgety et al., 2003). The CAEQ has been used to show how students who have more positive attitudes towards chemistry are more likely to pursue further chemistry after their first semester of general chemistry (Dalgety and Coll, 2006). In the context of POGIL, the CAEQ has been used to show POGIL's lack of influence on students’ attitudes in one study (Chase et al., 2013) while another study showed POGIL's efficacy in promoting positive attitudes using the CAEQ (Vishnumolakala et al., 2017).
Both the ASCI and the CAEQ operate as semantic differential scales meaning that each item of these scales consists of two opposing adjectives. Respondents are required to choose a position between these adjectives. However, a Likert scale is an alternative scale which requires respondents to state how much they agree or disagree to a statement. The Attitude toward Chemistry Lessons Scale (ATCLS) is a Likert attitudinal scale developed by Cheung (2009a). The ATCLS comprises of four dimensions in evaluating chemistry students’ attitudes; liking for chemistry theory lessons, liking for chemistry laboratory work, evaluative beliefs about school chemistry, and behavioural tendencies to learn chemistry. The ATCLS has been used to show a deterioration in the attitudes of secondary school male students in Hong Kong but little change in the attitudes of their female classmates (Cheung, 2009b). Another study qualitatively reported on how a teacher reacted to their students’ ATCLS scores (Cheung, 2011).
The Test of Science Related Attitudes (TOSRA) was first developed by Fraser (1978) to measure the attitudes of 7th grade students towards science. Villafañe and Lewis (2016) used a shortened version of the TOSRA to indicate how the attitudes of students in introductory college chemistry toward inquiry and career interest in science had a small but significant influence on their exam performance. Shamuganathan and Karpudewan (2017) used an adapted version of the Environment Attitude Inventory (Milfont and Duckitt, 2010) to show the positive influence that embedding science writing heuristics in a green chemistry curriculum had on students’ attitudes. Antonoglou et al. (2011) used a scale adapted from a number of different attitudinal scales to show the positive influence that an online course on molecular symmetry had on students’ attitudes. Kousa et al. (2018) adapted the Fennema–Sherman test to measure the attitudes of low achieving students in Finish secondary schools. The Fennema–Sherman test was originally developed to measure students’ attitudes towards mathematics (Fennema and Sherman, 1976). Here, Kousa et al. (2018) shows how the low achieving students tended to have more positive attitudes towards teaching methods which included visiting companies, institutes, museums and exhibitions, using the internet, videos, magazines and books for studying.
Berg (2005) developed a scale that indicated both positive and negative changes in students’ attitudes towards learning chemistry in their first year of university chemistry education. Nieswandt (2007) developed a scale to measure German high school students’ attitude toward chemistry which found no significant effects of attitudes on students’ conceptual understanding of chemistry. A scale has also been developed to measure the attitudes of high school students in Greece towards learning chemistry (Salta and Tzougraki, 2004). Penn and Ramnarain (2019) used this scale by Salta and Tzougraki (2004) to show the positive influence of a virtually simulated learning environment on students’ attitudes.
Abdullah et al. (2009) used a scale developed by Hofstein (1976) to show no significant influences of individualized microscale chemistry experimentation on students’ attitudes. Using a scale developed by Geban et al. (1994) to measure Turkish students’ attitudes towards chemistry, Tarkin and Uzuntiryaki-Kondakci (2017) showed the positive influence of case-study electrochemistry instruction on students’ attitudes. Positive influences on students attitudes towards case-study instruction on the topic of states of matter were noted by Ayyıldız and Tarhan (2013) using a scale developed as part of a doctoral dissertation by Acar and Tarhan (2008).
The CCSS scale measures college students’ belief in their ability to perform essential tasks in chemistry with respect to three dimensions; their self-efficacy for cognitive skills, for psychomotor skills and for everyday applications (Uzuntiryaki and Aydın, 2009). The CCSS and its adapted versions have been used throughout five studies included in this review. Graham et al. (2019) used the CCSS to show how an intervention designed to promote first year undergraduate students’ metacognition when studying chemistry improved their self-efficacy. The CCSS revealed how self-efficacy for cognitive skills significantly predicted the performance of students in a South African university (Ramnarain and Ramaila, 2018). A modified version of the CCSS was used to reveal improvements in general chemistry students’ self-efficacy throughout a semester, and how the gap between chemistry majors demonstrating a higher sense of self-efficacy at the beginning of a semester compared to non-chemistry majors can narrow (Ferrell and Barbera, 2015). The CCSS was modified to assess the self-efficacy of high school students (HCSS) by Çapa-Aydın and Uzuntiryaki (2009). The HCSS was used to show how case-based instruction on electrochemistry had no significant effect on students’ self-efficacy compared to chemistry traditional instruction (Tarkin and Uzuntiryaki-Kondakci, 2017).
As well as measuring attitudes, the CAEQ is another scale that is used to measure chemistry students’ self-efficacy (Dalgety et al., 2003). The CAEQ has been used throughout three studies included in this review. The CAEQ has shown how students’ self-efficacy with respect to their beliefs about learning and applying chemistry theory can deteriorate during their first year of chemistry (Dalgety et al., 2003) and that the self-efficacy of students throughout a college preparatory chemistry course can fluctuate when their race or ethnicity is taken into account (Villafañe et al., 2014). POGIL classes and a Problem-Based Laboratory Learning unit have also been shown to improve students’ self-efficacy using the CAEQ (Mataka and Kowalske, 2015; Vishnumolakala et al., 2017).
The CSCI has been used to show that while there is a strong correlation between higher self-concept and higher scores on ACS exams, up to 20% of general chemistry students have low self-concept with just 11% indicating high self-concept (Lewis et al., 2009). Chan and Bauer have used the CSCI in conjunction with a number of other affective scales throughout various studies (Chan and Bauer, 2014, 2015, 2016). Chan and Bauer (2014) used the CSCI to partially account for various the differences between high, medium and low affective groupings of first year general chemistry students. Here, the high affective group performed significantly better in exams than the low affective group. Similar groupings were also studied in respect to the learning strategies they used. Here, the high affective group reported to understand the notes they took in lecture more than the low affective group (Chan and Bauer, 2016). The CSCI was also used by Chan and Bauer (2015) to show no significant differences in the self-concept of students in Peer-Led Team Learning compared to students in a control instructional environment t. The CSCI has now been validated for use with high school students (Nielsen and Yezierski, 2015).
The MLLI is designed to measure students’ expectations of undergraduate lab work with respect to their cognitive and affective learning domains (Galloway and Bretz, 2015a). The MLLI is underpinned by Ausubel and Novak's theory of meaningful learning (Ausubel, 1968; Bretz, 2001; Novak, 2010). The MLLI has indicated that while organic chemistry students may have lower expectations for lab work compared to general chemistry students, all students acknowledged that their undergraduate lab experiences failed to meet their expectations (Galloway and Bretz, 2015a). A follow up study with over 3500 students enrolled in both organic and general chemistry in institutions across the United States also indicated unfulfilled expectations (Galloway and Bretz, 2015b). A cluster analysis of data retrieved from the implementation of the MLLI revealed that what students expected to think and feel in their laboratory course framed what they actually experienced (Galloway and Bretz, 2015c). George-Williams et al. (2018) modified the MLLI for implementation in universities across Australia. Here, considerable disparities in the expectations of various teaching, academic and student cohorts were found. While students tended to have positive expectations for lab work which were relatively stable as time progressed, both teaching and academic staff had a pessimistic outlook on their expectations for student learning in the lab.
The CHEMX survey measures students’ cognitive expectations for learning chemistry in areas such as effort, concepts, math link, reality link, outcome, laboratory, and visualization learning (Grove and Bretz, 2007). The CHEMX survey has been used to reveal significant differences between the expectations that faculty and students have for learning chemistry as well as improvements and prolonged declines in the expectations of chemistry majors and non-chemistry majors respectively (Grove and Bretz, 2007).
Irish students’ expectations and intentions towards university life and study were explored using the Motivation, Preparedness and Expectations (MPE) tool in a study by Lovatt and Finlayson (2013). Here is was found that 64% of students believed they knew what was expected of them academically at university and 80% of students also expected to have a part time job as they fulfilled their undergraduate studies. In one qualitative study, Chopra et al. (2017) revealed students default expectations of university lab work to situate within expository modes of lab education and that students’ expectations of lab work can influence their behaviours as either recipients of instructions and knowledge, or as inquirers and producers of knowledge.
Values are closely linked to expectations in line with the widely accepted expectancy-value theory of motivation. Here, this theory posits that an individual will behave in a way if they value the outcome they expect to achieve having behaved in that way (Eccles, 1983; Eccles and Wigfield, 1995). On closer inspection, values comprise a specific subset of beliefs which are associated with what should be desired, what is important and what standards of conduct are acceptable, which influence or guide behaviour (Bauer, 2005; McCoach et al., 2013). The extent to which a student perceives the value of their learning is considered to be an important contributor to interest, performance and future plans (Updegraff et al., 1996; Jacobs et al., 2002; Hulleman and Harackiewicz, 2009).
Just two studies included in this review have sought to learn more about what students value when learning chemistry and what implicates these values. González and Paoloni (2015) adapted various expectation and value scales to reveal how students' perception of their interactions with supportive teachers enhanced their cognitive expectations for learning in chemistry and their perceptions of value associated with various aspects of learning chemistry. Nieswandt (2007) developed a scale to measure how much grade 9 chemistry students in Germany perceived the importance of chemistry as a dimension of their value beliefs. Although students in this study valued the role of chemistry in their lives, this sense of value was not strongly associated with their conceptual understanding. Overall, this review did not find any considerable or exclusive qualitative research on chemistry students’ values.
Situational interest is a type of interest that emerges spontaneously in response to features of an environment (Hidi and Renninger, 2006). Using scales developed by Harackiewicz et al. (2008), Ferrell et al. (2016) found that general chemistry students with higher levels of feeling-related situational interest performed better than those with lower levels of interest. Using a situational interest questionnaire developed by Fechner (2009), Habig et al. (2018) showed how secondary school students who had low content and subject-related individual interest were found to have higher situational interest when what they learnt was made relevant to their everyday lives. The scale developed by Fechner (2009) was also used by van Vorst (2018) to indicate the positive influence of the Ladders of Learning tool on students’ chemistry related interest. Nieswandt (2007) developed a situational interest subscale as part of an affective construct questionnaire which found that in German secondary schools, situational interest had a positive effect on conceptual understanding. Basso et al. (2018) developed an interest questionnaire that indicated the positive influences of a crime scene investigation learning activity on students’ situational interest.
Provided that situational interest is maintained, interest then evolves into maintained situation interest. Using scales developed by Harackiewicz et al. (2008), Ferrell and Barbera (2015) found that chemistry majors had higher initial and maintained interest which was sustained from the beginning to the end of a semester of general chemistry compared to non-chemistry majors. Using the RIASEC + N instrument (Dierks et al., 2014), Höft et al. (2019) revealed considerable deteriorations in students’ interests in school science activities across grades in Germany and small to middle-sized positive correlations between conceptual understanding and interest in school science activities. In a follow up study, Höft and Bernholt (2019) used the RIASEC + N instrument again to reveal that students’ students’ interest in enterprising activities predicted their subsequent conceptual understanding. Apart from just two studies that included both quantitative and qualitative research methods to explore interest (Ferrell and Barbera, 2015; van Vorst, 2018), this review did not find any considerable or exclusive qualitative research on chemistry students’ interest.
Liu et al. (2017) adapted and validated the Academic Motivation Scale (AMS) (Vallerand et al., 1992) for use in chemistry (AMS-Chem). The AMS-Chem is designed to measure various aspects of student's motivation towards chemistry. Using the AMS-Chem with general chemistry students, it was found that males can be more motivated than their female counterparts (Liu et al., 2017). It was also found that intrinsic motivation (behaviour driven by internal rewards) can be positively associated with academic achievement (Liu et al., 2017). The AMS-Chem has shown the significantly positive influence of intrinsic motivation on student achievement in organic chemistry throughout various lecture-based, flipped and peer-led team teaching and learning environments (Liu et al., 2018). Juriševič et al. (2012) developed a scale to measure different components of motivation for learning chemistry in high school students in Slovenia and Poland. Here it was found that hands-on laboratory work with supportive teachers can enhance motivation and learning. The Intrinsic Motivation Inventory (IMI) (McAuley et al., 1989) was used by Southam and Lewis (2013) which revealed the positive influence of POGIL on students’ intrinsic motivation in a third year undergraduate spectroscopy course. A study by Ogunde et al. (2017) with students in Australia, the United Kingdom and New Zealand revealed the notable influence of students’ intrinsic motivation on their decisions to pursue chemistry after their graduation from undergraduate degree programmes. A motivation scale developed by Tuan et al. (2005) and implemented by Cicuto and Torres (2016) showed that the implementation of an active learning environment which involved study periods and discussion groups between biochemistry students and teachers had a positive influence on students motivation (Cicuto and Torres, 2016). Overall, this review did not find any considerable or exclusive qualitative research on chemistry students’ motivation.
The MSLQ measures students’ self-efficacy for learning and performance, control of learning beliefs, task value, test anxiety and intrinsic and extrinsic goal orientation in respect to various learning strategies (Pintrich et al., 1993). In the context of an organic chemistry course, Lynch and Trujillo (2011) found that self-efficacy was the strongest and most consistent MSLQ factor associated with performance. Although it was found that males exhibited significantly higher levels of self-efficacy than females, it was claimed that females tend to “rise above such disabling feelings to produce acceptable work” (Lynch and Trujillo, 2011, p. 1360). Gibbons and Raker (2019) used the self-efficacy subscale of the MSLQ to test the relationship between self-efficacy and self-concept in organic chemistry. Here it was found that self-efficacy was predicted by exam achievement whereby self-concept was found to be predictive of first examination achievement in organic chemistry. Using the MSLQ, Kırık and Boz (2012) showed the positive influence that cooperative learning had on students’ task value, self-efficacy and their extrinsic and intrinsic goal orientations compared to traditional modes of teacher instruction. Zusho et al. (2003) showed prolonged declines in students’ motivational beliefs throughout introductory college chemistry. Chan and Bauer (2014) included the MSLQ in conjunction with a number of other affective scales throughout various studies to account for various high, medium and low affective groupings of first year general chemistry students. Here, the high affective group performed significantly better in exams than the low affective group. In another study, Chan and Bauer (2016) also included the MSLQ in conjunction with a number of other affective scales to show how students who had stronger positive affective characteristics were more likely to be autonomous learners, relying less on tutors when preparing for exams and understanding the notes they took in class. The MSLQ was used by Abdullah et al. (2009) to show how individualised microscale chemistry experiments only had a positive influence on students’ self-efficacy and not on any other subscale of the MSLQ.
The SMQ measures students’ intrinsic motivation and personal relevance, self-efficacy and assessment anxiety, self-determination, career motivation and grade motivation (Glynn and Koballa, 2006; Glynn et al., 2009). Using a SMQ that was translated into Turkish by Çetin-Dindar and Geban (2010), Tarkin and Uzuntiryaki-Kondakci (2017) used the SMQ to show the positive influence that case-based instruction on electrochemistry had on students’ motivation to learn chemistry.
The SMQ was followed by the Science Motivation Questionnaire-II (SMQ-II) which measures five motivation components: intrinsic motivation, self-determination, self-efficacy, career motivation, and grade motivation (Glynn et al., 2011). Ardura and Pérez-Bitrián (2019) used the SMQ-II to reveal variances in the motivation profiles of students who decided to pursue or not to pursue chemistry and physics after secondary school. Here it was found that while self-efficacy was a strong predictor of academic performance, the grade motivation and self-determination in those who drop chemistry and physics are more important predictors of performance than for those who choose to continue with the subject. Austin et al. (2018) modified the SMQ-II for use in an organic chemistry context were it was found that despite students being highly motivated by their grade, this motivation was not strongly correlated with performance. Salta and Koulougliotis (2015) adapted the SMQ-II for use in the Greek secondary school context where it was found that younger students had higher grade motivation than older students and girls had higher self-determination than boys across all age groups.
The SPiCE was developed by Winkelmann et al. (2015) to measure students’ self-efficacy and attitudes about various aspects of chemistry lab work. Here, while undergraduate research modules were shown to improve students’ confidence, these modules had less impact on their attitudes (Winkelmann et al., 2015). The SPiCE also revealed the positive influences of Argumentation-Driven-Inquiry on students’ self-efficacy (Eymur, 2018).
Mujtaba et al. (2018) developed a scale to measure students’ chemistry career aspirations, motivation in science, self-concept and their general beliefs about learning science in school. Following a survey of almost five thousand secondary school students in England (ages 11–13), it was found that students’ extrinsic motivation (perceived utility of science) had a strong association to their aspirations to continue in chemistry.
Despite all this, the mechanisms in which we can positively influence affect have not been seriously considered by chemistry education research. Many learning interventions have been shown have a positive influence on student affect (Scott, 2014; Vishnumolakala et al., 2017; Penn and Ramnarain, 2019). However, the promotion of affective states was not the utmost priority of these interventions. The primary intention of these interventions was to improve learning. Improving affective states in these studies occurred as a bi-product of the research pursuit and not through the deduction of such from affective theoretical frameworks. Chemistry education research has not explored ways to influence affect in line with relevant affect literature.
A lack of insight on explicit mechanisms to positively influence affect is not just an issue for chemistry education research (Pekrun, 2006; Usher, 2016; National Academy of Sciences, 2018). The recently published How People Learn II report devotes considerable attention on students’ beliefs, values and goals (National Academy of Sciences, 2018). Taking students’ mind-set as “the set of assumptions, values, and beliefs about oneself and the world that influence how one perceives, interprets, and acts upon one's environment” (National Academy of Sciences, 2018, p. 111), the report calls for more experimental research to determine whether interventions designed to influence mind-sets can benefit learners.
While the enhancement of emotions can be deduced from theoretical assumptions, academic and achievement emotions remain largely under-researched with respect to their dimensions, antecedents, and functions in different academic settings (Pekrun, 2006). Although there are evidence-based principles for alleviating test anxiety (Zeidner, 1998), there is a need for educational intervention studies demonstrating the ways in which we can also influence positive emotions as opposed to alleviate negative emotions like anxiety (Pekrun, 2006). However, educational interventions continue to prioritise the enhancement of skills and performance on exams (Usher, 2016). It light of this Usher (2016) argues that the field of educational psychology would be “served by research that can clarify how and under what conditions learners come to form and alter their beliefs and how those beliefs change in magnitude and influence over time.” (Usher, 2016, p. 155). These calls for research on how we can influence affect, as opposed to measure it, perhaps stand true within the realm of chemistry education research also. Despite some studies running various structural equation modelling analyses on affect theoretical frameworks (Brandriet et al., 2013; Ross et al., 2018; Ross et al., 2019), there was little indication of efforts to progress affect theory specifically in the context of chemistry education throughout the studies reviewed in this article.
There are some examples throughout literature that evidences how affect can be positively influenced towards improving performance (Hulleman et al., 2010; Rozek et al., 2019). Indicating that “the extrinsic nature of utility value [usefulness for other tasks or aspects of their life] makes it particularly amenable to situational interventions from teachers or parents”, Hulleman et al. (2010, p. 882) designed an intervention to test whether the degree to which students’ valued what they learned because it was relevant to their lives could enhance their performance. In the intervention, the treatment group of undergraduate psychology students wrote letters to a significant person in their lives describing the relevance of a psychology topic they studied to this person. The control group of students were asked to just summarise a psychology topic. Compared to the control group of students, the treatment group of students reported significantly more interest in psychology at the end of the course and this effect was heightened for students with lower exam scores. In relation to performance, the changes in students’ perceptions of utility value predicted by the intervention, led to increases in graded performance (Hulleman et al., 2010).
More recently, Rozek et al. (2019) designed a writing intervention to help students from lower income backgrounds to regulate their performance anxiety in exams. Before exams, students were asked to think about their emotions and thoughts and express them in writing. Some students were assigned writing activities that encouraged them to write freely about their thoughts, others were assigned writing activities that encouraged them to see the positives of feeling anxious before exams while some students were assigned a combination of both of these writing activities. Compared to a control group of students also from lower income backgrounds, Rozek et al. (2019) found that the students who participated in the intervention performed better in their exams and also were more able to positively appraise their test anxiety.
Beliefs, values and attitudes are inherently susceptible to change (Mezirow, 1991, 1997, 2000). Such susceptibility has been hinted at throughout chemistry education research. We know that exams influence students’ personal capability beliefs (Ferrell et al., 2016; Gibbons and Raker, 2019), and instructors too are thought to have the just as much of an influence on these beliefs as Ferrell et al. (2016) states; “An instructor cannot control what beliefs students' hold when they enter the classroom, but our results suggest that instructors could target interest and self-efficacy in their teaching strategies, which could impact course performance.” (2016, p. 1062). However, faculty could a have a more influential role, particularly in respect of deteriorating expectations for learning throughout introductory chemistry courses (Grove and Bretz, 2007). In progressing affective chemistry education research, we may be encouraged to consider how we can influence affect as opposed to developing and implementing scales to measure affective states.
In 1931, Louis Leon Thurstone wrote; “An attitude is a complex affair which cannot be wholly described by any single numerical index. For the problem of measurement this statement is analogous to the observation than an ordinary table is a complex affair which cannot be wholly described by any single numerical index. So is a man [sic] such a complexity which cannot be wholly represented by a single index. Nevertheless we do not hesitate to say that we measure the table.” (Thurstone, 1931, p. 260). Following similar sentiments, Schwarz (2007) claims that the empirical analysis of attitudes using structural equation models and estimating it from multiple attitude measures is theoretically inconclusive as; “Any observed similarity of responses across attitude measures suffers from the same ambiguity as observed similarity of responses over time or consistency of judgment and behaviour. In each case, the observed similarity may either derive from a “true” enduring attitude or may merely reflect that respondents arrived at similar judgments for the reasons discussed above.” (Schwarz, 2007, p. 649). As such, “there is no empirical answer to whether people “have” attitudes or construct evaluative judgments on the spot” (Schwarz, 2007, p. 649). Krosnick et al. (2005) contend that what gets measured in attitude research is not actually attitude at all, rather it is the manifestation of attitude as the product of complex cognitive processes that gets measured. This should concern chemistry education research since only a small number of studies on chemistry students’ attitudes have sought to include qualitative research in the form of mixed methods research. This review did not find any exclusively qualitative studies published on chemistry students’ attitudes.
Issues are also thought to arise for the empirical analysis of personal capability beliefs (such as self-efficacy and self-concept). Given that the development and operation of an individual's personal capability beliefs are so unique to every individual, Usher (2016) claims that “The utility of testing complex analytic models should be weighted against their potential benefit to theory and practice” (2010, p. 154). Despite the wealth of chemistry education research on self-efficacy and self-concept, the relationship between these two self-belief constructs is ‘murky’ (Marsh et al., 2019). A recent study provides evidence that generalized self-efficacy, outcome expectations and self-concept, are essentially indistinguishable (Marsh et al., 2019). These authors claim that in order to respond to a generalised self-efficacy item (ex, ‘I am good at chemistry’), students must to adopt a frame of reference to evaluate their ability beliefs in chemistry, either by comparing themselves to their fellow classmates, or by comparing their performance in chemistry to their performance in other subjects. Framing ability in respect to one's peers or to other subjects is more reflective of the development of a self-concept, rather than of self-efficacy (Marsh et al., 2019). Marsh et al. (2019) go on to claim “…generalized outcome expectancy and self-concept measures reflect a jangle fallacy, in which two scales with apparently dissimilar labels actually measure similar constructs” (p. 338). However, the measures of self-efficacy used in PISA 2003 were more task-specific compared to more general measures used in PISA2000. PISA2003 required students to report their confidence in relation to functional mathematics tasks such as using a train timetable and calculating the price of a product after a 30% discount. Huang (2012) found that the specific self-efficacy measurements used in PISA 2003 were better predictors of achievement than the PISA2000 measures of self-efficacy. This should concern chemistry education research since all but one self-efficacy scale used throughout the literature reviewed in this article includes generalised items which are not acutely task-specific (Ex, “To what extent can you explain chemical laws and theories?” And “How confident are you at applying theory learn in a lecture for a laboratory experiment?”). The only self-efficacy scale that is task specific is the Organic Chemistry Self-Efficacy scale, developed by Villafañe et al. (2016) which includes items such as “Determine the most stable chair conformation of a substituted cyclohexane?” and “Convert between a bond line structure and a Fischer projection?”.
Chemistry education research may also be encouraged to reflect more on the influence that measurement techniques may have on what is actually being measured. Although most affective scales reviewed in this article take the form of self-reports, it should be noted that it has been argued that self-reports can be considered a fallible source of data since the smallest change in wording or format can result in a major change in the obtained results (Schwarz, 1999). While a function of measurement scales is to provide information about participants, these scales can also function as a source of information for participants to make an informed response to the scale items (Schwarz, 1999). The danger that is presented here is that researchers are often not fully aware of the information that questionnaires provide to respondents and as a result, researchers “miss the extent to which the questions we ask determine the answers we receive” (Schwarz, 1999, p. 103). In the context of affective research, Feldman and Lynch (1988) points out that belief, attitude, or intention can be created by measurement if the measured constructs do not already exist in long-term memory. In other words, if a researcher asks a question about a belief or attitude, there is an inherent assumption that the subject will already have such a belief or attitude. If they don’t, they will be compelled to give an answer and as such, create a non-existent belief or attitude (Feldman and Lynch, 1988). Throughout the publications reviewed in this article, there was no evidence of considerable critical reflection on how the data collection tools used could have influenced the nature of the data that was collected.
While the implications being set out by this review may be extensive, this article does not wish to argue against the analysis of variables which are exceedingly complex and unique to each individual in any one setting and at any one time. Rather, it is encouraged for chemistry education research to fulfil a research brief that seeks to capture the fantastic complexity of the very reason an individual student would seek to learn at all;
“When we educators fail to appreciate the importance of students’ emotions, we fail to appreciate a critical force in students’ learning. One could argue, in fact, that we fail to appreciate the very reason that students learn at all.”
Immordino-Yang and Damasio (2007, p. 9)
This journal is © The Royal Society of Chemistry 2020 |