Umesh
Ramnarain
* and
Sam
Ramaila
University of Johannesburg, South Africa. E-mail: uramnarain@uj.ac.za
First published on 5th October 2017
This study investigated the self-efficacy of first-year Chemistry students at a South African university. The research involved a quantitative survey of 333 students using the College Chemistry Self-Efficacy Scale (CCSS) developed by Uzuntiryaki and Capa Aydin (2009). Descriptive statistics on data for the CCSS scales suggested that students have positive beliefs in their capability to accomplish chemistry tasks. The students scored more strongly on the self-efficacy constructs of cognitive and psychomotor skills than on everyday application. There was a significant difference between students of different professional orientations for cognitive skills and everyday applications, with students enrolled for Chemical Engineering having the highest mean scores for these constructs. A multiple regression analysis was run in order to explore the relationship between chemistry self-efficacy and performance in a chemistry examination. The analysis indicated that cognitive skills significantly predicted chemistry performance, while psychomotor skills and everyday applications had no significant impact. The implications for research and instruction are discussed in terms of the relationship between chemistry self-efficacy and performance.
Today, more Black students, especially from disadvantaged communities, are pursuing physical sciences at secondary school and, subsequently, studies in science, engineering, and technology at universities (Department of Education, 2010). This is the prevailing trend at all universities in South Africa, but more especially at the particular South African city university that forms the primary focus of this study, where the student headcount in the Faculty of Science was 2382 in 2008 and 3410 in 2017. Chemistry 1 is a key module as it is not only an elective for students enrolled for a Bachelor of Science degree, but is also a crucial requirement for studies in other degree programs such as the Bachelor of Engineering and the Bachelor of Optometry. Over the past ten years there has been a rapid increase in the number of students registered for first-year chemistry (Chemistry 1) with the enrolment being 472 in 2017. However, despite the larger number of students pursuing studies in the sciences, the number of graduates does not correlate well with this trend. This is partly due to the high drop-out rate which is a function of failure in certain first-year modules.
Research in science education has focused largely on cognition with little consideration of affective constructs and its relationship with academic performance (Koballa and Glynn, 2007; Fortus and Vedder-Weiss, 2014) Hence, there is a need to turn our attention to affective constructs such as student motivation, goal orientation, and self-efficacy (Schunk et al., 2008) with a view to addressing a dearth of research investigating affective factors related to science education in the literature (Gungor et al., 2007; Fortus and Daphna, 2017). Certainly, in South Africa, student affective constructs such as self-efficacy have been under-researched (Harry and Coetzee, 2011; Sofowora, 2014).
Accordingly, the research was guided by the following questions.
(1) How do first-year students perceive their self-efficacy in chemistry?
(2) Is there a significant difference in chemistry self-efficacy for students of various professional orientations?
(3) How well do students' perception of their chemistry self-efficacy predict and explain their performance in chemistry?
Bandura (1986, 1997) postulates that there are four sources of self-efficacy. According to him, a student's sense of self-efficacy emanates from mastery experiences, vicarious experiences, social persuasion as well as emotional and psychological states. Mastery experiences have the greatest impact on student self-efficacy. As students become successful, their self-efficacy is enhanced, while failure lowers self-efficacy. Vicarious experiences are related to the observation of role models such as teachers, parents, peers or characters in films with whom students can identify. Social persuasion can influence students positively and make them work harder towards achieving desired outcomes in science. The last source, emotional and psychological states, refers to the anxiety and stress that a person faces when performing a given task. If a student experiences high anxiety or stress when performing a given task, it can be interpreted by students as a lack of skill or ability to complete the task and will likely have a negative impact on students' self-efficacy (Usher and Pajares, 2008).
Self-efficacy is a construct that is context and task dependent (Pajares, 1996; Bong, 2006). Self-efficacy is therefore domain-specific. For example, a student may have a high self-efficacy with respect to knowledge and skills in chemistry and a low self-efficacy with respect to knowledge and skills in biology. The research reported in this paper recognises this domain-specificity in self-efficacy and is focussed on the self-efficacy and achievement in chemistry.
In particular, self-efficacy has been found to predict achievement amongst college students and high school students doing science (Zusho et al., 2003; Cavallo et al., 2004; Lalich et al., 2006; Jansen et al., 2015; Uzuntiryaki-Kondakci and Senay, 2015; Villafañe et al., 2016). Jansen et al. (2015) found that self-efficacy had a stronger predictive impact on the achievement of high school students in general science than self-concept. Cavallo et al. (2004) showed that self-efficacy significantly predicted physics achievement in college physics students in the United States. Villafañe et al. (2016) found a significant positive relationship between organic chemistry self-efficacy and chemistry performance. Lalich et al. (2006) found a correlation between self-efficacy and final achievement in a general chemistry course. In college chemistry, Zusho et al. (2003) conducted research with United States students enrolled in introductory college chemistry, and found self-efficacy to be the best predictor of grades, even controlling for prior achievement. They found that students' ratings of their levels of self-efficacy were better predictors of final course performance than the SAT-mathematics scores. A study reported by Uzuntiryaki-Kondakci and Senay (2015) with Turkish grade 11 students revealed that chemistry self-efficacy for cognitive skills was a significant positive predictor of chemistry achievement.
The research reported in this article purports to build upon the emerging knowledge domain of self-efficacy of college students studying science and the relationship of this construct with achievement. This research takes on particular significance in light of the increasing enrolment in science at South African universities and the disturbing trend of high failure in science courses.
Self-efficacy for cognitive skills (SCS) is a construct on students' beliefs in their ability to deal with intellectual operations in chemistry. This includes both lower and higher levels of understanding in the cognitive domain. An example of an item from this dimension is, “How well can you interpret chemical equations?” Self-efficacy for psychomotor skills (SPS) relates to students' beliefs in their ability to apply psychomotor skills. A sample item is, “How well can you construct laboratory apparatus?” Self-efficacy for everyday applications (SEA) describes students' beliefs in their ability to use the learned chemistry concepts in daily life situations. For example, such beliefs can be elicited by the item, “How well can you understand the news/documentary you watched on television related to chemistry?”
Professional orientation | N | Percent |
---|---|---|
Chiropractic | 57 | 17.1 |
Optometry | 38 | 11.4 |
Mechanical Engineering | 30 | 9.0 |
Chemical Engineering | 31 | 9.3 |
Analytical Chemistry | 93 | 27.9 |
Food Technology | 84 | 25.2 |
Total | 333 | 100.0 |
The factor analysis revealed the presence of three components with eigenvalues exceeding 1, explaining 52.37% of the variance. The interpretation of the three factors was consistent with previous research on the College Chemistry Self-Efficacy Scale (CCSS). The internal reliabilities of scales were evaluated by calculating Cronbach's alpha for each scale. Cronbach's alpha is used in this study as an indicator of scale reliability or internal consistency (Taber, 2017). This is the degree to which the items that made up the scale are all measuring the same underlying attribute (Pallant, 2007), which in this case is dimensions of self-efficacy. Table 2 presents the constructs assessed by the questionnaire used in this study, the items clustered in each construct, and Cronbach's alpha for each construct. The Cronbach's alpha of the scales compared favourably with the statistics obtained when the instrument was developed by Uzuntiryaki and Çapa Aydın (2009). This suggested satisfactory reliability. The Kolmogorov–Smirnov statistic assessed the normality of distribution. A non-significant result of more than 0.05 indicated normality.
Dimension | Cronbach alpha | Number of items | Items |
---|---|---|---|
Self-efficacy for cognitive skills (SCS) | 0.68 | 5 |
To what extent can you explain chemical laws and theories?
How well can you describe the structure of an atom? How well can you interpret chemical equations? How well can you interpret data during the laboratory sessions? How well can you solve chemistry problems? |
Self-efficacy for psychomotor skills (SPS) | 0.76 | 4 |
How well can you work with chemicals?
How well can you construct laboratory apparatus? How well can you use the equipment in the chemistry laboratory? How well can you carry out experimental procedures in the chemistry laboratory? |
Self-efficacy for everyday applications (SEA) | 0.69 | 3 |
To what extent can you propose solutions to everyday problems by using chemistry?
To what extent can you explain everyday life by using chemical theories? How well can you understand the news/documentary you watched on television related to chemistry? |
The first research question was addressed by computing the mean, standard deviation and bivariate correlations for the three self-efficacy constructs under consideration. The second research question was investigated by conducting a one-way analysis of variance (ANOVA). The third research question was investigated by constructing a hypothesized model describing the relationship between students' self-efficacy constructs and their achievement in chemistry. Based on previous research findings already reported in this article, we hypothesized that students' self-efficacy for cognitive skills, psychomotor skills and everyday applications will predict their achievement in chemistry. A multiple regression analysis was run in order to explore the relationship between self-efficacy constructs and chemistry achievement. For chemistry achievement, we used the students' semester examination mark in chemistry because the semester examination is considered a high stakes assessment. The examination is composed of questions that test conceptual understanding and application of chemistry concepts.
Self-efficacy constructs | ||||
---|---|---|---|---|
Cognitive skills | Psychomotor skills | Everyday application | ||
Cognitive skills | Pearson correlation | 1 | 0.575 | 0.538 |
Sig. (2-tailed) | 0.000 | 0.000 | ||
N | 333 | 333 | 333 | |
Psychomotor skills | Pearson correlation | 0.575 | 1 | 0.437 |
Sig. (2-tailed) | 0.000 | 0.000 | ||
N | 333 | 333 | 333 | |
Everyday application | Pearson correlation | 0.538 | 0.437 | 1 |
Sig. (2-tailed) | 0.000 | 0.000 | ||
N | 333 | 333 | 333 | |
Mean | 3.4469 | 3.5766 | 3.0255 | |
Standard deviation | 0.54850 | 0.62194 | 0.73229 |
The results shown in Table 3 indicate that students scored more strongly on the self-efficacy constructs of cognitive (M = 3.45, SD = 0.54) and psychomotor skills (M = 3.58, SD = 0.62) than on everyday application (M = 3.02, SD = 0.73). The overall mean was 3.35. From this score, it can be interpreted that the students' rating of their self-efficacy is slightly above neutral, with students exhibiting a favourable rating of self-efficacy. Correlation analyses revealed that students' perceptions of self-efficacy for cognitive skills and psychomotor skills were moderately related (r = 0.58, n = 333, p < 0.01). There was also a moderate correlation between everyday application and cognitive skills (r = 0.54, n = 333, p < 0.01) and a weak correlation between everyday skills and psychomotor skills (r = 0.44, n = 333, p < 0.01). The results per qualification are shown in Table 4 below.
N | Mean | Std deviation | Minimum | Maximum | ||
---|---|---|---|---|---|---|
Cognitive skills | Chiropractic/homoeopathy | 57 | 3.27 | 0.55 | 2.25 | 4.50 |
Optometry | 38 | 3.41 | 0.6 | 2.25 | 4.75 | |
Mechanical Engineering | 30 | 3.37 | 0.57 | 2.25 | 4.50 | |
Chemical Engineering | 31 | 3.66 | 0.58 | 2.75 | 4.75 | |
Analytical Chemistry | 93 | 3.53 | 0.53 | 2.00 | 4.50 | |
Food Technology | 84 | 3.43 | 0.49 | 1.67 | 4.50 | |
Overall | 333 | 3.44 | 0.55 | 1.67 | 4.75 | |
Psychomotor skills | Chiropractic/homoeopathy | 57 | 3.56 | 0.57 | 2.00 | 5.00 |
Optometry | 38 | 3.53 | 0.65 | 2.00 | 5.00 | |
Mechanical Engineering | 30 | 3.42 | 0.62 | 2.25 | 4.75 | |
Chemical Engineering | 31 | 3.83 | 0.68 | 2.50 | 5.00 | |
Analytical Chemistry | 93 | 3.53 | 0.63 | 1.50 | 5.00 | |
Food Technology | 84 | 3.63 | 0.60 | 2.50 | 5.00 | |
Overall | 333 | 3.58 | 0.62 | 1.50 | 5.00 | |
Everyday application | Chiropractic/homoeopathy | 57 | 2.8 | 0.73 | 1.00 | 5.00 |
Optometry | 38 | 2.84 | 0.75 | 1.50 | 4.50 | |
Mechanical Engineering | 30 | 2.92 | 0.76 | 1.50 | 4.50 | |
Chemical Engineering | 31 | 3.21 | 0.91 | 1.00 | 5.00 | |
Analytical Chemistry | 93 | 3.11 | 0.61 | 1.50 | 5.00 | |
Food Technology | 84 | 3.14 | 0.73 | 1.50 | 5.00 | |
Overall | 333 | 3.03 | 0.73 | 1.00 | 5.00 |
The differences between the qualification groups were investigated by conducting the analysis of variance (ANOVA) test. The ANOVA showed that the effect of the qualification type on self-efficacy was significant for cognitive skills and everyday applications. There was a statistically significant difference at the p < 0.05 level in cognitive skills: F(5, 330) = 2.844, p = 0.016 and everyday applications: F(5, 330) = 2.706, p = 0.021. The students enrolled for Chemical Engineering had the highest mean score for both cognitive skills and everyday applications. The effect size for cognitive skills, calculated using eta squared, was 0.04, and for everyday applications, it was 0.05. This suggests that despite reaching statistical significance, the actual difference in mean scores between the groups was quite small.
An analysis of standard residuals was carried out, which showed that the data contained no outliers. The histogram of standardised residuals indicated that the data contained approximately normally distributed errors, as did the normal P–P plot of standardised residuals, which showed points that were not completely on the line, but close. The scatter plot of standardised residuals showed that the data met the assumptions of homogeneity of variance and linearity. Correlation analyses revealed that students' perceptions of their self-efficacy for cognitive skills, psychomotor skills and everyday applications were weakly to moderately related to each other (0.44 ≤ r ≤ 0.58; p < 0.01) (Table 3). As correlations between most exogenous variables (independent variables) were lower than 0.6, the possibility of multicollinearity between these variables was excluded (Grewal et al., 2004) and all variables were retained for multiple regression analysis.
The R2 value indicates how much of the variance in student performance (dependent variable) can be explained by the model. In this case a value of 0.32 was attained and that means this model explains 32% of variance in the performance of students. The statistical significance of this result was then established through the ANOVA, F(3, 354) = 43.56, p < 0.001. The regression weight (beta) of the cognitive skills (0.35) was greater than 0.1 and significant (p < 0.05). This suggested that this variable positively predicted the performance. Both psychomotor skills (beta = 0.21) and everyday application (beta = 0.23) had no significant impact on student performance with p greater than 0.05.
There was a significant difference between students of different professional orientations for cognitive skills and everyday applications, with students enrolled for Chemical Engineering having the highest mean scores for these constructs. This finding may be attributed to students who register for Chemical Engineering needing to meet a higher pre-requisite score for this qualification than students registered for other qualifications. This is suggestive of a causative relationship where prior chemistry performance influences self-efficacy constructs. This can be explained in terms of the reciprocal relationship between chemistry self-efficacy and achievement that was revealed in the study by Villafañe et al. (2016).
A hypothesised model tested the effects of chemistry self-efficacy constructs on performance in a chemistry semester examination. In testing the model, multiple regression analysis indicated that cognitive skills predicted chemistry performance, while psychomotor skills and everyday applications had no significant impact. This result adds to the growing literature on psychological constructs within the chemistry education domain by identifying variables related to motivation that have a significant relationship to chemistry performance (Zusho et al., 2003; Cavallo et al., 2004; Lalich et al., 2006; Jansen et al., 2015; Uzuntiryaki-Kondakci and Senay, 2015; Villafañe et al., 2016).
The findings of this research highlight the significance of self-efficacy as an affective factor in the classroom for chemistry performance in particular. While other studies have investigated the role of various psycho-social and study skill variables in academic performance, self-efficacy has been identified as a strong predictor of performance (Robbins et al., 2004). The implication of this for chemistry teaching is that learning experiences ought to be tailored to improve the self-efficacy of students. For example, Vishnumolakala et al. (2016) reported that first-year undergraduate chemistry process-oriented guided inquiry learning (POGIL) classes positively influenced students' self-efficacy. Mataka and Kowalske (2015) found that students experiencing problem-based learning (PBL) in chemistry classes had enhanced self-efficacy scores. Nbina (2012) concluded that instruction in meta-cognitive self-assessment strategies has a significant effect on the Chemistry self-efficacy of secondary school students. In addition, a metacognitive self-assessment strategy and self-explanation have been shown to enhance students' chemistry self-efficacy and achievement at the levels of senior secondary school and college (Crippen and Earl, 2007). The findings of this study take on a particular significance within the South African education landscape where previously disadvantaged communities have under-performed in science, especially at the tertiary level. The significant relationship between the cognitive skills component of self-efficacy and chemistry performance suggests that cognitive skills are a good predictor of achievement, and are something that should be further promoted. However, no such relationship could be established between the other two self-efficacy constructs and chemistry achievement. Students appeared to overestimate their psychomotor skills. These findings are consistent with the claim that the overestimation of performance abilities is a result of not knowing one's limitations (Lawson et al., 2007). Thus, when students lack competencies, their judgment of what they can do is impaired, and they will often overestimate their competence.
Further research may address some of the limitations of this study. This research was conducted with students at a single university, and so future research may be carried out with more students at other universities in order to improve the generalizability of the findings. The research focussed primarily on the relationship between self-efficacy and achievement. Future studies may examine how student characteristics such as age, gender, race or ethnicity shape students' self-efficacy. This model can be tested with other factors that have been hypothesized to affect performance such as motivation, interest, and effort beliefs.
Future research can also investigate the interconnected reciprocal causation relationship between self-efficacy and performance chemistry. A recent study by Villafañe et al. (2016) was the first to introduce a reciprocal causation model in examining the interrelated relationship between students' chemistry self-efficacy in chemistry and performance. The findings of that study lent support to the hypothesis that self-efficacy has a consistent effect on examination performance and vice versa. There is a need for confirmation of such findings that examine the effect of chemistry performance on self-efficacy.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c7rp00110j |
This journal is © The Royal Society of Chemistry 2018 |