Self-efficacy as a predictor of academic performance in science

Download Self-efficacy as a predictor of academic performance in science

Post on 06-Jul-2016




0 download


<ul><li><p>Journal of Advanced Nursing, 1998, 27, 596603</p><p>Self-efficacy as a predictor of academicperformance in science</p><p>Sharon Andrew BAppSc MSc(Hons) RNDoctoral Candidate, Faculty of Education, University of Wollongong, Wollongong,NSW 25252, Australia</p><p>Accepted for publication 25 February 1997</p><p>ANDREW S. (1998) Journal of Advanced Nursing 27, 596603Self-efficacy as a predictor of academic performance in scienceNursing students have traditionally experienced difficulties with the sciencesubjects in nursing curricula, and irrespective of the institution conducting anursing programme, this trend appears to be continuing. A satisfactory means ofpredicting academic performance in these subjects will facilitate thedevelopment of educational strategies designed to assist students overcometheir difficulties. In this study, an instrument called the Self-Efficacy for Science(SEFS) was developed and tested. The SEFS was designed to predict academicperformance in the science areas of a first-year undergraduate nursing course. Acohort of first-year students enrolled in a bachelor of nursing course weresurveyed by questionnaire. Students academic scores for two first-year sciencesubjects were obtained and used as the criterion measure for the study.Principal component factor analysis revealed the SEFS contained six instead ofthe hypothesized four factors. These six factors could explain 70% of studentsself-efficacy for science. Cronbach alpha of the SEFS was 09. The SEFS couldpredict 24% of the cohorts academic performance in a physical science subjectand 185% for a bioscience subject. Studying science in the final year at highschool was not statistically significantly related to the SEFS. Implications forstudents and future research are discussed.</p><p>Keywords: Self-efficacy, science, academic performance, nursing, academicprediction, undergraduate education and gender</p><p>factors contributing to students responses to science inINTRODUCTION</p><p>the nursing curricula (Akinsanya &amp; Hayward 1980, Bishop1990, Courtenay 1991, Caon &amp; Treagust 1992, 1993,Although nursing curricula may vary (Barclay &amp; Neill</p><p>1987, Wharrad et al. 1994), nursing students will be Chapple et al. 1993).Due to the multiplicity of factors influencing studentsexpected, at some stage of a nursing course, to study biosci-</p><p>ence and possibly physical sciences. Nursing students in perceptions about science, academic performance in theseareas of the nursing curricula has been hard to predict.traditional and undergraduate courses have consistently</p><p>been reported as having difficulties with these areas of The development of specific educational strategies,designed to improve students academic achievements andtheir nursing courses. The science background of students,</p><p>the difficulty of the subject, perceived relevance to nursing reduce their anxieties, is dependant upon a satisfactorymethod of identifying nursing students self-expectationsand teaching methods employed have been suggested asabout the biological or physical sciences. Therefore, theaim of this study was to establish whether BandurasCorrespondence: Sharon Andrew, 28 Terrie Avenue, Figtree, NSW 2525,</p><p>Australia. (1977, 1986) theory of self-efficacy could be used in the</p><p>596 1998 Blackwell Science Ltd</p></li><li><p>Self-efficacy</p><p>prediction of nursing students academic performance in between self-efficacy expectations and academic perform-ance and course persistence, found that better results werethe science subjects of a first-year undergraduate nursing</p><p>course. To measure students self-efficacy for science an obtained if the measure/tool was more specific to the stud-ent group being assessed (Multon et al. 1991). Therefore,instrument called the Self-Efficacy for Science (SEFS) was</p><p>developed and tested. In this paper, the term science to measure the science self-efficacy expectations of nursingstudents, it is important that the instrument used isshould be interpreted as referring to the bio-and physical</p><p>sciences (chemistry, physics). specifically designed for this purpose. Hence, to do this,a research instrument called the Self-Efficacy for Science(SEFS) was developed and tested. The SEFS was designed</p><p>LITERATURE REVIEWto measure nursing students self-efficacy for science, andto determine its ability to predict academic performanceStudents high school subjects (particularly biology) have</p><p>been found to be a weak indicator of academic perform- in nursing bioscience and physical science subjects of afirst year bachelor of nursing course.ance in first-year bioscience subjects of undergraduate</p><p>nursing courses (Kershaw 1989, Caon &amp; Treagust 1992).Studies have frequently concluded that motivation is a THE STUDYpossible explanation for students academic achievements(Higgins &amp; Leelarthaepin 1986, Bishop 1990, Mills et al. The aim of the study was two-fold: firstly, to develop a</p><p>research instrument capable of measuring nursing stud-1992).Learning and motivation from the social cognitive per- ents self-efficacy for science and secondly, to determine</p><p>if that instrument was a predictor of students aca-spective are viewed in terms of cognitive processes thatare based on self-evaluations of past experiences and demic performance in the science subjects of a first year</p><p>undergraduate nursing course.involve the setting of internal standards or goals for behav-ioural tasks (Bandura 1977). Self-efficacy in this theory isa personal expectation about ones ability to successfully</p><p>Hypothesesperform a specific task or behaviour (Bandura 1986). Thesepersonal expectations in term influence the effort and per- To test the predictive validity of the SEFS and to investi-</p><p>gate the relationship between students science back-sistence that individuals will expend in the behaviouraldomain (Bandura 1977, 1986). ground and science self-efficacy, the following research</p><p>hypotheses for the study were proposed:Although Banduras self-efficacy theory was developedto assist in the understanding and treatment of phobias, itwas recognized as a theory with wider implications. It has $ nursing students self-efficacy for science is related</p><p>to academic performance in science-based first-yearbeen shown to be predictive of academic performance andpersistence in disciplines other than nursing (Lent et al. subjects.</p><p>$ students who study science in the final year of high1984, 1986, 1987, Brown et al. 1989). Whilst self-efficacyhas been applied to nursing practice (Redman 1985, Moore school will have a higher self-efficacy for science than</p><p>students who did not study science.1990, Mowat &amp; Laschinger 1994), very few studies wereidentified involving the application of self-efficacy theoryto nursing education or academic performance in nursing. Sample and design</p><p>Self-efficacy was found to be related, in one study, toacademic achievement in an introductory nursing course Undergraduate nursing students from a tertiary edu-</p><p>cational institution were surveyed by questionnaire during(Chacko &amp; Huba 1991). Self-efficacy, in turn, was found tobe influenced by students language and mathematical the first year of their nursing course. The questionnaire</p><p>included the SEFS, students demographic details/aca-ability, motivation and concentration/preparation forclass. Self-efficacy was reported to account for 8% of vari- demic background and research tools not covered in this</p><p>paper (Andrew 1995). The questionnaire was completedance in students academic achievement (Chacko &amp; Huba1991). voluntarily by 81 respondents, representing 94% of the</p><p>cohort. With ethics approval, consent to access studentsTo identify factors influencing students retention andattrition from a nursing course, two tools measuring nurs- academic scores for their first year subjects was sought</p><p>in the questionnaire. Approval was given by 77% (66ing and academic self-efficacy were developed and tested(Harvey &amp; McMurray 1994). Although not concerned with respondents) to this request.</p><p>Students in the undergraduate nursing course beingacademic performance per se, one of the findings indicatedthat students with a low mean academic self-efficacy and examined study two science subjects, one each session, in</p><p>the first year of their bachelor of nursing course. The firstgrade point average were more likely to withdraw from anursing course (Harvey &amp; McMurray 1994). session science subject (SCIE110) contained aspects of</p><p>physics and chemistry considered relevant to nursing. TheA meta-analysis of studies concerning the relationship</p><p>597 1998 Blackwell Science Ltd, Journal of Advanced Nursing, 27, 596603</p></li><li><p>S. Andrew</p><p>second science subject (SCIE111) included an introduction McMurray 1994). This method involves summing thestrength score for the individual items on the tool to get ato the biological functions of the body. Students academic</p><p>scores for these subjects were used to determine academic total strength score, and then dividing by the number ofitems to get a mean strength score for the tool. A meanperformance in science.score of five, for example, indicates strong self-efficacy forscience, whilst a low score indicates low self-efficacy for</p><p>Self-efficacy for sciencescience.</p><p>The face validity of the SEFS items were verified by theThe Self-Efficacy for Science (SEFS) a researcher-developed instrument, originally named the SS (Andrew experts in the teaching of physics, chemistry or human</p><p>bioscience to nursing students. One expert was also a1995), and later re-named the SEFS, was devised to deter-mine a students strength of self-efficacy for science. Self- nurse. Minor changes were made to the wording of a few</p><p>items. The science tasks were considered appropriate forefficacy strength refers to students degree of confidencein their ability to perform a task (Bandura 1986). It was nursing students.anticipated that a measure of the strength of a studentsself-efficacy for science would be related to academic RESULTSperformance in the science subjects.</p><p>The SEFS included science tasks, many of which were A SAS (1988) computer package was used in the analysisof the results.every day science tasks, however, some included course-</p><p>specific science-orientated tasks. The idea to use practical The gender composition of the respondents were foundto be 86% female and 14% male, with 71% of the femaleseveryday science tasks to measure self-efficacy was based</p><p>on Betz and Hacketts (1983) research regarding mathemat- and 100% of the males giving consent for access to theirfirst year academic marks.ics self-efficacy expectations. These researchers found</p><p>that females were not confident in performing mathe-matics tasks, and those tasks that were gender-stereo-</p><p>Academic performancetyped resulted in higher self-efficacy expectations for theappropriate gender (Betz &amp; Hackett 1983). Students academic performance was measured by their</p><p>score for each of the science subjects being examined. TheAlthough not identifiable to respondents, the SEFS wasdesigned to contain four sub-sections of neutral (5 items), descriptive statistics for these subjects, given in Table 1,</p><p>show that the mean for SCIE111 is lower and the standardmasculine (5 items), feminine (5 items) and mathematics(6 items) science tasks. Accordingly, these sub-sections deviation (SD) higher than SCIE110. However, a Spearman</p><p>Rank Order correlation found that students scores forwere termed: neutral science (NS), masculine science(MS), feminine science (FS), and mathematics science the both subjects were statistically related to each other</p><p>(P=00001).(MAS).The masculine and feminine items were chosen to reflect</p><p>societal gender-specific tasks. It is stressed, however, that Science backgroundoverall the science items were not designed to be gender-biased, rather to test the science self-efficacy expectations In the questionnaire, students were requested to list their</p><p>high school certificate (entry level for university) subjectsof all undergraduate nursing students. Neutral scienceitems were developed to cover general science tasks. and scores obtained for those subjects. The data regarding</p><p>students subject scores were incomplete for many stud-Masculine, feminine and neutral items were based onchemistry and physics principles which were applied to ents, and was not used in subsequent analyses. The science</p><p>subjects studied by students in their final year of higheveryday tasks, some of which were also directly appliedto nursing. The mathematics science sub-section includedmathematics items applied to science (physics, chemistry</p><p>Table 1 Summary statistics for SCIE110 and SCIE111and bio-) although some of these items were also specificto nursing. References for the tasks, except for one</p><p>Summary statistics(Walpole 1990), came primarily from material writtenspecifically for nursing students (DiMichael &amp; Raynor</p><p>SCIE110 SCIE1111988, Cree &amp; Rishmiller 1989, Marieb 1992).(n=64) (n=60)</p><p>Respondents were asked to indicate on a scale of one(not confident) to five (very confident), their confidence in Mean 6901 6603their ability to successfully perform each of the tasks. SD 940 1700Scoring for the SEFS was based on the method used by Median 6800 6950Betz and Hackett (1983) and other researchers (Lent et al.1984, 1986, 1987, Phillips &amp; Russell 1994, Harvey &amp; * Statistically significant.</p><p>598 1998 Blackwell Science Ltd, Journal of Advanced Nursing, 27, 596603</p></li><li><p>Self-efficacy</p><p>Table 2 Science subjects studied in final year of high school (%) could explain 12% of the variance. Factor three, calledand relationship between students science background and the lifestyle (LS), could also explain 12% variance and con-SEFS tained two items, one each from MS and FS. Factor four,</p><p>termed science principles (SP), contained four items oneScience subjects (%) each NS, MS and two FS items. It could explain 11% vari-</p><p>ance. Factor five, labelled practical science (PS), containedChemistry Physics Biology General science</p><p>three items, one each M, F and N items. It could explain10% of the variance. Finally, factor six explained 9% vari-</p><p>Females (n=70) 26 5 37 3ance and contained three items, two MAS and one MS.Males (n=11) 45 18 18 0These items were related to knowledge of physics conceptsCohort (n=81) 28 7 35 2applied to the home and bioscience and hence was termedphysics applied (PA).SEFS</p><p>t-Test results, shown in Table 5, indicate that there wasSciencebackground (%) Mean SD t-test: P only a statistically significant gender difference for the</p><p>PA sub-section; however, there were no gender differ-No science 43 390 064 ences for the remaining sub-sections, or for the SEFSDid science 57 414 061 009 overall.</p><p>Predictor validity of SEFSschool are shown in Table 2. For the cohort, students aremore likely to have studied biology than chemistry or As the SEFS was devised specifically for the prediction of</p><p>academic performance in the first year of an undergraduatephysics. Students were divided into two groups accordingto whether they had studied science (chemistry, physics, nursing course, it therefore required prediction validation</p><p>with the SEFS regarded as the predictor variable, and aca-biology or g...</p></li></ul>


View more >