Research in Science Education (2006) 36: 337Y353 * Springer 2006

DOI: 10.1007/s11165-005-9007-0

Sources of Self-efficacy in a Science Methods Course for Primary Teacher

Education Students

D.H. PALMER
School of Education, University of Newcastle
E-mail: David.Palmer@newcastle.edu.au

Abstract. Self-efficacy has been shown to be an issue of concern for primary teacher education
students Y many of them have low self-efficacy and this can negatively affect their future teaching
of science. Previous research has identified four factors that may contribute towards self-efficacy:
enactive mastery experiences, vicarious experiences, verbal persuasion and physiological/affective
states. It could also be argued that there are additional sources of self-efficacy that apply to primary
teacher education students, namely cognitive content mastery, cognitive pedagogical mastery and
simulated modelling. The main purpose of the present paper was to investigate the relative impor-
tance of the various sources of self-efficacy in a primary science methods course. Data on changes in
self-efficacy and sources of self-efficacy were collected throughout the course using formal and
informal surveys. It was found that the main source of self-efficacy was cognitive pedagogical
mastery.

Key Words: self-efficacy, science, primary teacher, elementary teacher

Bandura (1982) proposed that self-efficacy represents a person’s belief in his/her

ability to perform a difficult task: BSelf-efficacy is concerned with judgements
about how well one can organise and execute courses of action required to deal
with prospective situations that contain many ambiguous, unpredictable, and often
stressful, elements^ (Bandura, 1982, pp. 200Y201). Self-efficacy has been shown to
be a powerful predictor of performance. People who have strong self-efficacy for a
particular task will make vigorous and persistent efforts and thus will be more
likely to succeed, but those with low self-efficacy will give up after minimal effort,
or even avoid the task altogether. Self-efficacy is very similar to self-confidence,
and in many studies the two terms have been used interchangeably (Appleton &
Kindt, 2002; Cannon & Scharmann, 1996; Rice & Roychoudhury, 2003; Settlage,
2000; Watters & Ginns, 2000).
Bandura (1997) described four main sources of self-efficacy: enactive mastery
experiences, vicarious experiences, verbal persuasion and physiological/affective
states. Enactive mastery experiences are authentic successes at dealing with a
particular situation. Bandura (1997) considered these to be the most influential
source of self-efficacy because they provide authentic evidence that the indi-
vidual has the ability to succeed at the task. Vicarious experiences are situa-
tions in which people estimate their capabilities in comparison to others who have
338 D.H. PALMER

modelled the desired behaviour. There are several modes of modelling influences:
(1) Feffective actual modelling_ occurs when one sees a person similar to oneself
perform the task successfully; (2) Fsymbolic modelling_ occurs when individuals
are exposed to effective models provided by television and other visual media; (3)
Fself-modelling_ occurs when individuals’ performances are videotaped for them to
watch, but only after the recordings have been edited to show only the favourable
aspects; and (4) Fcognitive self-modelling_ occurs when individuals visualise them-
selves performing successfully at the task. Verbal persuasion refers to situations in
which individuals are given positive feedback from others Y if a person is told that
he/she does possess the capabilities to succeed in the task, then that person will be
encouraged to try hard to succeed. Physiological and affective states refers to
individuals’ responses to their own stress, fear and anxiety Y moderate levels of
stress can energise high achievers but can debilitate low achievers.
One group of people for whom science teaching self-efficacy is an issue of con-
cern is primary teacher education students. Many of these students had neg-
ative experiences of science in high school (Jarrett, 1999; Mulholland & Wallace,
1996) and they lack confidence in their ability to teach the subject (Tosun, 2000).
Teacher lack of confidence is a serious problem because it impacts on class-
room behaviour. Appleton and Kindt (1999) found that beginning teachers with
low confidence avoided hands-on science and used strategies based on reading and
writing.
As a result of findings such as these, a number of studies have investigated how
best to enhance the science teaching self-efficacy of primary teacher education
students. It has been found that science methods courses that involve extensive use
of hands-on activities can be very successful in this regard (Jarrett, 1999; Schoon
& Boone, 1998; Watters & Ginns, 2000). However, the reason why hands-on
activities can positively affect students’ self-efficacy remains to be established. In
order to resolve this issue it is necessary to interpret the content of courses from the
perspective of Bandura’s sources of self-efficacy.
A few authors have taken the approach of identifying the sources of self-efficacy
in their methods courses. Cantrell, Young, and Moore (2003) found that mastery
experiences, consisting of time spent teaching science to children in a primary
classroom, were associated with increases in personal science teaching efficacy
in their methods course. Wingfield, Freeman, and Ramsey (2000) found that a site-
based preservice education program enhanced self-efficacy in several ways: mas-
tery experiences occurred when students assisted in small group activities with
children, then planned and implemented whole class science lessons; vicarious
experiences were provided by peer reviews, critiques and modifications of lessons;
and verbal persuasion was provided by the site-based teachers, the university
coordinator and peers. Settlage (2000) measured self-efficacy before and after a
science methods course. He found a significant improvement and proposed that
BThe microteaching would be classified as a performance accomplishment, the
classroom videos as vicarious experiences, lectures and discussion as verbal
 SOURCES OF SELF-EFFICACY IN A SCIENCE METHODS COURSE 339

persuasion, and visits to the classrooms . . . as emotional arousal^ ( p. 49). It thus

appears that there are several ways in which mastery experiences, vicarious ex-
periences and verbal persuasion could possibly be included in these courses. How-
ever, the relative importance of each of these sources is still an open question.
Bandura (1997) proposed that enactive mastery experiences are usually the most
powerful source of self-efficacy, but corroborating evidence that this is the case in
science methods courses is yet to be obtained.
It could also be argued that, with respect to primary teacher education students,
there are other sources of self-efficacy in addition to those proposed by Bandura
(1997). For example, in addition to enactive mastery (in which students actually
teach primary children) other forms of mastery could be possible. Firstly, science
content knowledge is one factor that has been linked with increased confidence and
self-efficacy of primary teacher education students (Schoon & Boone, 1998), so it
is possible that mastery experiences in understanding science content are im-
portant. Understanding science can be a difficult and daunting task for many stud-
ents, particularly at high school level when science is compulsory. The fact that
preservice primary teachers’ memories of their school science experiences arouse
feelings of fear, anxiety and intimidation (Bell, 2001) suggests that understanding
science content represents a significant efficacy issue for them. Successes in
mastering understandings of science subject matter could therefore be expected to
enhance their feelings of efficacy for teaching science. This type of mastery ex-
perience is distinct from enactive mastery because it involves success in under-
standing something rather than success in doing something. It could therefore be
referred to as Fcognitive content mastery._
A second factor that has been linked to increased confidence of these students
is science pedagogical knowledge. For example, Appleton (1995) found that a
constructivist/interactive science methods course resulted in positive changes in
confidence even in content areas that were not in the course Y implying that
pedagogical knowledge was important to these students. Similarly, Settlage (2000)
found that instruction about the learning cycle (a pedagogical technique) con-
tributed to preservice elementary teachers’ self-efficacy. It is therefore possible that
mastery of specialist pedagogical knowledge may provide a source of science
teaching self-efficacy. Again, it is not easy for primary teacher education stud-
ents to develop a confident knowledge of how to teach science, because there
are comparatively few positive role models available. For example, Mulholland
and Wallace (2001) found that the low status of science in primary schools and
the inexperience of the teachers in the subject meant that there was a lack of
positive role models for new teachers. It is therefore not uncommon for primary
teacher education students to enter their science methods courses with very little
understanding of how to teach science in a way that will capture the imaginations
of the children. Success in mastering an understanding of some motivating and
effective techniques for teaching science could therefore be expected to make an
important contribution towards developing their science teaching self-efficacy.
340 D.H. PALMER

This type of experience could be referred to as Fcognitive pedagogical mastery_ for

science teaching.
There is also a third possible alternative source of self-efficacy. Bandura (1997)
described vicarious experiences as those in which individuals observed other
people modelling the target behaviour. For primary teacher education students,
a vicarious experience could be provided by watching a video of a teacher or peer
successfully teaching a science lesson to primary students. However, there is
another type of modelling that is also associated with improved self-efficacy.
Posnanski (2002) found that personal science teaching efficacy was improved in a
course that included a type of modelling in which the tutor assumed the role of
a primary teacher and the students assumed the role of children, as they par-
ticipated in hands-on activities in preparation for real experiences in schools. Rice
and Roychoudhury (2003) argued that this type of modelling was an important
component of a science methods course for elementary teaching. This type of
modelling does not fit neatly into any of Bandura’s (1997) categories of vicarious
experiences because it does not involve any observed teaching of real children.
Instead, the tutor and the students simulate the conditions of a primary classroom
by a type of role playing. This type of vicarious experience could therefore be
referred to as Fsimulated modelling._
In summary, it has been argued that cognitive content mastery (successes in
understanding science content), cognitive pedagogical mastery (successes in un-
derstanding how to teach science) and simulated modelling (in which teaching is
role played) could be sources of self-efficacy in addition to those proposed by
Bandura (1997). However, the relative importance of each of these in comparison
to the other sources remains to be established. The purpose of the present study is
to investigate this issue. The research questions are:

1. In one particular primary science methods course, what is the relative im-
portance of each of the factors that can influence self-efficacy?
2. To what extent can these sources enhance the science teaching self-efficacy of
the students?

Materials and Methods

Participants

The research was carried out at a regional university in south-eastern Australia.

The participants were primary teacher education students who were enrolled in a
one-semester science methods course. There were about 190 students enrolled, but
as explained below, not all students participated in each phase of the data gath-
ering. About 84% of them were female, and most were in the 20Y30 years age
group. These students were in the third year of their four-year undergraduate pro-
gram and had previously completed a one-semester science foundations course.
 SOURCES OF SELF-EFFICACY IN A SCIENCE METHODS COURSE 341

The foundations course covered science concepts in areas such as physics, chem-
istry and biology.

The Methods Course

The main purpose of the methods course was to teach the students how to teach
science to primary children (i.e., children of ages 5Y12) but it also aimed to
reinforce the students’ science content knowledge and to develop positive attitudes
towards science. It ran for one semester of 13 weeks, and each week there was a
1-h lecture to the whole group and a 1-h workshop to subgroups of about 30
students. All the lectures and workshops were presented by the same tutor.
The lecture content covered topics such as process skills, student misconcep-
tions, lesson planning, programming, safety issues and teaching techniques such as
demonstrating, investigating and problem solving. In order to interest the students
and arouse their enthusiasm for teaching science, the lectures involved many de-
monstrations of hands-on activities, and the use of numerous examples of lessons
and lesson sequences that were relevant to primary classrooms. The tutor also
modelled simple explanations of science concepts, at a similar level to that which
could be used with primary children.
The workshop content covered topics such as sound, water, air, magnets, static
electricity, plants and animals. The emphasis was on the use of hands-on activities
to teach these topics. The tutor modelled how to provide teacher explanations of
science concepts and how to provide instructions for the hands-on activities. The
students usually did the hands-on activities themselves and also participated in
investigating and problem solving. The hands-on activities were carefully selected
and in many cases they had surprising or unexpected results, so as to appeal to the
students. They also used everyday materials rather than specialised scientific
equipment, and this was intended to make the activities more relevant to primary
schools, where specialised equipment is often not available.
As one of the assessment tasks for the course, students were required to teach a
hands-on science lesson to a single child of primary age (i.e., a 1:1 teaching sit-
uation) and self-evaluate the experience. This was to be done outside of course
time and in informal settings such as private residences, rather than in schools.
This was intended to give students an opportunity to have a successful experience
in teaching science to a child.
The teaching of the course was not modified in any way to meet the req-
uirements of this study. The course did not contain any experiences in a real
school. This means that there were no opportunities for effective actual modelling
(i.e., seeing a person teaching in a primary school). Neither did it contain any
videos, pictures or films of teaching, so there were no opportunities for symbolic
modelling or self-modelling. Thus, this particular course did not contain the full
range of Bandura’s sources of self-efficacy.
342 D.H. PALMER

Data Collection

The study involved both quantitative and qualitative methodologies. Data were
obtained through the use of two formal surveys and three informal surveys, as
follows.

Formal Surveys

These were used to measure students’ self-efficacy levels. The instrument used was
the Science Teaching Efficacy Belief Instrument Form B (STEBI B). This was
developed by Enochs and Riggs (1990) for use with preservice primary teachers. It
consisted of 23 items, each of which was linked to a 5-point Likert-type scale
consisting of the choices Fstrongly agree,_ Fagree,_ Funcertain,_ Fdisagree_ and
Fstrongly disagree._ Some of the items were worded positively and others
negatively. The items were scored 5, 4, 3, 2 or 1 in which 5 was the maximum
positive response and 1 was the most negative response. The instrument contained
two scales: the Personal Science Teaching Efficacy Belief Scale (PSTEB) con-
tained 13 items, and the Science Teaching Outcome Expectancy Scale (STOE)
consisted of 10 items. Enochs and Riggs (1990) showed that the PSTEB had an
alpha coefficient of 0.9, and the STOE alpha coefficient was 0.76. All items loaded
highly with their own scale, under factor analysis. The instrument was validated for
use with Australian students by Ginns, Watters, Tulip, and Lucas (1995) and has
been widely used in recent years (Bell, 2001; Cannon & Scharmann, 1996;
Cantrell, et al., 2003; Schoon & Boone, 1998; Settlage, 2000; Watters & Ginns,
2000; Wingfield, et al., 2000).
The STEBI B was administered to the students as a pre-test, on the first day
of the course, and as a post-test on the final day of the course, 13 weeks later.
Students were asked to write an anonymous identifier (e.g., their mother’s given
names) so their pre-test and post-test results could be paired.
Analysis of the STEBI B data was carried out using paired t-tests. For each
student, the two scales were summed and compared separately. It was recognised
that the use of two paired t-tests could possibly increase error margins, so the
decision was made to adopt a lower significance level, of 0.01 instead of 0.05,
to compensate.

Informal Surveys

The purpose of the three informal surveys was to provide data about the sources of
self-efficacy and the relative importance of each source. The relative importance of
each source would be indicated by the number of people who mentioned it. The
three informal surveys were designed to focus on different components of the
 SOURCES OF SELF-EFFICACY IN A SCIENCE METHODS COURSE 343

course. The course consisted of two main types of experiences: the lectures, in
which students were relatively passive members of a large audience; and the
workshops, in which they actively participated in group work, hands-on activities
and discussion. It was decided to seek feedback on each of these separately, as
well as feedback about the course as a whole, so three surveys were needed. In
addition, the disadvantage of having a survey at the end of the course was that
students would be forced to rely on their memories of what had happened over
the previous 13 weeks, which may have introduced some error or bias into the
results, so it was decided that the other two surveys would be carried out during the
course itself. The three informal surveys therefore complemented each other by
focussing on different aspects of the course and by being carried out at differ-
vent times.
The first informal survey was carried out in Week 5 of the course, at the lecture.
This lecture addressed the topic of Fair_ Y the tutor presented background
information about the composition of air and demonstrated several hands-on
activities that could be used to show the properties of air. The tutor also described a
student-centred investigation on the topic of bubbles, and outlined a sequence that
could be used to structure student investigations. At the end of the lecture, each
student was provided with a sheet of paper and asked to write an answer to the
question BHas anything in today’s lecture helped to make you more confident to
teach science? Please write as much detail as possible.^
The second informal survey occurred in the Week 8 workshop. The topic for this
workshop was magnets, and students were presented with a description of several
lessons that represented a short unit of work on magnets. For each lesson, the tutor
described the knowledge that the children would be expected to achieve, a hands-
on activity that could be used to help them understand, and one or two ideas for
assessment tasks that the children could do to demonstrate their learning. The
students in the workshop actually did most of the hands-on activities. At the end of
the workshop they were asked to answer the question BWas there anything in
today’s workshop that helped to make you more confident to teach science? Please
write as much detail as possible.^
The third survey was carried out at the end of the semester, when students
were asked to reflect back on the course as a whole and BWrite something that
stands out as being a useful or valuable aspect of the course in giving you more
confidence to teach science.^ In each survey, the time provided for writing was
5Y10 min.
Analysis of each of the surveys firstly involved the creation of categories repre-
senting sources of self-efficacy. Students’ responses were then allocated to the
categories and the proportion of students in each category was calculated. To check
the reliability of the categories, a representative sample of 79 student statements
taken from all three surveys was independently coded by the author and a second
person who had a higher degree in education. Agreement was found in 86%
of cases.
344 D.H. PALMER

Results

Formal Surveys

Paired pre- and post-test responses were obtained from 108 students. The results of
the paired t-test analysis (Table 1) showed there had been a significant im-
provement in both scales. Effect size was calculated by finding the difference
between the group means and dividing it by the mean standard deviation (Cantrell
et al., 2003). This showed that for both scales, the effect size was large (using this
calculation, any effect size above 0.8 is considered large). These results indicated
that students’ self-efficacy had improved by a considerable amount over the period
of the course.
The remainder of students in the course either missed the pre-test or the post-
test, or their pre-tests and post-tests could not be reliably paired, or they did not
answer all the questions in the survey. To check whether the responses of these
students were different in any way to the 108 who provided complete responses,
the following analysis was carried out. The results were recorded for any student
who provided a complete response for at least one of the two scales, even if they
answered only one survey. On this basis, 52 additional students’ responses were
obtained for the pre-test and 42 additional responses for the post-test. The means
and standard deviations for these additional students were as follows: PSTEB pre-
test = 43(5.0), post-test = 51(6.4); STOE pre-test = 34(4.0), post-test = 38(3.5).
Comparison with the results in Table 1 indicates that the corresponding means
were quite similar. This implies that the students for whom full responses were
obtained were not substantially different from other students in the course.

Informal Surveys

After initial reading of the survey responses, a list of sources of self-efficacy was
created. This included categories described by Bandura (1997) as well as the
additional categories of cognitive content mastery, cognitive pedagogical mastery
and simulated modelling. An additional category of unspecified cognitive mastery
was included, as some responses indicated a successful learning experience, but

Table 1
Mean (Standard Deviation) for STEBI B Scales.

Pre-test Post-test t df Probability Effect size

PSTEB 42 (6.8) 53 (4.8) 16.6453 107 G0.01 1.9 (large)

STOE 34 (4.4) 38 (4.3) 8.6262 107 G0.01 0.9 (large)
 SOURCES OF SELF-EFFICACY IN A SCIENCE METHODS COURSE 345

were too vague as to whether it was content or pedagogy. The categories of ef-
fective actual modelling, symbolic modelling and self-modelling were not included
as there were no opportunities for them to participate in these activities in the
course, and none of the students’ responses referred to them. The final list therefore
contained the following categories:

1. enactive mastery (i.e., a successful experience teaching a child);

2. cognitive content mastery (i.e., a successful learning experience involving the
understanding of science concepts);
3. cognitive pedagogical mastery (i.e., a successful learning experience involving
the understanding of science teaching techniques);
4. unspecified cognitive mastery (i.e., a successful learning experience was indi-
cated, but whether it was content or pedagogy could not be established);
5. cognitive self-modelling (i.e., students imagined themselves teaching);
6. simulated modelling (i.e., role playing a primary class);
7. verbal persuasion (i.e., students received feedback that their teaching was
successful);
8. physiological /affective states (i.e., coping with stress, fear and anxiety);
9. other (i.e., students whose responses could not be categorised).

The numbers of students responding to the three surveys varied as not all stud-
ents attended the lectures or tutorials. In the first survey, responses were received
from 124 students, 175 responded to the second survey and 163 responded to the
third.
The survey responses supported the results from the STEBI B as many students
stated that their confidence had increased. The following are some selected re-
sponses. The first two (from Survey 1) are especially interesting because they in-
dicate that significant shifts in confidence had occurred as early as Week 5 of
the course.

I didn’t like science before these lectures. Now I feel confident to teach and feel confident the kids will
enjoy. (Survey 1)

These are excellent lectures. I am now feeling confident to teach science. I would not have before this
series of lectures. (Survey 1)

I use [sic] to dislike science but now I think it will be a great subject to teach. (Survey 3)

Several sources of self-efficacy were indicated by the responses, and these are
shown in Table 2. For each survey, percentages add up to more than 100 because
some students made more than one type of response.
In each of the three surveys, the majority of students made statements that were
categorised as cognitive pedagogical mastery (i.e., the students indicated they had
learnt information about how to teach science). Responses were included in this
category if they indicated that students had learnt or been shown how to do science
346 D.H. PALMER

Table 2
Frequency of Sources of Self-Efficacy (percent).

Survey 1 Survey 2 Survey 3

(n = 124) (n = 175) (n = 163)

Enactive mastery 0 0 0
Cognitive content mastery 18 19 9
Cognitive pedagogical mastery 59 88 75
Unspecified cognitive mastery 15 2 4
Cognitive self-modelling 21 26 26
Simulated modelling 8 5 10
Verbal persuasion 0 0 0
Physiol./affect. states 2 0 2
Other 7 6 12

lessons, activities, explanations, demonstrations or procedures for teaching science.

Responses were also counted in this category if they wrote that science lessons,
activities, explanations, demonstrations or procedures for teaching science had
been clear, easy, helpful or good, as this was taken to imply that they had learnt
something about these techniques. The following are some examples of the types
of responses included in this category (where appropriate, the critical words and
phrases have been italicized):

Yes, learning a range of practical things you can do to develop children’s understandings of air. i.e.,
demonstrations, activities etc (Survey 1)

Tutorials are prescriptive and therefore we learn the procedures necessary to teach a science lesson
effectively. (Survey 2)

The thing that was most helpful in my understanding of science and teaching it was the excellent
teaching activities and strategies that were provided in each lesson. I feel more confident to teach this
subject as I have a background knowledge of science topics and have learnt great strategies. (Survey 3)

Table 2 shows that fewer students gave responses indicating cognitive content
mastery. Responses were placed in this category if they implied improved under-
standing of science concepts or improved ability to answer children’s questions
about science. The comments indicated that understanding science content had
been facilitated not only by the explanations given by the tutor but also by the
hands-on activities. For example,

Simple explanations of why things happen help my own understanding of science. (Survey 1)

I never understood magnetic pulls now I have a basic understanding. You can’t teach without un-
derstanding yourself. (Survey 2)
 SOURCES OF SELF-EFFICACY IN A SCIENCE METHODS COURSE 347

I found that hands-on activities extremely helpful. They explained science concepts that I didn’t fully
understand . . . (Survey 3)

In each survey, small numbers of responses were categorised as unspecified

cognitive mastery (Table 1). Responses were placed in this category if they in-
dicated that the lecture/tutorial/course had been informative or had improved
their understanding, but the nature of the learnt material was not clear. For
example,

I have more confidence now because I have a better knowledge base. (Survey 2)

The results provided no evidence that enactive mastery was a source of self-
efficacy in the course. The students had an opportunity to experience enactive
mastery through the assessment task that required them to teach a hands-on science
lesson to an individual student. However, in their written comments, none of the
students specifically mentioned the 1:1 teaching experience, and the comments that
referred to the assessment tasks were too vague to be reliable.
The most frequent type of modelling in each survey was cognitive self-model-
ling (i.e., imagining oneself teaching) and this was the second most important
source of self-efficacy in all three surveys (Table 2). Responses were placed in
this category if they referred to future teaching, or future use of ideas for teach-
ing or potential use of ideas or techniques. Responses were also placed in this
category if they stated that the course had provided resources for teaching, as
this was taken to indicate they had thought about whether they could use the
ideas in a classroom:

Yes. The use of practical examples . . . makes me believe that I can teach using these practical activities.
(Survey 1)

Yes. Real stuff that we can see is going to be useful when we go into schools. (Survey 2)

The way in which each new lesson/concept was presented. It was practical and I could see myself
teaching science in the same way. (Survey 3)

Interestingly, many of the students wrote responses that indicated both cognitive
self-modelling and cognitive pedagogical mastery:

Yes Y the activities and demonstrations [the tutor] gave, I think I could use them in a classroom.
(Survey 1)

By giving us examples of lessons and lesson sequences, I feel that I could apply them to a class.
(Survey 3)

These types of responses suggest that the students had been shown teaching
techniques that they had understood, and were then imagining themselves teaching
using these techniques.
348 D.H. PALMER

Table 2 shows that small numbers of students provided responses that were
categorised as simulated modelling (i.e., role playing of teaching). Responses were
placed in this category if they indicated that the tutor had taught them as if they
were children in a primary class, or the tutor had taught them in the same way that
they would teach children. In addition, some responses indicated that individual
students had experienced science learning or motivation from activities in the
course, so they believed that primary children would also learn from and be
motivated by those activities Y these types of responses were also categorised in
this group because they implied that the student was taking the place of primary
children. Examples of this category are:

These lectures do make me more confident to teach science. In this lecture, as others, the way [the tutor]
explains to us as we would explain to students, and examples as to demonstrations etc. fantastic!
(Survey 1)

Because I am a hands-on learner I find the tutorials excellent for deep learning and enjoyable. I think
kids would enjoy to be taught the same way. I have always disliked science and now I am quite happy to
teach this to primary school students. (Survey 2)

The fact that [the tutor] would take us through an actual lesson, do the experiment, let us also do
experiment and generally simulate the classroom procedures by asking questions at various intervals of
exercise showed us how to teach in science. I am much more confident as a result. (Survey 3)

In all three surveys there were very small numbers (less than 5% in all cases) of
students who made comments about coping with fear, anxiety and other negative
physiological/affective states:

Yes, because [the tutor] gave us specific activities to do with a class and then explained etc which makes
approaching science less threatening and more fun. (Survey 1)

I now know not to fear science in a classroom because already I will have captured their interest which
is half the battle. (Survey 3)

These types of comments implied that learning about science teaching techniques
had helped these students overcome their fears.
The surveys contained no evidence that verbal persuasion had been a significant
factor in enhancing confidence. Only one student made a comment about the
feedback given from the tutor, but it was too vague to be reliable.

Discussion

Sources of Self-Efficacy

The results provided evidence that cognitive content mastery, cognitive pedagog-
ical mastery and simulated modelling can be sources of self-efficacy in addition to
 SOURCES OF SELF-EFFICACY IN A SCIENCE METHODS COURSE 349

those proposed by Bandura (1997). It could be rightly be argued that content

knowledge and pedagogical knowledge are prerequisites for enactive mastery in
teaching, as a successful teaching experience would partly depend on having a
good knowledge of content and teaching methods. However, the students’
comments in this study suggested that they gained confidence directly from suc-
cess in understanding content and pedagogy, which thus makes them distinctive
mastery experiences in their own right, as well as being prerequisites for enactive
mastery. In addition, simulated modelling is a type of vicarious experience which
differs from other types because the modelling process does not involve real
children. Instead, the tutor and students are to some extent role playing the teacher
and children in a primary classroom. Through this type of modelling, the students
could still be shown how to teach children, rather than just being told, and this
helped to improve their confidence that they could teach science effectively.
The results also showed that cognitive pedagogical mastery (i.e., successes in
understanding how to teach science) and cognitive self-modelling (i.e., imagining
oneself teaching) were the two most common sources, and many students appeared
to use the two in combination.
Cognitive content mastery (i.e., success in understanding science content) was a
source of self-efficacy for 9%Y19% of students, and this appeared to have been
partly facilitated by the use of hands-on activities. It thus appears that extensive
use of hands-on activities can enhance efficacy in two ways Y firstly, by providing
effective instructional strategies (cognitive pedagogical mastery), and secondly, by
consolidating the science content understandings of the students themselves (cog-
nitive content mastery).
Simulated modelling was a source of self-efficacy for only 5%Y10% of students.
These responses indicated that self-efficacy had been enhanced through role play
situations in which they had been taught as children would be taught in a school.
Some of these students wrote the following types of responses:

Asking if the students have any questions, lets us find out the sorts of questions that we want to know
and so do the kids. (Survey 2)

The practical ideas are great. Kids love it hell I love it. (Survey 2)

Comments such as these suggest that the students had directly experienced mo-
tivation and content learning under the simulated conditions at university, and this
led them to believe that the same techniques would also be effective in the primary
classroom. Whilst it is to be hoped that those techniques would in fact be effective
in the primary classroom, there is a potential problem with this type of modelling.
Teacher education students are at a vastly different educational level to primary
aged children, so it should not be assumed that a technique that can promote
motivation and learning in adult tertiary students would be just as effective with
young children. Hence there is the potential that the use of this type of modelling
could create false expectations of efficacy. Perhaps one solution to this problem
350 D.H. PALMER

would be for the tutor to make students aware of the issue by discussing it with
them in conjunction with the simulated experiences.
The results also indicated that several of Bandura’s sources of self-efficacy were
not significant in this study. These included enactive mastery, actual modelling and
verbal persuasion. A small number of students made comments about coping with
stress, fear or other physiological/affective states, and their comments indicated
that these feelings had been reduced by learning about the activities, demonstra-
tions or experiments that could be used to teach science. This suggests that cog-
nitive pedagogical mastery may play a role in reducing negative visceral arousals.
This is roughly in agreement with the view of Bandura (1997) who stated that
anxiety could be diminished by modelling or mastery experiences.

Changes in Self-Efficacy

The results of the formal surveys indicated that students’ science teaching self-
efficacy had significantly improved over the period of the course. Large effect sizes
were found for both Personal Science Teaching Efficacy Belief and Science
Teaching Outcome Expectancy. One issue however, is that the course did not
provide enactive mastery experiences, and as these are usually considered to be the
most powerful source of self-efficacy (Bandura, 1997) it raises the question of
whether the other sources of self-efficacy were able to raise efficacy to the same
degree. It other words, it is possible that students in this course may have been
disadvantaged Y they may not have experienced as much growth in self-efficacy as
they may have had if the course had contained opportunities for them to teach in a
school. To investigate to extent to which this may have been the case, the STEBI B
results from this study were compared to those from other studies that have
contained in-school experiences. Three recent studies of this type were found, and
their results are compared in Table 3. This table shows that the post-test scores for
PSTEB and STOE were only slightly below the highest values recorded in the
other studies, and were well above some of the others. This implies that the sources
of efficacy identified in the present study (i.e., cognitive pedagogical mastery,
cognitive content mastery, cognitive self-modelling and simulated modelling) did
compensate for a lack of enactive mastery.

Implications

The study has raised an issue about the content of future primary science methods
courses, as it showed that significant gain in students’ self-efficacy can occur in the
absence of enactive mastery experiences. However, this should not be construed as
implying that enactive mastery experiences should be omitted from courses.
Rather, it is possible that if authentic enactive mastery experiences had been
 SOURCES OF SELF-EFFICACY IN A SCIENCE METHODS COURSE 351

Table 3
Comparison of STEBI B Scores with Other Studies.

PSTEB STOE

Pre-test Post-test Pre-test Post-test

Study 1 46.77 53.78 36.05 39.32

Study 2 45.1 46.1 33.5 34.9
Study 3 46.33* 53.58** 25.42 26.00
This study 42 53 34 38

Study 1 = Wingfield et al. (2000).

Study 2 = Ginns et al. (1995).
Study 3 = Cantrell et al. (2003).
*Prior to the methods course semester
**At the end of the methods course semester

provided in this course then self-efficacy would have been improved even further.
An additional advantage of including authentic science teaching experiences in real
schools is that it may facilitate verbal persuasion Y constructive feedback by
teachers and peers will also contribute towards student efficacy Y and this was also
a factor missing from the present course.
However, the findings should be interpreted with the following limitations in
mind. Firstly, the present study applied the concept of self-efficacy only to primary
teacher education students. It is possible that for other groups of people, the
sources of self-efficacy may be quite different. Cognitive pedagogical mastery for
example, would obviously not apply outside of the specialist world of teacher
education. Secondly, it should be remembered that the sources of self-efficacy
identified as important in this particular course, may not equally apply in other
courses. Other primary science methods courses have achieved comparable levels
of self-efficacy through use of other sources such as enactive mastery experiences
in schools, vicarious experiences observing peers teaching in schools, and au-
thentic verbal persuasion from practicing teachers (Wingfield et al., 2000). The
extent to which the findings from this study can be applied to other methods
courses should therefore be determined on a case by case basis, according to course
structure and content.
Finally, there are several avenues for further research in this area. Other authors
have rightly argued that studies such as this provide no information about how long
the positive changes in self-efficacy will persist over time, after completion of the
course (Morrisey, 1981). It should also be emphasised that high efficacy does not
necessarily result in effective teaching Y it is a necessary first step, but the other
factors influencing the process should be clearly mapped. In addition, it is tempting
to speculate whether some sources of self-efficacy might have longer-lasting
352 D.H. PALMER

effects than others. Do successful enactive mastery experiences for example, have
more long-term effect than successful cognitive mastery experiences? The role of
perceived relevance should also be analysed, as it could be argued that this factor
was common to all the sources of efficacy identified in the present study. Research
that can provide the answers to these types of issues will potentially play an
important role in the design of future preparatory courses for primary teacher
education students.

References

Appleton, K. (1995). Student teachers’ confidence to teach science: Is more science

knowledge necessary to improve self-confidence? International Journal of
Science Education, 17, 357Y369.
Appleton, K., & Kindt, I. (1999). Why teach primary science? Influences on
beginning teachers’ practices. International Journal of Science Education, 21,
155Y168.
Appleton, K., & Kindt, I. (2002). Beginning elementary teachers’ development as
teachers of science. Journal of Science Teacher Education, 13, 43Y61.
Bandura, A. (1982). Self-efficacy mechanism in human agency. American
Psychologist, 37, 122Y147.
Bandura, A. (1997). Self-efficacy: The exercise of control. NY: W.H. Freeman
and Co.
Bell, G. L. (2001, March). Reflective journal writing in an inquiry-based science
course for elementary preservice teachers. Paper presented at the Annual Meet-
ing of the National Association for Research in Science Teaching, St. Louis, MO.
Cannon, J. R., & Scharmann, L. C. (1996). Influence of a cooperative early field
experience on preservice elementary teachers’ science self-efficacy. Science
Education, 80, 419Y436.
Cantrell, P., Young, S., & Moore, A. (2003). Factors affecting science teaching
efficacy of preservice elementary teachers. Journal of Science Teacher
Education, 14, 177Y192.
Enochs, L. G., & Riggs, I. M. (1990). Further development of an elementary
science teaching efficacy belief instrument: A preservice elementary scale.
School Science and Mathematics, 90, 695Y706.
Ginns, I. S., Watters, J. J., Tulip, D. F., & Lucas, K. B. (1995). Changes in pre-
service elementary teachers’ sense of efficacy in teaching science. School
Science and Mathematics, 95, 394Y400.
Jarrett, O. S. (1999). Science interest and confidence among preservice elementary
teachers. Journal of Elementary Science Education, 11, 47Y57.
Morrisey, J. T. (1981). An analysis of studies on changing the attitude of elemen-
tary student teachers toward science and science teaching. Science Education,
65, 157Y177.
 SOURCES OF SELF-EFFICACY IN A SCIENCE METHODS COURSE 353

Mulholland, J., & Wallace, J. (1996). Breaking the cycle: Preparing elementary
teachers to teach science. Journal of Elementary Science Education, 8, 17Y38.
Mulholland, J., & Wallace, J. (2001). Teacher induction and elementary science
teaching: Enhancing self-efficacy. Teaching and Teacher Education, 17,
243Y261.
Posnanski, T. J. (2002). Professional development programs for elementary science
teachers: An analysis of teacher self-efficacy beliefs and a professional de-
velopment model. Journal of Science Teacher Education, 13, 189Y220.
Rice, D. C., & Roychoudhury, A. (2003). Preparing more confident preservice
elementary science teachers: One elementary science methods teacher’s self-
study. Journal of Science Teacher Education, 14, 97Y126.
Schoon, K. J., & Boone, W. J. (1998). Self-efficacy and alternative conceptions of
science of preservice elementary teachers. Science Education, 82, 553Y568.
Settlage, J. (2000). Understanding the learning cycle: Influences on abilities to
embrace the approach by preservice elementary school teachers. Science
Education, 84, 43Y50.
Tosun, T. (2000). The beliefs of preservice elementary teachers towards science
and science teaching. School Science and Mathematics, 100, 374Y379.
Watters, J. J., & Ginns, I. S. (2000). Developing motivation to teach elementary
science: Effect of collaborative and authentic learning practices in preservice
education. Journal of Science Teacher Education, 11, 301Y321.
Wingfield, M. E., Freeman, L., & Ramsey, J. (2000, AprilYMay). Science teaching
self-efficacy of first year elementary teachers trained in a site based program.
Paper presented at the Annual Meeting of the National Association for
Research in Science Teaching, New Orleans, LA.