See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/233492478

Measuring academic behavioural confidence: The ABC scale revisited

Article  in  Studies in Higher Education · February 2009

DOI: 10.1080/03075070802457058

CITATIONS READS

39 4,945

2 authors:

Paul Sander Lalage Dorothy Sanders

Soon to be Teesside University Cardiff Metropolitan University
64 PUBLICATIONS   1,072 CITATIONS    59 PUBLICATIONS   1,169 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Mejora de la autorregulación en estudiantes universitarios a través de estrategias de enseñanza reguladas on-line View project

Gender differences in approach to university study View project

All content following this page was uploaded by Lalage Dorothy Sanders on 30 November 2015.

The user has requested enhancement of the downloaded file.

This article was downloaded by: [UWIC University of Wales Inst - Cardiff]
On: 29 January 2009
Access details: Access Details: [subscription number 789759699]
Publisher Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Studies in Higher Education

Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t713445574

Measuring academic behavioural confidence: the ABC scale revisited

Paul Sander a; Lalage Sanders a
a
University of Wales Institute, Cardiff, UK

Online Publication Date: 01 February 2009

To cite this Article Sander, Paul and Sanders, Lalage(2009)'Measuring academic behavioural confidence: the ABC scale
revisited',Studies in Higher Education,34:1,19 — 35
To link to this Article: DOI: 10.1080/03075070802457058
URL: http://dx.doi.org/10.1080/03075070802457058

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article may be used for research, teaching and private study purposes. Any substantial or
systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or
distribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss,
actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly
or indirectly in connection with or arising out of the use of this material.
 Studies in Higher Education
Vol. 34, No. 1, February 2009, 19–35

Measuring academic behavioural confidence: the ABC scale

revisited
Paul Sander* and Lalage Sanders

University of Wales Institute, Cardiff, UK

Taylor
Studies
10.1080/03075070802457058
CSHE_A_345873.sgm
0307-5079
Original
Society
602008
33
[email protected]
PaulSander
00000December
and
in
for
Article
Higher
(print)/1470-174X
Francis
Research
Education
2008
Ltd
into Higher
(online)
Education

The Academic Behavioural Confidence (ABC) scale has been shown to be valid
Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

and can be useful to teachers in understanding their students, enabling the design
of more effective teaching sessions with large cohorts. However, some of the
between-group differences have been smaller than expected, leading to the
hypothesis that the ABC scale many not be unidimensional and that inherent
subscales may be behaving in different ways, reducing the size of anticipated ABC
effects. This study aimed to analyse the factor structure of the ABC scale. Pre-
existing data sets were combined into a large composite data set (n = 865) of
undergraduate student respondents to the ABC scale. Exploratory factor analyses
using SPSS, and confirmatory factor analysis in AMOS, were carried out. A
reduced, 17-item ABC scale can be considered as having four factors, grades,
verbalising, studying and attendance. From the data sets, the discriminative power
of the factor structure has been confirmed, with the results providing further
criterion validity of the ABC scale.

Introduction
The Academic Behavioural Confidence (ABC) scale was first published as the
Academic Confidence Scale (Sander and Sanders 2003), with the suggestion that the
scale consisted of six subscales. The scale, renamed because it focuses on confidence
in actions and plans related to academic study (Sander and Sanders 2006a), is a
psychometric means of assessing the confidence that undergraduate university
students have in their own anticipated study behaviours in relation to their degree
programme.
Academic behavioural confidence is conceptualised as being how students differ
in the extent to which they have a ‘strong belief, firm trust, or sure expectation’ (Sander
and Sanders 2003, 3) of how they will respond to the demands of studying at university.
As such, ABC is distinct from the academic performance aspirations that students may
have, although the two may be related to some extent. The question of the relationship
between ABC and academic aspirations will be considered later.
The ABC scale has been contrasted with other measures of confidence (see Sander
and Sanders 2006a), including Bandura’s self efficacy (Bandura 1977, 1986, 1993),
expectancy value theory (Eccles and Wigfield 2002; Wigfield and Eccles 2000) and
self-concept and self-esteem models (McCoach and Siegle 2003; Skaalvik and Rankin
1995). Sander and Sanders (2006a) argue that the ABC scale is designed to provide a
global measure of academic confidence, with academic confidence being seen as a

*Corresponding author. Email: [email protected]

ISSN 0307-5079 print/ISSN 1470-174X online

© 2009 Society for Research into Higher Education
DOI: 10.1080/03075070802457058
http://www.informaworld.com
 20 P. Sander and L. Sanders

broader term than the more focused concept of self-efficacy. Thus, ABC measures
seek to provide a general measure of a student’s confidence in undertaking their
academic course, although an analysis of individual statements can also be useful in
providing more focused measures more akin to efficacy measures (see Sander and
Sanders 2005, 2006b). As such, the ABC scale measure would be responsive to posi-
tive interventions, such as personal tutor support or revised study habits, and adversely
affected by negative study experiences, and thus ABC scores may not be stable over
time. Should confidence measures be required within different aspects of a course,
self-efficacy measures would be more appropriate. As such, ABC is seen to be concep-
tually different from self-efficacy, as discussed by Sander and Sanders (2006a).
Academic confidence may stem from the same four sources proposed for self-
efficacy: mastery experience, vicarious experience, verbal persuasion and physio-
Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

logical states (Bandura 1977, 1986, 1993). It is likely to be subject to change as

experience impinges upon expectation (Sander and Sanders 2006a). Indeed, it is
likely that this explains, at least in part, the ABC differences between the medical
students and the psychology students used in the original validation of the ABC
scale, although clearly there are other factors involved (Sander and Sanders 2003;
Sanders and Sander 2006b).
The ABC scale was developed within an ethos of using survey techniques to try to
understand students within the large student groups that many university lecturers
have to teach (Cassidy and Eachus 2000). The rationale for this approach has been
discussed in some detail by Sander (2004), but essentially it is argued that, with large
classes, there is little or no opportunity for the informal interactional discourse possi-
ble within small groups, that allows the teacher to understand her or his students, or
help and guide them by effective teaching. Rather than teachers assuming that
students have a particular outlook, knowledge base, orientation to learning or what-
ever, psychometric techniques can be used to help understand a cohort of students. It
was for this reason that a global or aggregate confidence measure was developed,
rather than context or task specific self-efficacy measures (Pajares 2002).
Profile information collected at the beginning of a university course could
provide a known and clearly understood starting position, from which learning envi-
ronments can be designed. This would be particularly important, because the transi-
tion from school or college to university places many demands on undergraduates
(Chemers, Hu, and Garcia 2001) and the culture in which university students find
themselves can make it difficult for many of them to begin effective studying at
degree level (Sander 2004). Indeed, school leavers entering university are likely to
need focussed input (Laurillard 1993; Northedge 2003; Scott 1999) to give them the
competence and confidence needed to become reflective, deep and independent
learners.
The ABC scale was developed through an iterative process, with colleagues iden-
tifying appropriate academic behaviours that students face. It comprises 24 state-
ments, each of which respondents rate on 5-point scale. The scale’s psychometric
properties were explored in a preliminary study of first-year undergraduates, 102 on a
psychology course in a new university, and 182 on a medical degree in a traditional
university (Sander and Sanders 2003, 2006a; Sanders and Sander 2007). The ABC
scale demonstrated a high level of internal reliability (Cronbach’s alpha 0.88). An
overall ABC score was computed as the mean response over the 24 items, and the
median for all students in the study was 3.8. A comparison of the overall ABC scores
showed that the medical students scored higher, i.e. were more confident, than the
 Studies in Higher Education 21

psychology students, (medians 3.9 and 3.7 respectively, z = 2.07, p < .05 one-tailed).
This is the expected outcome, showing that ABC scores were significantly greater for
the medical students, which prompted the development of the ABC scale (Sander and
Sanders 2006a). Therefore, it offers construct validity of the scale, although the differ-
ence could well have been expected to be greater.
The scale’s concurrent validity was assessed by asking respondents to estimate their
final-year degree mark. This correlated significantly (p < .05) with their ABC score,
indicating that those who were confident that they could produce the behaviours
required for academic study were those who felt they would do well academically
(Sander and Sanders 2006a). Further, this significant correlation indicates an answer
to a question posed earlier, suggesting that there may indeed be a relationship between
ABC and academic aspirations.
Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

The relatively small but significant difference in ABC scores between the medical
and psychology students, and the situation where, looking at individual ABC statement
scores, the psychology students showed more confidence than the medical students on
some items (Sander and Sanders 2003), highlighted the need to understand the struc-
ture of the scale better. Were there underlying factors in the scale for which the medical
students were more confident, and others that showed no difference or even a differ-
ence in favour of the psychology students? With a robust understanding of the factor
structure of the ABC, changes in factor scores could be used to detect, more sensi-
tively, confidence changes across time, situations and with individuals against cohort
norms. This article focuses specifically on the factor structure of the ABC scale, using
exploratory and confirmatory factor analysis through structural equation modelling to
explore further the properties of the scale.

Method
Design
This study used pre-existing data sets to explore the invariance (Martin 2006) of any
emergent factor structure of the ABC scale.

Materials
All data were collected using the ABC scale as originally published (Sander and
Sanders 2003, 2006a; see Table 1) within the last six years.

Participants
The analysis presented in this article was conducted on three data sets that were
combined into one composite data set. The three data sets were:
1. Psychology undergraduates. This data set consisted of data collected on five differ-
ent occasions between 2001 and 2006 in a post-1992 university in South Wales. The
total number of participants was 507. Typically, the average A (Advanced) level score
for these psychology students is C, C, D (A level points are gained from examinations
at the end of school, aged 18 years. Entry to university courses in the UK is usually
by attaining a minimum number of points set by the course’s admission team. The
scoring details for entry qualifications can be found at: http://merlinhelpstudents.com/
studentlife/beforeuniversity/ucas/ucaspointstable.asp)
 22 P. Sander and L. Sanders

Table 1. Academic behavioural confidence.

How confident are you that you will be able to:
1. Study effectively on your own in independent/ Not at all Very
private study confident confident






2. Produce your best work under examination Not at all Very
conditions confident confident






3. Respond to questions asked by a lecturer in front of Not at all Very
a full lecture theatre confident confident






4. Manage your workload to meet coursework Not at all Very
Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

deadlines confident confident

5. Give a presentation to a small group of fellow Not at all Very
students confident confident






6. Attend most taught sessions Not at all Very
confident confident






7. Attain good grades in your work Not at all Very
confident confident






8. Engage in profitable academic debate with your Not at all Very
peers confident confident






9. Ask lecturers questions about the material they are Not at all Very
teaching, in a one-to-one setting confident confident






10. Ask lecturers questions about the material they are Not at all Very
teaching, during a lecture confident confident






11. Understand the material outlined and discussed with Not at all Very
you by lecturers confident confident






12. Follow the themes and debates in lectures Not at all Very
confident confident






13. Prepare thoroughly for tutorials Not at all Very
confident confident






14. Read the recommended background material Not at all Very
confident confident






15. Produce coursework at the required standard Not at all Very
confident confident






16. Write in an appropriate academic style Not at all Very
confident confident






17. Ask for help if you don’t understand Not at all Very
confident confident






 Studies in Higher Education 23

Table 1. (Continued).
How confident are you that you will be able to:
18. Be on time for lectures Not at all Very
confident confident






19. Make the most of the opportunity of studying for a Not at all Very
degree at university confident confident






20. Pass assessments at the first attempt Not at all Very
confident confident






21. Plan appropriate revision schedules Not at all Very
Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

confident confident






22. Remain adequately motivated throughout Not at all Very
confident confident






23. Produce your best work in coursework assignments Not at all Very
confident confident






24. Attend tutorials Not at all Very
confident confident

2. Medical students. This data set comprised data on 182 first-year medical students
studying at a pre-1960 university in the English Midlands. These data were
collected in 2001. The average A level score for this group is in excess of A, A, A.
Given the difference in both A level and General Certificate of Secondary Educa-
tion (GCSE) points (national statistics for 2005 show that the qualifications of
medical students were half as good again as those of psychology students: see http:/
/www.ucas.com/figures/reports/tariff.xls) between the medical and psychology
students, both statistically significant at p < .001, it would not be unreasonable to
predict considerable differences in ABC scores between the two groups. Indeed this
was the starting point for the development of the ABC (see Sander and Sanders
2003, 2006a).
3. Health Care students. The data in this data set came from students on six health care
courses (Speech and Language Therapy, n = 37; Nutrition, n = 36; Podiatry, n = 32;
Dental Technology, n = 13; Health and Social Care, n = 34 and Housing, n = 24) in
the new university in South Wales in 2003. The total number of participants was 176.
Students in this data set had very variable A level scores (which is addressed in the
results).
These three data sets combined into one composite data set gives a total n of 865.

Method of analysis
Exploratory factor analysis (principal components) in SPSS 14.0 was conducted
using oblique rotation, as there was good reason to believe that the factors would
be correlated. Possible resultant factor models for the ABC scale were subject to
 24 P. Sander and L. Sanders

confirmatory factor analysis (CFA) in AMOS 6.0 and compared. Following Miles
and Shevlin (2007) and Prosser and Trigwell (2006), a range of fit indices in addi-
tion to the chi-squared statistic was used to assess model fit. In addition, the ECVI
(expected cross validation index) statistic was used to make direct comparisons of
models in CFA.
To facilitate the two stages of factor analysis, the data set (n = 865) was randomly
split into two. File_EFA (n = 432) was used for the exploratory factor analysis. Models
suggested here were then considered by confirmatory factor analysis using file_CFA
(n = 433). All tests of difference when considering the discriminative power of the
subscales in the resultant model employed multivariate analysis of variance. Post hoc
testing was with the Scheffe test.
Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

Results
Overview
There are three sections to the results presented here. The first section presents the
outcome of exploratory factor analysis on the ABC scale. The second section presents
the outcome of confirmatory factor analysis. Finally, the discriminative power and the
invariance of the factor structure were considered.

1. Exploratory factor analysis

Table 2 shows the output from exploratory factor analysis and suggests that the 24-
item ABC scale could be reduced to six factors or subscales. This will be one model
that will be considered through confirmatory factor analysis.
Inspection of the scree plot (Figure 1) suggests that a four-factor explanation may
be more parsimonious. This will be the second model to be considered.
Two further models being put forward for confirmatory factory analysis are the
Figure 1. Scree plot from exploratory factor analysis on file_EFA.

original six-factor model (Sander and Sanders 2003) and a unidimensional model with
all 24 ABC statements loading onto a single confidence factor.

2. Confirmatory factor analysis

Using the CFA file, confirmatory factor analysis was conducted using AMOS on the
four models. The results from CFA were considered using the absolute fit statistics;
CFI (comparative fit index), TLI (Tucker-Lewis coefficient), RMSEA (root mean
square error of approximation) and chi-squared, and also the ECVI as a comparative
fit index to facilitate the comparison of the four models. To interpret the output statis-
tics, CFI and TLI values close to 1 indicate a very good fit; RMSEA values less than
0.05 indicate a good fit and values as high as 0.08 represent reasonable errors of
approximation in the population (Byrne 2001), and the lower the ECVI value the
better the model (Mackenzie et al. 2004). Table 3 summarises these statistics for the
four models.
Inspection of Table 3 shows that the four-factor model is superior to the others,
although the absolute fit statistics suggest that it may warrant some further develop-
ment. The 17 ABC statements that contribute to this preferable four-factor model are
shown in Table 4.
Figure 2 shows the outputs from AMOS for this 4 factor model.
 Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

Table 2. Pattern matrices of the ABC scale from EFA.

Grades Studying Verbalising Attendance Understanding Requesting
[20] Pass assessments at the first attempt 0.83 0.12 −0.11 −0.14
[15] Produce coursework at the required standard 0.74 −0.10 0.16 0.12
[16] Write in an appropriate academic style 0.67 −0.13 0.14 0.16
[7] Attain good grades in your work 0.66
[23] Produce your best work in coursework assignments 0.55 0.14 0.28 −0.16
[2] Produce your best work under examination conditions 0.51 0.16 0.16 −0.18 −0.11
[21] Plan appropriate revision schedules 0.80 −0.13
[1] Study effectively on your own in independent/private study 0.72 −0.30 0.22
[22] Remain adequately motivated throughout 0.62 0.19 0.11 −0.10
[4] Manage your workload to meet coursework deadlines 0.22 0.56
[3] Respond to questions asked by a lecturer in front of a full lecture theatre 0.11 0.85 −0.16
[5] Give a presentation to a small group of fellow students 0.81 0.18 −0.21
[8] Engage in profitable academic debate with your peers 0.67 0.29
[10] Ask lecturers questions about the material they are teaching, during a 0.58 −0.26 0.19 0.38
lecture
[24] Attend tutorials 0.86
[6] Attend most taught sessions 0.82 0.13
[18] Be on time for lectures −0.14 0.35 0.40 0.18
[19] Make the most of the opportunity of studying for a degree at university 0.17 0.24 0.29 0.21
[13] Prepare thoroughly for tutorials −0.18 0.25 0.15 0.73
[12] Follow the themes and debates in lectures 0.30 −0.20 0.13 0.72
[11] Understand the material outlined and discussed with you by lecturers 0.35 −0.10 0.68
[14] Read the recommended background material −0.19 0.28 −0.12 0.10 0.68
[9] Ask lecturers questions about the material they are teaching, in a one-to- 0.85
Studies in Higher Education

−0.13
one setting
[17] Ask for help if you don’t understand 0.13 0.83
25
 26 P. Sander and L. Sanders
Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

Figure 1. Scree plot from exploratory factor analysis on file_EFA.

3. Discriminative power of the proposed subscales

Figure 2. The four factor model.

To assess the discriminative power of the favoured four-factor model, the three
constituent data sets from the composite data set for factor analysis were considered
separately.
The subscale differences between the different data sets (medical students,
psychology students and health care students) are shown in Figure 3 and Table 5.
It can be seen that there is some variability between the medical students, psychol-
Figure 3. Mean ABC subscales scores for the three data sets.

ogy students, and all the aggregated health care students, for the subscales.
Multivariate analysis of variance shows significant differences for the factors
studying (F(2, 859) = 4.9, p < .01, r = .14) and attendance (F(2, 859) = 9.6, p < .001,
r = .3). Post hoc testing shows that the significant effects for studying are that the
medical students were more confident than the health care group. For attendance, post

Table 3. Results for confirmatory factor analysis for the 4 models, ordered by the ECVI
statistic.
Model CFI TLI RMSEA Chi-squared DF p ECVI
4-factor model .92 .889 .065 276 98 p < 0.001 .893
6-factor model from Sander .902 .868 .063 416 155 p < 0.001 1.491
and Sanders (2003)
6-factor model from .884 .851 .065 602 215 p < 0.001 1.789
exploratory factor analysis
Uni-dimensional model .656 .590 .106 1475 252 p < 0.001 3.757
 Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

Table 4. The ABC factor structure.

Grades Verbalising Studying Attendance
[2] Produce your best work under [3] Respond to questions asked by a [1] Study effectively on your [6] Attend most taught
examination conditions lecturer in front of a full lecture own in independent/private sessions
theatre study
[7] Attain good grades in your work [5] Give a presentation to a small group [4] Manage your workload to [18] Be on time for lectures
of fellow students meet coursework deadlines
[15] Produce coursework at the [8] Engage in profitable academic [21] Plan appropriate revision [24] Attend tutorials
required standard. debate with your peers schedules
[16] Write in an appropriate [10] Ask lecturers questions about the [22] Remain adequately
academic style. material they are teaching, during a motivated throughout
lecture
[20] Pass assessments at the first
attempt
[23] Produce your best work in
coursework assignments
Studies in Higher Education
27
 28 P. Sander and L. Sanders

3.18
0, .86
1
e10 ABC2
3.73
0, .23 .87
1
e9 ABC7 3.90 1.06
0, .27 0, .25
1 1.10
e8 ABC15 3.75
0, .44 1.13 G
1
e7 ABC16 1.15
3.67
0, .26
1 1.00
e6 ABC20
3.89
0, .37
1
e5 ABC23
Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

.23

3.88
0, .52
1
e22 ABC1
3.82 .69
0, .41
1 0, .36
.71
e21 ABC4 .14
3.55
0, .53 1.03 S
1
e20 ABC21 1.00
3.76
0, .35
1
e19 ABC22

.09
.15
2.75
0, .48
1
e4 ABC3
3.39 1.19
0, .85 0, .59
1 .94
e3 ABC5
0, .48
3.61
.67 V .16
1
e2 ABC8 1.00
3.03
0, .63
1
e1 ABC10

.09

4.67
0, .13
1
e14 ABC6 .89 0, .25
4.37
0, .30 1.02
1 A
e13 ABC18
4.56 1.00
0, .15
1
e11 ABC24

Figure 2. The four factor model.

hoc testing shows that the medical students were more confident than both the
psychology students and the health care students.
However, the student groups in the aggregated health care data set are very varied
not only in terms of career aspiration but also in A level points (F(5, 172) = 12.303,
 Studies in Higher Education 29

Subscale

34 755 Grades
Studying
Verbalising
Attendance

4
Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

588

209 744
555 691

73 302 571
2 742
534
31 571
646

358 358 618 747

Psychology Medical Healthcare

Figure 3. Mean ABC subscales scores for the three data sets.

p < .001; r = 0.51), so it is not unreasonable to suppose that there would be differences
in ABC scores between the disciplines this data set comprises (Figure 4).
Given the small sample for each of these constituent student groups they could
Figure 4. Mean ABC subscales scores for the student groups within the Health Care data set.

not each be readily compared directly with the medical and psychology students.
Comparisons within this data set using multivariate analysis of variance showed
significant effects between the different courses within the health care group for
grades (F(5, 172) = 3.22, p < .05; r = 0.25)), with post hoc testing showing that that
the housing students were significantly lower in confidence than both the nutrition
and speech and language therapy students for grades. There was also an overall
significant effect for studying (F(5, 172) = 2.44, p < .05, r = .2
Differences in ABC factor scores were also considered by gender (Table 6) as
there is good reason to believe that there are meaningful gender differences in
approaches to higher education (Sander and Sanders 2006b, 2007). Analysis by
gender could only be carried out for the healthcare students, as this was the only data
set for which gender was recorded. Multivariate analysis of variance shows that men
were significantly more confident in each of three scales, grades (F(1, 173) = 10.22,
p < .005, r = 0.22), verbalising (F(1, 173) = 9.58, p < .005, r = 0.22) and studying (F(1,
173) = 3.94, p < .05, r = 0.13).
 30 P. Sander and L. Sanders

Table 5. Descriptive statistics for ABC sub scales for each subject group, ordered by mean
ABC.
Student group Mean ABC Grades Verbalising Studying Attendance
Medical Mean 3.88 3.73 3.31 3.88 4.66
n = 182 SD .46 .59 .69 .67 .46
Psychology Mean 3.81 3.66 3.17 3.76 4.47
n = 507 SD .44 .57 .86 .61 .56
Speech and Language Mean 3.68 3.75 3.46 3.83 4.55
therapy SD .77 .46 .64 .46 .453
n = 37
Nutrition Mean 3.71 3.82 3.05 3.77 4.29
Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

n = 37 SD .10 .65 .83 .58 .67

Podiatry Mean 3.56 3.69 3.07 3.65 4.47
n = 31 SD .11 .58 .75 .53 .49
Dental Mean 3.54 3.67 3.46 3.85 4.51
n = 13 SD .14 .51 .76 .59 .55
Health and Social Care Mean 3.45 3.56 3.17 3.52 4.47
n = 34 SD .11 .68 .85 .635 .565
Housing Mean 3.11 3.24* 2.97 3.43 4.36
n = 24 SD .13 .69 .92 .64 .69

Discussion
Confirmatory factor analysis suggests that the ABC scale can best be considered as
having four factors, grades, verbalising, studying and attendance. In the process, the
original 24-item scale is reduced to 17 items as shown in Table 4. The discriminative
power and, importantly, the invariance (Martin 2006) of the factor structure has been
confirmed. In total, these findings provide further criterion validity of the ABC scale.
The factors and the discriminatory analyses, though, require further consideration.

Factors: under student control or part of the student–teacher dyad?

In discussing the exploratory factor analysis, attention needs to be drawn to two
types of factors emergent from this analysis. Firstly, there are those that refer to
behaviours that were substantially under the student’s control, like attendance and
studying (although it is acknowledged that teachers will have some influence on
their likelihood). In contrast, grades and, to a lesser extent, verbalising are behav-
iours and processes that are more obviously the product of the student–teacher dyad.
The significant differences are grouped in Table 7. In each of the four factors,
differences have been detected, although the fewest were for verbalising, which
only showed an interesting gender effect (discussed later).

Healthcare students. When exploring the differences in ABC at factor level in the six
health care courses, statistically significant differences in the grades and studying
subscales were found. Given the difference in qualifications at entry of the health care
groups, a difference in grades confidence could be expected (however, grades does not
 Studies in Higher Education 31

Subscale

Grades
5.00 95 139
Attendance
Studying
Verbalising

4.00
Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

3.00 100
126

126
55 84
69
2.00

1.00

Health and Housing Nutrition Dental SLT Podiatry

social care

Figure 4. Mean ABC subscales scores for the student groups within the Health Care data set.

show reasonably expected differences elsewhere). In the same data set, men come out
with higher confidence for grades, verbalising and studying. This accords with other
research, which shows that men are more likely than women students to rate their
academic abilities highly, controlling for differences in performance, and are less

Table 6. Descriptive statistics for ABC sub scales by sex in the Health Care data set.
Sex Grades Verbalising Studying Attendance
Male Mean 3.93 3.56 3.85 4.31
n = 34 SD .46 .69 .59 .57
Female Mean 3.56 3.09 3.63 4.47
n = 141 SD .63 .81 .58 .57
 32 P. Sander and L. Sanders

Table 7. Summary of statistically significant factor differences.

The product of the student–teacher dyad Substantially under the student’s control
Grades Verbalising Studying Attendance
Nutrition > Housing *men > women *men > women Medics > Psychologists
Speech & Language Medics > Health Care Medics > Health Care
Therapy > Housing
*men > women
*Analysis only possible in Health Care data set.
Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

likely to be adversely affected by the transition into higher education, perhaps in part
because men may be more self-centred and less attuned to social interaction issues
than women (Jackson 2003).
The difference in verbalising between the male and female students is supported
by a substantial literature considering the difference in verbal behaviour between male
and female students, particularly in tutorial settings (Read, Archer, and Leathwood
2003; Somners & Lawrence 1992; Sternglanz & Lyberger-Ficek 1997).
Interestingly, the health care data set did not show a difference in confidence for
attendance between men and women. Woodfield, Jessop, and McMillan (2006)
argued that attendance itself is important in explaining the variance in degree perfor-
mance, in that students gain something from the formal teaching situation. Indeed, in
their study, attendance explained degree performance over and above measures of
cognitive ability and personality variables, and, interestingly, men students were more
likely to be absent and more likely to under-report their absenteeism. Woodfield and
colleagues also suggest that the difference in attendance rates can be explained by
women students’ greater compliance to institutional requirements. Certainly, it has
been established that diligence and conscientiousness are traits taken up by girls as
part of their construction of femininity (Francis, Robson, and Read 2001). Confidence
and gender are discussed further by Sander and Sanders (2006b, 2007). In hindsight,
collecting gender data for all participants would have been preferable, but at the start
of this project we were unaware of its importance.

Medical students. The significantly greater confidence in studying expressed by the

medical students against both the psychology students and the health care students
would be expected, given the medical students’ superior results in school. Indeed,
mastery experience has been identified as the main source of self-efficacy (Bandura
1977, 1986, 1993; Pajares 2002).
The medical students had a significantly greater confidence than the psychology
students and the health care students in their likelihood to attend (attendance score),
which, following Woodfield, Jessop, and McMillan (2006), might determine, at least
in part, the academic performance of the two student groups. However, there was not
a significant different in grades, which could reasonably be expected to show a differ-
ence, given that the medical students had almost, on average, twice the A level points
in comparison to the psychology group. If the grade confidence was linked to actual
performance, the expectations between the two groups should be very different.
Maybe grade confidence, as measured in the ABC subscale grades, is different from
actual performance? Longitudinal research is required to monitor ABC through a
 Studies in Higher Education 33

degree programme, and map it and its component scales onto academic performance
and attendance. From this, the symbiotic nature of the relationship between ABC,
performance and attendance should become clearer. It is possible that both medical
and psychology students were expressing a confidence – a confidence that was not
statistically significantly different – in their ability to perform to their expectations.
Both student groups had succeeded in meeting their expectations, as both were on their
chosen university courses. That a significant difference in studying was found between
these two groups supports this argument.
This explanation would need to be reconciled with the differences in grades confi-
dence between the nutrition and housing students, speech and language therapy
students and housing students and between men and women. Research is required to
explore the frames of reference used in rating confidence in grades when engaging
Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

with the ABC scale. A qualitative strategy could be best in this context. Another
confounding factor that should be considered is motivation. Whilst ABC, like self-
efficacy, is seen as a motivational construct (Pajares 2002), explicit measures of moti-
vation might help to unravel the complexities of interpreting the contrasting findings.
The low effect sizes will be a product of the greater variability within groups than
between groups, which needs to be borne in mind when interpreting the results and
considering their implications. Further research, in addition to longitudinally tracking
ABC scores, could usefully look at students on other quite different courses to estab-
lish the validity of the ABC factorial structure for university undergraduates in
general, rather than just those used in this study.
An additional test of the validity of the ABC scale would come from comparing it
with other similar psychometric scales, although, as discussed by Sander and Sanders
(2006a), there is no directly comparable scale. Indeed, if there had been, the ABC
would not have been developed! However, it could be useful to compare ABC scores
with more general measures of self-esteem, such as the Rosenberg Self-Esteem Scale
(1989), and also to see how the scores compare with measures of academic self-
concept using, for example, the SDQIII scale extensively developed by Marsh (2003).

Conclusion
The three discriminatory comparisons made (health care students by course; health
care students by sex; psychology, medical and health care students collectively) have
shown statistically significant differences across the analyses in each of the four
factors (Table 7).
The results further reinforce the validity of the ABC scale, although the lack of a
statistical significance in grades between the psychology students and the medical
students demands further consideration; what exactly are the students rating when they
are responding to the statements in the grades factor?
Likewise, a difference in attendance for the sex comparison in the health care data
set might have been expected. It would be informative to consider the correspondence
between confidence in attendance and actual attendance. Alternatively, the respon-
dents may have been responding to create a socially desirable image given that the
ABC was administered in all cases by members of the academic staff. The high mean
attendance confidence ratings provide some support for this suggestion.
Consideration of the comparison between the medical students and the psychology
students suggests a motivational basis to academic confidence, in that the psychology
students seem resigned to a poor attendance and study record. This is despite these
 34 P. Sander and L. Sanders

behaviours being largely under the student’s control, and more commitment in
this respect possibly leading to better academic performance. There could be a self-
fulfilling prophecy of underperformance here. If so, teachers working with student
groups comparable to the psychology students in this investigation should consider
carefully how best to further these students’ educational careers. They could be
dismissed as feckless, and any additional help given to them could serve to underline
their low academic behavioural confidence. Maybe greater efforts have to be made to
teach such students to aspire to high targets and help the students to reach them? If
the ‘weaker’ psychology students could be encouraged to ‘set their sights higher’ and
supported to achieve at this higher level, through mastery and vicarious experience,
greater confidence could emerge. In this way, in terms of both academic achievement
and personal growth, students really would be benefiting from our educational system.
Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

References
Bandura, A. 1977. Self-efficacy: Toward a unifying theory of behavioural change. Psycholog-
ical Review 84, no. 2: 191–215.
Bandura, A. 1986. Social foundations of thought and action. Englewood Cliffs, NJ: Prentice-
Hall.
Bandura, A. 1993. Perceived self-efficacy in cognitive development and functioning.
Educational Psychologist 28, no. 2: 117–48.
Byrne, B. 2001. Structural equation modelling in AMOS: Basic concepts, applications, and
programming. Mahwah, NJ: Lawrence Erlbaum.
Cassidy, S., and P. Eachus. 2000. Learning style, academic belief systems, self-report student
proficiency and academic achievement in higher education. Educational Psychology 20,
no. 2: 307–20.
Eccles, J.S., and A. Wigfield. 2002. Motivational beliefs, values and goals. Annual Review of
Psychology 53: 109–32.
Chemers, M.M., L. Hu, and B.F. Garcia. 2001. Academic self-efficacy and first-year college
student performance and adjustment. Journal of Educational Psychology 93, no. 1: 55–64.
Francis, B., J. Robson, and B. Read. 2001. An analysis of undergraduate writing styles in the
context of gender and achievement. Studies in Higher Education 26, no. 3: 313–26.
Jackson, C. 2003. Transitions into higher education: Gendered implications for academic self-
concept. Oxford Review of Education 29, no. 3: 331–46.
Laurillard, D. 1993. Rethinking university teaching: A framework for the effective use of
educational technology. London: Routledge.
Mackenzie, C.S., V.J. Knox, W.L. Gekoski, and H.L. Macauley. 2004. An adaptation and exten-
sion of the Attitudes Towards Seeking Professional Psychological Help Scale. Journal of
Applied Social Psychology 34, no. 11: 2410–35.
Marsh, H.W. 2003. Self Description Questionnaire III (SDQIII). http://self.uws.edu.au/Instru-
ments/packages.htm
Martin, A.J. 2006. Personal bests. British Journal of Educational Psychology 76, no. 4: 803–25.
Miles, J., and M. Shevlin. 2007. A time and a place for incremental fit indices. Personality
and Individual Differences 42: 869–74.
McCoach, D.B., and D. Siegle. 2003. The structure and function of academic self-concept in
gifted and general education students. Roeper Review, Winter: 61–5.
Northedge, A. 2003. Rethinking teaching in the context of diversity. Teaching in Higher
Education 8, no. 1: 17–32.
Pajares, F. 2002. Overview of social cognitive theory and of self-efficacy. http://
www.emory.edu/EDUCATION/mfp/eff.html
Prosser, M., and K. Trigwell. 2006. Confirmatory factor analysis of the Approaches to Teaching
Inventory, British Journal of Educational Psychology 76, no. 2: 405–19.
Read, B., L. Archer, and C. Leathwood. 2003. Challenging cultures? Student conceptions of
‘belonging’ and ‘isolation’ at a post-1992 university. Studies in Higher Education 28,
no. 3: 261–77.
 Studies in Higher Education 35

Rosenberg, M. 1989. Society and the adolescent self-image. Revised ed. Middletown, CT:
Wesleyan University Press.
Sander, P. 2004. Increasing student numbers: Diminishing tutor insight? Psychology Learning
and Teaching 4, no. 1: 15–21.
Sander, P., and L. Sanders. 2003. Measuring confidence in academic study: A summary
report. Electronic Journal of Research in Educational Psychology and Psychopeda-
gogy 1, no. 1. http://www.investigacion-psicopedagogica.org/revista/articulos/1/english/
Art_1_1.pdf
Sander, P., and L. Sanders. 2005. Giving presentations: The impact on students’ perceptions.
Psychology Teaching Review 11, no. 1: 25–39.
Sander, P., and L. Sanders. 2006a. Understanding academic confidence. Psychology Teaching
Review 12, no. 1: 29–39.
Sander, P., and L. Sanders. 2006b. Rogue males: Sex differences in psychology students.
Electronic Journal of Research in Educational Psychology and Psychopedagogy. http://
Downloaded By: [UWIC University of Wales Inst - Cardiff] At: 14:24 29 January 2009

www.investigacion-psicopedagogica.org/revista/articulos/8/english/Art_8_89.pdf
Sander, P., and L. Sanders. 2007. Gender, psychology students and higher education. Psychology
Learning and Teaching 6, no. 1: 33–6.
Sanders, L., and P. Sander. 2007. Academic behavioural confidence: A comparison of medi-
cal and psychology students. Journal of Research in Educational Psychology and
Psychopedagogy. http://www.investigacion-psicopedagogica.org/revista/articulos/13/
english/Art_13_150.pdf
Scott, S.V. 1999. The academic as a service provider; Is the customer ‘always right’?’ Journal
of Higher Education Policy and Management 21, no. 2: 193–202.
Skaalvik, E.M., and R.J. Rankin. 1995. A test of the internal/external frame of reference
model at different levels of math and verbal self-perception. American Educational
Research Journal 32, no. 1: 161–84.
Somners, E., and S. Lawrence. 1992. Women’s ways of talking in teacher-directed and
student-directed peer response groups. Linguistics and Education 4: 1–36.
Sternglanz, S.H., and S. Lyberger-Ficek. 1977. Sex differences in student–teacher interactions
in the college classroom. Sex Roles 3, no. 4: 345–52.
Wigfield, A., and J.S. Eccles. 2000. Expectancy-value theory of achievement motivation.
Contemporary Educational Psychology 25: 68–81. http://www.idealibrary.com
Woodfield, R., D. Jessop, and L. McMillan. 2006. Gender differences in undergraduate atten-
dance rate. Studies in Higher Education 31, no. 1: 1–22.

View publication stats