Antecedents and consequences of accounting students' approaches to learning: A cluster analytic approach

Accepted Manuscript Antecedents and consequences of accounting students’ approaches to learning: A cluster analytic approach Angus Duff, Rosina Mladenovic PII:

S0890-8389(14)00046-8

DOI:

10.1016/j.bar.2014.06.003

Reference:

YBARE 668

To appear in:

The British Accounting Review

Received Date: 12 February 2013 Revised Date:

28 May 2014

Accepted Date: 3 June 2014

Please cite this article as: Duff, A., Mladenovic, R., Antecedents and consequences of accounting students’ approaches to learning: A cluster analytic approach, The British Accounting Review (2014), doi: 10.1016/j.bar.2014.06.003. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Antecedents and consequences of accounting students’ approaches to learning: A cluster analytic approach

Accounting and Finance Research and Knowledge Exchange Institute University of the West of Scotland University Avenue Ayr Campus Ayr KA8 0SX Scotland

2

M AN U

[email protected]

SC

1

RI PT

Authors: Angus Duff 1 * and Rosina Mladenovic 2

TE D

Discipline of Accounting University of Sydney Business School The University of Sydney NSW 2006 Australia

AC C

EP

[email protected]

* Corresponding author

ACCEPTED MANUSCRIPT

Antecedents and consequences of accounting students’ approaches to learning: A cluster analytic approach

Abstract

RI PT

A dominant theoretical model in the approaches to learning literature is the PresageProcess-Product Model (3P). This model, in its various forms depicts the complex interrelationships between what students bring to their studies (presage), how they engage in learning (process), and the outcomes of their learning processes (product). While an

SC

extensive literature addresses accounting students’ approaches to learning, relatively few studies consider all aspects of the 3P model simultaneously. This study explores a comprehensive number of antecedents including: accounting students’ expectations of

M AN U

learning, prior study of accounting, gender, major, and language (presage); their reflections on learning (processes); and their performance outcomes (product). A cluster analytic approach is employed to: (i) identify students’ expectations of learning at an intra-individual level; and (ii) examine their relationship to approaches to learning and academic outcomes. Two inventories were administered to a diverse sample of 1,553 first-year undergraduate students studying accounting at two universities in Australia. One inventory measured their

TE D

expectations of learning accounting; the other, their study processes. Three distinct groups of students were identified. The cluster with the most optimistic expectations of learning accounting had the most positive approaches to learning. The cluster with the most The

EP

pessimistic expectations of learning accounting had a maladaptive learning profile.

implications of this investigation include that it: (i) provides support for the holistic exploration of presage, process, and product factors; (ii) highlights the key interrelationships

AC C

between all factors for student learning in context; and (iii) sheds light on the plethora of inconsistent findings in previous research with respect to gender, language, major and previous study of accounting. A further implication of these findings is that providers of accounting education should consider having two types of introductory accounting course: a technical one for accounting majors; and for non-accounting majors a conceptual one based on how to use accounting information.

Keywords: Pathways to accounting; the first course in accounting; expectations of learning accounting; approaches to learning; academic performance. 1

ACCEPTED MANUSCRIPT

Acknowledgement

RI PT

We are grateful for the helpful and constructive comments of the two anonymous reviewers and the editors. In addition we wish to express our thanks to our colleague Ursula Lucas who’s inspiration and groundbreaking work provided us with the foundation for this interesting research journey into students’ expectations of learning accounting.

1. Introduction

Students’ approaches to learning (SAL) have been the focus of vigorous research efforts Marton & Saljo’s (1976) pioneering work identified two

SC

over the past four decades.

contrasting approaches in Scandinavian students’ approach to reading academic articles and

M AN U

texts, namely a ‘deep’ approach and a ‘surface’ approach. A deep approach entails looking for meaning in the matter being studied and relating it to other experiences and ideas with a critical approach. By contrast, a surface approach describes a reliance on rote-learning

TE D

and memorization in isolation to other ideas.

The imperial position SAL research holds in the higher education literature has been largely motivated by constructivist theories of learning that seek to make teaching more student-

EP

centred (Baeten, Kynt, Struyven & Dochy, 2010). Specifically, constructivist approaches to learning have four key characteristics: learners construct their own meaning; new learning

AC C

builds on prior knowledge; learning is enhanced by social interaction; and learning develops through “authentic” tasks (see Cooperstein & Kocevar-Weidinger, 2004 p.141).

As a consequence, notions of deep and surface approaches to learning have become relatively ingrained in how learning is understood in the context of higher education (e.g., Haggis, 2009; Webb, 1997). Almost uniformly, a deep approach is labelled as desirable and

2

ACCEPTED MANUSCRIPT

a surface approach is characterised as undesirable or maladaptive. In an enlightening review of learning approaches research, Haggis (2009, p.377-8) suggests:

RI PT

In response to the repeated finding that large numbers of students appear not to be taking a deep approach, the question implied by the research seems to be ‘why do so many students take a surface approach to learning?’ Despite nearly 40 years of concentrated research activity, this question appears to remain still largely unanswered.

Arguably, the continuing interest in SAL research can be interpreted as an expression of the

SC

massification of higher education (HE) project in the western world. This has encouraged high rates of participation in HE, an increasingly diverse student population, with many

M AN U

students coming from overseas and from non-traditional backgrounds. A common solution to the challenge this creates is for academics to help meet students’ needs (e.g., Haggis, 2006) and to remedy deficiencies in study skills needed for academic success at university (e.g., Tait, Entwistle & McCune, 1998). Accordingly, a range of generic and subject-specific tools have been developed with the aim of assisting ‘at risk’ students. Accounting educators

TE D

have also been enthusiastic adopters of the approaches to learning literature, evidenced by recent literature reviews (Duff & McKinstry, 2007; Lucas & Mladenovic, 2006).

EP

What are frequently overlooked are students’ expectations of, and preconceptions about, the subject they will be studying. Within constructivist approaches, students construct their

AC C

own meaning and that new learning builds on prior knowledge, so it is imperative that accounting educators understand the implications of students’ expectations and preconceptions of the subject matter as they are central to their learning. This investigation is subject-specific and located in the discipline of accounting, which is a popular vocational subject in most anglicised western countries. The accounting profession is often the subject of powerful negative stereotypes (e.g., Bourgen, 1994; Dimnik & Felton, 2006; Friedman & Lyne, 2001; Mladenovic 2000). Phenomenographic work identifies that 3

ACCEPTED MANUSCRIPT

these negative stereotypes are associated with negative expectations of learning accounting (Lucas, 2000).

RI PT

Lucas (2000) identified two categories (or ‘worlds’) of students’ expectations of learning accounting that can be considered polar opposites. A ‘world of engagement’ is populated by intrinsically motivated students who view the process of learning accounting as personally relevant, meaningful and inherently useful. By contrast, a ‘world of detachment’

SC

is reality for extrinsically motivated students who see studying accounting as consisting of techniques to be learned, rather than something integral to their development.

These

M AN U

worlds have been conceptually and empirically linked to students’ approaches to learning (Lucas, 2001). A deep approach to learning accounting, like the world of engagement, is characterised by a desire to engage with concepts and ideas and a quest for personal meaning. By contrast, a world of detachment, like a surface approach is characterised as

TE D

learning by rote, memorisation, and academic anxiety (Lucas, 2001, Lucas & Meyer, 2005). 1 The purpose of this study is to consider how students’ expectations of learning influence

EP

their approaches to learning accounting and, in turn, their academic outcomes using the presage-process-product (3P) model (Biggs, 2003; Ramsden, 2003). It does so by sampling It also examines the

AC C

first-year undergraduate students of accounting in Australia.

relationship between these measures and a number of demographic variables (presage factors) namely: gender, English as a first or second language, subject majored in, and prior experience of study. Prior research has demonstrated that each of these demographic variables are influential in determining how students learn. The contribution of this paper lies in four areas. First, in contrast to earlier work in the accounting field that tends to focus on one or two elements of the 3Ps model, this study 4

ACCEPTED MANUSCRIPT

provides a comprehensive investigation of all three stages of the 3Ps model. Second, by the use of three assessment tools: an inventory assessing students’ expectations of learning accounting; a generic measure of approaches to learning adapted to an accounting context; Third, the use of a relatively

RI PT

and an overall mark for the unit achieved by the students.

large and diverse sample of international students enrolled in accounting classes at two institutions in Australia. The large sample allows testing for statistical significance; effect

replication.

SC

sizes are reported to allow an assessment of the magnitude of the effect and to facilitate Furthermore the diverse nature of the sample permits a comprehensive

M AN U

assessment of presage factors (expectations of learning accounting, gender, English as a first or second language, subject majored in, and prior experience of study) when prior work has tended to focus on only a few specific variables such as subject studied or gender. Fourth, the use of a cluster analytic statistical method enables the identification of distinct

TE D

subgroups of students with similar expectations of learning accounting profiles and highlights the interrelationships between the multiple presage factors students bring to the learning environment. This method is in contrast to the use of aggregate measures largely

EP

adopted by other scholars in the field and allows an assessment of variation within the

AC C

sample to assess the learning profiles and outcomes of the identified subgroups.

The next section of this paper describes the students’ approaches to learning literature and the accounting discipline that forms the contextual setting for the study. The research questions are outlined in section three. The methodology is described in section four and includes a description of the research instruments. The results are presented in section five. The final section presents a discussion of the findings and concludes the paper.

5

ACCEPTED MANUSCRIPT

1. Theoretical Framework

A framework widely used to understand SAL research is the presage-process-product (3Ps)

RI PT

model (Biggs, 2003; Duff & McKinstry, 2007; Dunkin & Biddle, 1974; Gibbs, 2010; Ramsden, 2003) – see Figure 1. Approaches to learning (the process of learning) are influenced by students’ perceptions of the requirements of the learning task. Perceptions reflect: (i) student perceptions of the learning context, i.e., teaching and learning activities, curriculum,

SC

assessment methods, classroom and institutional climate; and (ii) their general orientation

M AN U

to learning. The presage stages describing their orientations to learning, or antecedents of learning, are determined by their prior educational experience and the current learning context. The gist of the 3Ps model is that if educators can understand students’ perceptions of the assessment, the curriculum, the teaching and support they receive, then better learning outcomes can be achieved (Ramsden 2003; Trigwell & Prosser 1991).

TE D

-----------------Figure 1 here ------------------

Consequently, educators have sought to adapt the context of learning to alter students’

EP

perceptions, and in turn adopt more ‘desirable’ approaches to learning (for example:

AC C

English, Luckett & Mladenovic, 2004; Hall, Ramsay & Raven, 2004). Thus, the process stage has been the source of most research activity. More desirable approaches to learning are theoretically associated with better outcomes, or academic performance: the consequences of student learning (Birkett and Mladenovic, 2009).

Active learning approaches are often touted by educationalists as the means of improving the context of student learning.2 Continually-assessed projects, portfolios, cooperative

6

ACCEPTED MANUSCRIPT

learning, and appropriate essay questions provide students with the opportunity to demonstrate the quality and integrity of their learning. The provision of timely and constructive feedback provided on progress is key to quality learning outcomes (Baeten et The corollary to promoting a deep approach to learning is the idea that

RI PT

al., 2010).

educators unconsciously promote a surface approach to learning, which results in low quality learning outcomes. Some of the many factors reported in the literature to promote

SC

a surface approach to learning include: one-way transmission of information, assessment methods that focus on examining content that can be acquired through rote-learning (e.g.,

M AN U

multiple-choice tests or essays which assess discrete quantities), assessments that create anxiety (e.g., emphasis on testing, norm-referencing marks, and closed-book examinations), and learning contexts where students have little independent choice in what and how they

TE D

study (Biggs, 2003; Ramsden, 2003).

However, the presage stage of the 3Ps model has been the source of much less research. Earlier phenomenographic research in the domain of accounting identifies how students’

(Lucas, 2000).

EP

expectations of learning vary from a ‘world of detachment’ to a ‘world of engagement’ Arguably, the presage stage of the 3Ps model provides a clue to solving

AC C

Haggis’ (2009) question concerning the prevalence of a surface approach in student learning: the role of prior educational experience and what students expect their experience of student learning will be.

A detailed review of prior literature documenting students’ approaches to learning in accounting identifies just three empirical studies that address all three stage of the 3Ps model (Davidson, 2002; Duff, 2004; Ramburuth & Mladenovic, 2004). Each of these three 7

ACCEPTED MANUSCRIPT

investigations addresses the issue in different ways. For example, Davidson (2002) uses different types of assessments to evaluate students’ learning processes; Ramburuth & Mladenovic (2004) undertake a qualitative examination of students responses’ to a

RI PT

comprehension task which they grade according to a learning taxonomy, which in turn is related back to learning orientations (presage) and academic performance (product). Duff (2004) uses cluster analysis to relate students’ approaches to learning to academic

SC

performance, and relates presage variables such as prior academic attainment to the clusters, which are in turn related to academic performance. The results of these three

instrumentations.

M AN U

studies are difficult to compare as the research designs vary with the different However, from Duff (2004) and Davidson (2002) we can establish that

prior academic achievement is a significant variable in predicting future academic performance, in relation to approaches to learning self-reports, and that approaches to

TE D

learning scores are broadly related to the quality of learning outcomes (Davidson, 2002; Duff, 2004; Ramburuth Mladenovic, 2004).

EP

For the purposes of this investigation, a comprehensive number of key presage factors previously identified in the literature are explored, namely: gender, language, major, prior

AC C

study of accounting, and students’ expectations of learning accounting.

Gender

There are mixed findings regarding the relationships between gender and student approaches to learning. For example, a study of 265 students about to commence their first first-year of accounting study revealed gender differences on the strategic scale and some surface subscales, e.g., female students were more organised but had a higher fear of failure 8

ACCEPTED MANUSCRIPT

(Byrne & Willis, 2009). Similarly, Lucas & Meyer (2005) found significant differences between male and female students (N=1,211) studying introductory accounting: female students were more likely to worry and males were more likely to be exam-focussed. Other

RI PT

studies of accounting students have found that female students scored significantly higher than males on the surface approach (Paver & Gammie, 2005; Flood & Wilson, 2008).

SC

Duff’s review of the literature revealed unclear and inconclusive findings, and suggested that “gender differences may be instrument specific, with some approaches to learning

M AN U

inventories more sensitive to gender than others” (2004: 414). More recent studies show the same pattern of varied findings. Some researchers have found no differences due to gender in the learning approaches of accounting students (e.g. Ballantine, Duff & McCourtLarres, 2008; Byrne & Flood, 2008). The indecisive direction of gender differences in

TE D

learning approaches is also recorded in disciplines other than accounting (Baeten et al., 2010; Chamorro-Premuzic & Furnham, 2009; Edmunds & Richardson, 2009; Watkins &

EP

Regmi, 1990; Wilson, Smart, & Watson, 1996; Zeegers, 2001, 2004).

Language and culture

AC C

In terms of student learning outcomes, Drennan & Rohde (2002) found no association between English as a first or additional language and the performance of students in introductory accounting (N=303). Similarly Rankin, Silvester, Vallely & Wyatt (2003) found no significant relationship between English as a first or additional language and performance in introductory accounting (N=652). However, students with English as an additional language had significantly poorer results when studying an advanced accounting course (Drennan & Rohde, 2002). 9

ACCEPTED MANUSCRIPT

Most studies of the effects of culture on accounting students’ approaches to learning show differences between Asian and local (Australian) students. A large (N=542) study of first

RI PT

year accounting students found significant differences between Chinese and local students, with Chinese students having deeper learning motives (Donald & Jackling, 2007). Abeysekera (2008) found that a higher percentage of international, compared to local,

SC

students studying 3rd year accounting preferred interactive lectures and group work to traditional modes of learning. Cooper’s (2004) two-year study of accounting students

M AN U

(N=120) found that Chinese students had higher scores for both surface and deep approaches. The Chinese students with surface scores also performed well academically, in what Cooper calls “the enigma of the Chinese learner” (p.289), i.e., Chinese students use

Major

TE D

memorisation to deepen their understanding.

Findings regarding the relationships between a major and student approaches to learning

EP

are inconclusive. A one-year longitudinal study of students in a management accounting course (N=286) found that changes over time in student approaches to learning are not

AC C

associated with a major (Ballantine et al., 2008). Other studies show that students who major in accounting are more likely than non-majors to have positive views of accounting and of studying accounting (Geiger & Ogilby, 2000; Heiat & Brown, 2007; Lucas & Meyer, 2005). Lucas & Meyer (2005) found significant differences between accounting and business students (N=1,211) studying introductory accounting: accounting majors had higher enjoyment and saw the relevance of the subject, whereas non-majors were more likely to be worried, exam-focussed and disinterested. Students majoring in accounting are more 10

ACCEPTED MANUSCRIPT

likely to receive higher marks than non-majors (Al-Twaijry, 2010; Bonachi, Mutiu & Mustata, 2010; Mo & Waples, 2011; Rankin et al., 2003).

RI PT

Collectively, the problems with first-year accounting and non-accounting majors have been recognised by the Pathways Commission (2012), under the sponsorship of the American Accounting Association and the American Institute of Certified Public Accountants.

SC

Pathways (2012) is motivated by the need to create the most talented and diverse community of accountants at all levels, which it identifies can only be met by attracting as

M AN U

wide a pool as possible to accounting employment, including non-accounting majors. The work of Pathways has been echoed by the International Federation of Accountants’ (IFAC) International Accounting Education Standards Board’s (IAESB) (2013) recognition that the profession should benefit from greater diversity of intake. The first-course in accounting

TE D

then becomes pivotal in encouraging, or conversely, dissuading these students from further study of accounting and eventual accounting employment.

Wygal (2014) provides a

stimulating history of the first course in accounting and educators’ and policymakers’ efforts

EP

to craft its content and pedagogy away from a didactic elements of accounting principles

AC C

course towards a vibrant pathway to and of accounting.

Prior study of accounting

As Baeten et al. (2010, p.251) suggest: since students do not enter a learning environment like ‘empty vessels’ but bring their own experiences with them, these previous experiences may influence students’ approaches to learning.

Consequently, the interaction between approaches to learning and prior study of a discipline would seem a fertile area for research. However, relatively little effort has been 11

ACCEPTED MANUSCRIPT

expended in researching how prior knowledge of a discipline affects learning approaches. Beckwith (1991) reports that prior experience is negatively correlated with adoption of a surface approach. Research on approaches to learning has been less blind to the effect of

RI PT

prior learning of accounting at school on university learning of accounting, yet the results to date are inconclusive. Al-Twaijry (2010) and Byrne & Flood (2008) found that prior study of accounting has no impact on student performance in accounting. By contrast, Rankin et al.

SC

(2003) and Guney (2009) reported that prior study of accounting has a positive impact, while Koh & Koh (1999) identified prior study of accounting had a negative effect. These

M AN U

mixed findings have been explained by varying degrees of similarity between prior accounting education and university accounting programmes (Tan & Laswad, 2008; Koh, 2014).

TE D

In summary, a critical examination of the prior research on the antecedents of learning (presage factors), including: gender, language, major and prior study of accounting reveals inconsistent and inconclusive results across all factors. Given the study of these presage

EP

matters has, in many cases been in isolation, a more comprehensive examination of the

AC C

factors may provide insights into the inconsistent results.

Approaches to learning and academic outcomes Prior research indicates that deep and strategic approaches are broadly positively associated with success in studying accounting (Byrne, Flood & Willis, 1999, 2002; Davidson 2002; Duff 2004; Flood & Wilson, 2008; Paver & Gammie, 2005; Tan & Choo, 1991). Similarly surface approaches are negatively correlated with performance in accounting (Booth, Luckett and Mladenovic, 1999; Byrne et al., 2002; Duff 2004; Ramburuth & Mladenovic, 12

ACCEPTED MANUSCRIPT

2004; Tan & Choo, 1991). Other work (Duff, 1996) finds that studying differences across different cohorts by year and subject studied does identify statistically significant differences, yet these may be due in part to what statisticians term ‘family-wise’ error

RI PT

whereby multiple univariate testing identifies differences by chance alone.

However, different assessment instruments (e.g., examinations, essays, experiential

SC

assignments) will measure different learning outcomes. Therefore, as many accounting scholars have concluded (e.g., Davidson, 2002; English et al., 2004; Ramburuth &

M AN U

Mladenovic, 2004) the use of aggregate assessments, measuring different outcomes, is potentially problematic. Furthermore, the predictive ability of SAL inventories varies across instruments as the primary function of these measures is not to predict performance but as

TE D

research that aims to better inform educators about learning processes.

Evans (2014, p. in press) considering the accounting education and professional training interface, identifies an “over-emphasis on technical skills and preparation for the CPA

EP

examination as well as the production of narrowly-trained graduates”. In the light of this concern there have been many calls for education reform by policy makers, the accounting

AC C

profession, employers and educators over the past four decades. Beattie, Collins & McInnes (1997) claim the suggested reforms share a common fundamental feature, namely the need to develop learning environments that motivate students to move away from procedural or surface approach to learning towards a conceptual or deep approach to learning. This is particularly challenging in accounting departments which are characterised by large numbers of student enrolments, large class sizes and high student/staff ratios and follow

13

ACCEPTED MANUSCRIPT

the traditional role of education to enable students to develop technical competence in accounting (Evans, Burritt & Guthrie, 2010; Wygal, 2014).

RI PT

Birkett and Mladenovic (2009) argue that the study of accounting by its very nature, may require some memorisation and rote learning at all levels from introductory accounting to senior level accounting units. Empirical evidence to support this proposition was found by Jackling (2005b) who reported no change in students’ surface approach scores over three

SC

years but an increase in their deep approach scores. Jackling (2005b) argued that while students’ capacity to also engage in higher-order cognitive tasks as measured by the deep

M AN U

approach to learning student increases in later years, their surface scores may stay the same as they learn new technical material (see Birkett & Mladenovic, 2009 for a discussion of the linkages between learning the technical aspects of accounting and the surface and deep

TE D

approaches to learning).

Thus, considerable research effort has considered the nature of learning in the formative years of accounting education. However, relatively little work considers the relationship

EP

between students’ expectations of learning accounting and their approaches to learning, an

AC C

exception being Lucas & Meyer’s (2005) exploration using an earlier version of the Expectations of Learning Accounting (ELAcc) questionnaire and Meyer’s (2000) Reflections on Learning Inventory (RoLI). Undertaking an exploratory factor analysis (EFA) of six ELAcc scales and 13 RoLI scales, expectations of relevance and enjoyment were linked with transformative learning processes, for both males and females.

However, for the

accumulative expectations scales (worry, numbers, exam focus, and lack of interest) the findings were less clear, with few of these scales loading onto surface/accumulative RoLI

14

ACCEPTED MANUSCRIPT

measures, and a few loading onto a third ‘pathology’ factor. Some evidence of gender differences were apparent. The nature of EFA is, as its name suggests, exploratory and consequently neither tests hypotheses via goodness-of-fit statistics nor assesses how

RI PT

different groups of cases (students) assembled via one inventory (e.g., the ELAcc) map onto the learning profile produced by another measure (e.g., the RoLI). Consequently, Lucas &

SC

Meyer’s (2005) pioneering work invites further development and interpretation.

This paper attempts to fill the gap in research relating to the comprehensive examination of

M AN U

the three stages of the influential 3Ps model. It achieves this by measuring expectations of learning accounting using the ELAcc inventory (Lucas & Meyer, 2005; Duff & Mladenovic, 2014), which can be considered as an antecedent of learning (presage factors). To provide a more comprehensive examination of the many key antecedents of learning identified as

TE D

significant in the literature, our study examines the background variables of gender, major, English as a first or second language, and prior study of the discipline being studied. ELAcc scores (presage stage) are then related to a well-known measure of students’ approaches to

EP

learning, the RoLI (Meyer, 2004) (process stage), then their academic outcomes (product stage) as a consequence of learning. Cluster analysis of ELAcc scores is undertaken to

AC C

identify different ‘study orchestrations’ Meyer (2004, p.491) which fits with the intended purpose of the RoLI to present student learning profiles in a graphic form to enhance their metalearning capacity.

The cluster analytic method provides an intra-individual

examination of ELAcc and RoLI scores and relates them to performance. Three clusters are identified, each with a distinct profile.

Significance variation is identified in RoLI Deep,

Surface, and Pathology scores and academic performance.

15

ACCEPTED MANUSCRIPT

Prior research has identified that confirmatory statistical procedures such as factor analysis may disguise important variations in approaches to learning within groups of students and that cluster analysis provides a useful analytic tool (Meyer, 2000). Cluster analysis has also

RI PT

been used successfully within the literature examining the context of higher education student learning (Meyer, 2000; Meyer, Parsons & Dunne, 1990). In particular cluster analytic work has found statistically significant variation in learning pathologies associated

SC

with academic performance. These associations are clearly indicated in two studies of first-

undergraduate students.

M AN U

year economics (Cowie, Shanahan & Meyer, 1997) and engineering (Meyer & Sass, 1993)

The foregoing literature review then prompts three research questions:

1. Are there clusters of first-year students with distinct profiles based on their

TE D

expectations of learning accounting?

2. What are the demographic characteristics of these clusters in terms of gender, language, major, and prior study of accounting?

EP

3. Are there significant differences between the clusters in terms of their approaches to

AC C

learning and academic outcomes?

4. Method and preliminary results

Participants

A total of 1,553 undergraduate first-year students studying accounting at two universities in Australia participated in the research in 2004. There were 750 male students and 802 female students. For 578 students, English was their first language (EFL). For 973 students, English was their second language (ESL). 646 students were majoring in accounting (Major) 16

ACCEPTED MANUSCRIPT

and 890 students did not intend to major in accounting (Non-Major). Finally 558 students had studied accounting at school, while 991 students had undertaken no prior study of

RI PT

accounting. 3

Both universities have entry requirements in the highest quartile.

They are AACSB

accredited4 and are also members of the Group of Eight (Go8) which is “a coalition of

SC

leading Australian universities intensive in research and comprehensive in general and professional education”. In both universities the student cohort consists of high achieving

M AN U

local and international students (having achieved entry requirements in the highest quartile). Not all students accepted on these courses will have prior study of accounting and no prior study of accounting is assumed.

TE D

The degree programs and introductory core accounting subjects at Institution 1 and Institution 2 are similar in many ways. The topic areas included in the curriculum in both institutions are similar. The main focus was on developing students’ foundational technical

EP

competence in accounting with the study of topics including: the accounting equation; debits and credits; the accounting cycle; internal control and cash; receivables, bad and

AC C

doubtful debts; non-current assets (cost, depreciation, disposal); payables; and equity. While the focus was on supporting students to develop a procedural and technical understanding of accounting, there was discussion of the social sphere and the competing interests of stakeholders in the business environment to provide students with the boarder context of accounting. The students experienced similar face-to-face teaching contact hours. The assessment regimes at the two institutions are similar with a large proportion of the assessment being examination-based (Institution 1, 75%; Institution 2, 70%). 17

ACCEPTED MANUSCRIPT

Partitioning the sample by institution yields similar findings in terms of: the magnitude and direction of the effects; and interpretation of the cluster analytic results.

RI PT

Ethical clearance was gained from both institutions to administer the ELAcc and RoLI to students of introductory accounting. The inventories were administered by one of the researchers within class-time. Students were instructed to ask questions if necessary,

SC

assured of the confidentiality of their answers, and that no right or wrong answers existed.

Expectations of Learning Accounting

M AN U

Measures

The Expectations of Learning Accounting (ELAcc) inventory is a 43-item inventory developed for use with accounting undergraduate students (Lucas & Meyer, 2005; Duff & Mladenovic, 2014). Nine expectations of learning accounting, along with their conceptual underpinnings

TE D

and descriptions are shown in Table 1. The development and validation of the ELAcc is reported in Duff & Mladenovic (2014).5

EP

ELAcc has nine subscales labelled: Enjoyment (ENJ); Achieving (ACH); Reality/meaning behind

accounting (REAL); Social and economic importance of accounting (SOC); Questioning (QN);

AC C

Exam focus (EXM); Lack of personal interest (LINT); Worry (WOR); Numbers (NUM). Participants indicated the extent to which they agreed with each item on a five-point scale from 0 (strongly agree) to 4 (strongly disagree). Confirmatory factor analysis on the ELAcc demonstrated adequate fit indices [χ2 (824) = 2540.288; p < .001; CFI = .924; NNFI = .913; RMSEA = .031 (90% confidence intervals = .030 - .033)].6,7 ----------------Table 1 here ----------------18

ACCEPTED MANUSCRIPT

Reflections of Learning Inventory The Reflections of Learning Inventory (RoLI) is a 90-item inventory that aims to identify

RI PT

variation in student learning within a subject or topic context (Meyer, 2004 p.491). It attempts to make students aware of their own learning by representing responses to the instrument by creating a personal learning profile.

SC

The RoLI is not conceived of as a fixed inventory, but as an instrument with a variety of

M AN U

subscales, that can be used in different contexts, e.g., different students or different subject areas (Meyer, 2004). In accounting, for example, Lucas & Meyer (2005) reported that ‘Thinking Independently’ did not operate as expected and is removed for the present investigation.

The RoLI version 10a used for the present study has three scales of

Deep/Transformative, Surface/Accumulative, and Pathology. 16 subscales measure these three scales.8

TE D

For the purposes of analysis each of these three scales is addressed

separately. Table 2 provides a description of each subscale relating to the three scales, along with their conceptual underpinnings. Students indicated the extent to which they

EP

agreed with each item on a five-point scale from 0 (strongly agree) to 4 (strongly disagree).

AC C

Confirmatory factor analysis using Amos v17 was used to assess the internal validity of scores produced by the instrument.9 Each scale was subjected to an item attrition exercise whereby pattern matrix coefficients (FPCs) were examined. When the FPC was less than .4, the item was removed and the factor structure coefficients examined to see if the item loaded onto another scale. The CFA item attrition exercise suggested the removal of 7 items from the inventory along with a subscale (DRP) which was expected to load onto the pathology scale. 19

ACCEPTED MANUSCRIPT

The Deep/Transformative scale consists of eight subscales, labelled: Knowledge Objects (KOB); Seeing Things Differently (SDI); Relating Ideas (RID); Memorise After Understanding

RI PT

(MAU); Rereading a Text; Memorise With Understanding (MWU) ; Knowing About Learning (KAL); and Repetition Aids Understanding (RAU). Each subscale is measured by five items. CFA on the Deep/Transformative scale and its eight associated subscales showed acceptable

SC

fit [χ2(586) = 1782.11; p < .001; CFI = .904; NNFI = .891; RMSEA = .031 (90% confidence

M AN U

intervals, .029 - .033)]. 10

The Surface/Accumulative scale consists of four subscales. These are labelled: Learning is Fact Based (FAC); Knowledge is Discrete and Factual; Learning is Experienced as a Duty (DUTY); and Memorising Before Understanding (MBU). CFA on the Surface/Accumulative scale and its four constituent subscales also fit the data [χ2(148) = 470.25; p < .001; CFI =

TE D

.963; NNFI = .952; RMSEA = .032 (90% confidence intervals, .029 - .035)].

EP

The Pathology scale consists of four subscales, labelled: Detail Related Pathology (DRP); Fragmentation (FRA); Memorising as Rehearsal (MAR); Learning by Example (LBE). The item

AC C

attrition exercise removed one subscale from the Pathology scale (Detailed Related Pathology) which produced a FPC value below .4. The Pathology scale and its final three subscales demonstrated adequate fit [χ2(51) = 148.47; p < .001; CFI = .970; NNFI = .953; RMSEA = .030 (90% confidence intervals, .025 - .036)]. The final model consists of three scales calibrated by 15 subscales. -----------------Table 2 here -----------------20

ACCEPTED MANUSCRIPT

4.

Results

Table 3 reports the means and standard deviations for the main variables used in the study.

RI PT

In general, students had high deep/transformative expectations of learning accounting (means, 2.51 to 3.04, on a scale from 0 to 4). Similarly, students had moderate low surface/accumulative expectations of learning accounting (means, 1.20 to 2.36, on a scale

SC

from 0 to 4). In terms of learning processes, they reported generally high scores on the Deep/Transformative measures, with the exception of moderate scores on KOB and low Both Surface/Accumulative and Pathology RoLI subscales scores were

M AN U

scores of SDI.

uniformly low (means < 2.16, on a scale from 0 to 4).

----------------Table 3 here -----------------

positively

correlated

TE D

As would be expected, Deep/Transformative ELAcc subscales were moderately to highly with

Deep/Transformative

RoLI

subscales.

Similarly,

Surface/Accumulative ELAcc subscales moderately to highly positively correlated with

EP

Surface/Accumulative RoLI subscales.

AC C

To explore distinct expectations of learning accounting within the students sampled, the nine subscales were subjected to a k-means cluster analysis using the log-likelihood distance measure and Schwarz’s Bayesian clustering criterion. As not all the nine subscales are measured using the same scale, standardization of the scores was undertaken.

The number of clusters was determined by examining: first, the within cluster variation plots, to determine the distance between the potential clusters across each measure case; 21

ACCEPTED MANUSCRIPT

and second, a Bonferoni-adjusted comparison of means between cluster scores on each measure. It was decided that a three-cluster solution was most appropriate for this data.

RI PT

A one-way MANOVA was conducted using the nine subscales as dependent variables and the clusters as the fixed factor. The results show significant differences between the three clusters on the dependent measures [Wilks’ λ = .216, F(18, 3056) = 396.89, p < .001, η2 =

SC

.54].11 Table 4 contains the standardised means and standard deviations on the nine subscales for the three clusters, in addition to the F-tests and partial effect sizes. The large

M AN U

F-ratio, small observed statistical significance level, and large effect sizes (Cohen, 1977) associated with each of the nine ELAcc subscales suggests there is high variability between the two clusters for each of these variables and allows us to conclude that the clusters are satisfactory descriptors of both types of learners.12 An examination of the individual F-ratios

TE D

identifies that Reality/Meaning Behind Accounting, Exam Focus, and Enjoyment contribute the most to the variability between the four clusters (all statistically significant at p <.001

----------------Table 4 here -----------------

AC C

EP

and η2 >.37).

Figure 2 displays the three distinct expectations of learning accounting profiles identified using the cluster analysis. The demographic membership of each cluster is reported in Table 5.13 Cluster 1 has an ‘average’ profile, with scores that lie in general terms between clusters 2 and 3, but with relatively high scores on the Numbers subscale. With 791 students, cluster 1 is the largest cluster, consisting of predominantly females (55%) and EFL (40%) students.

22

ACCEPTED MANUSCRIPT

Cluster 2 is characterised by positive expectations or a ‘world of engagement’ (Lucas, 2000) with high scores on the five deep/transformative subscales and low scores on the four surface/accumulative measures. Students in this cluster reported that they had a strong

RI PT

desire to excel in their academic achievement (ACH); expected to enjoy their accounting studies (ENJ); wanted to see the reality or meaning behind accounting numbers in a business context (REAL); expected that they would question the basis on which accounting

SC

techniques are founded (QN); and would be interested in exploring the social and economic importance of accounting (SOC). In contrast, students in the ‘world of engagement’ did not

M AN U

expect to focus on simply ‘passing’ the exam (EXM); did not report high scores for a lack of personal interest in accounting (LINT); but did expect that it would have some focus on numbers (NUM); and were the least worried about their learning in accounting (WOR). This cluster, with 274 students, comprises predominantly students majoring in accounting (57%),

gender and EFL/ESL.

TE D

with nearly half having studied accounting at school, and an even distribution of both

EP

Cluster 3 is characterised by negative expectations or a ‘world of detachment’ (Lucas, 2000), with low scores on the deep/transformative measures, and high scores on the

AC C

surface/accumulative subscales. Students in this cluster reported that: they had a lower desire to excel in their academic achievement (ACH); did not expect to enjoy their accounting studies (ENJ); were less interested in exploring the reality or meaning behind accounting numbers in a business context (REAL); did not expect to question the basis on which accounting techniques are founded (QN); and did not anticipate that they would be interested in exploring the social and economic importance of accounting (SOC). In contrast, students in the ‘world of detachment’ did expect to do only enough work to simply ‘pass’ 23

ACCEPTED MANUSCRIPT

the exam (EXM); reported a lack of personal interest in accounting expecting it to be dull and boring (LINT); with a focus on numbers (NUM); and anticipated that they would be worried about their learning in accounting (WOR). The 488 members of cluster 3 are

individuals; gender is evenly distributed.

RI PT

dominated by students not majoring in accounting (72%) and an over-representation of ESL

SC

Notably, as an examination of Figure 2 highlights, is that exam (EXM) and worry (WOR) scores are relatively low for all three clusters. This is likely to be an expression of the

M AN U

relatively high abilities of our sample. Thus, students with positive prior experiences of education, regardless of their exposure to accounting, are likely to be more self-efficacious and less concerned with the prospect of passing exams or other academic neurotic behaviours.

TE D

-----------------Figure 2 here --------------------------------Table 5 here ----------------

EP

The analysis then examined the related responses of the three clusters using criteria

AC C

variables, i.e., the RoLI subscales that were not included in the cluster analysis. This is undertaken in three stages, using the subscales of the three scales. First, a MANOVA was conducted with cluster group as the fixed factor and the eight deep/transformative RoLI subscales as the dependent variables. The multivariate statistics identified statistically significant differences among the three clusters [Wilks’ λ = .871, F (16, 3060) = 13.58, p < .001, η2 = .07]. To control for type 1 error, the alpha level was reduced to

24

ACCEPTED MANUSCRIPT

.0167 (i.e., .05/3). The between subjects effects revealed statistically significant differences between the three clusters on all eight dependent variables (Panel A, Table 6).

RI PT

Pairwise comparisons of the three clusters were undertaken applying post-hoc tests, Tukey’s HSD, see Table 6. As expected, clusters 2 and 3 had significantly higher scores than cluster 1 on seven of the eight deep/transformative subscales, and cluster 1 a significantly higher score than 3 on the one remaining measure (RAU). However, a small effect size is

SC

indicated by the partial η2 statistics. Figure 3 shows the profiles for the three clusters on the

M AN U

Deep/Accumulative subscales. ----------Figure 3 -----------

Second, the procedure was repeated for the four surface/accumulative subscales. MANOVA

TE D

with cluster as the fixed factor and the four surface/accumulative measures as dependent variables revealed statistically significant differences among the three clusters [Wilks’ λ = .921, F (8, 3080) = 16.23, p < .001, η2 = .04]. The univariate cases were consequently

EP

examined and the between subjects effects were statistically significant for all four dependent variables – see Panel B of Table 6. Pairwise comparisons of the three clusters

AC C

were undertaken applying post-hoc tests, Tukey’s HSD (Table 6). As expected, cluster 1 had significantly higher scores than cluster 3 on all four of the surface/accumulative subscales. However, in each instance the partial η2 statistics indicate a small effect size. The profiles for each of the three clusters on the Surface/Accumulative subscales are shown in Figure 4. -----------------Figure 4 here ------------------

25

ACCEPTED MANUSCRIPT

The same procedure was applied to the three pathology measures. Undertaking a MANOVA with cluster as the fixed factor and the three pathology subscales as the dependent variables revealed statistically significant variation among the three clusters [Wilks’ λ = .848,

RI PT

F (6, 3056) = 43.66, p < .001, η2 = .08]. The between subjects effects revealed statistically significant differences in FRA and MAR. Applying Tukey’s HSD as a post-hoc test, clusters 1 and 2 produced statistically significant (p < .0167) higher scores than cluster 3. Similarly

SC

cluster 1 had a statistically significant higher score than cluster 2. The partial η2 of .10 indicates a medium effect size. Figure 5 shows the profiles for the three clusters on the

M AN U

Pathology subscales.

-----------------Figure 5 here -----------------

TE D

Finally, the concurrent validity of the three clusters was examined by a one-way ANOVA with the overall assessed mark for the course as the dependent variable and cluster as the fixed factor. Statistically significant differences were identified [F(2) = 14.53, p < .001]

EP

where Tukey’s HSD post-hoc tests identified clusters 1 and 2 produced a higher mark than cluster 3 (p <.001). The differences between the three clusters were as anticipated, with

AC C

the clusters associated with positive expectations of learning accounting, producing higher levels of performance and demonstrating satisfactory concurrent validity. ----------------Table 6 here -----------------

Discussion

26

ACCEPTED MANUSCRIPT

This comprehensive study provides a unique contribution to the literature as it is concerned with all three stages of the 3P model as outlined in the three research questions:

expectations of learning accounting (presage)?

RI PT

1. Are there clusters of first-year students with distinct profiles based on their

2. What are the demographic characteristics of these clusters in terms of gender, language, major, and prior study of accounting (presage)?

SC

3. Are there significant differences between the clusters in terms of their approaches to

M AN U

learning (process) and academic outcomes (product)?

The main aim of this study was to look at students’ intra-individual expectations of learning accounting as well as a comprehensive number of other key presage factors and to examine their links to reflections of learning profiles and outcomes in performance. This was

TE D

achieved using cluster analysis. Three expectations of learning profiles were established. The ‘high expectations’ cluster with very positive views of learning accounting (Cluster 2) and the cluster with ‘moderate expectations’ (Cluster 1) had the most adaptive profiles. A

EP

student from the ‘high expectations’ cluster (Cluster 2) is equally likely to be a male or a female, is more likely to be majoring in accounting, be more likely to be an English as a first

AC C

language student relative to the rest of the cohort, and more likely to have studied accounting at school. The ‘moderate expectations’ cluster (Cluster 1) is similar in most respects to cluster 2, expect that they are marginally more likely to be female than cluster 1, but much less likely to have studied accounting at school. Cluster 3 has a relatively maladaptive profile, with moderate scores on the nine ELAcc variables, has the highest proportion of ESL students, marginally fewer female students and the smallest proportion of students majoring in accounting. 27

ACCEPTED MANUSCRIPT

These findings show that looking at the multitude of presage variables in isolation when exploring the relationships with approaches to learning may be too simplistic. If we look at the relationship between gender and approaches in isolation, we see that females are

RI PT

almost equally likely to fall into cluster 2 (adaptive approaches) as they are cluster 3 (maladaptive approaches). This is consistent with the inconclusive findings in the literature on gender. However, looking at the key presage variables together, provides some insights

SC

into the mixed findings. When we look at expectations, gender, language, major, and prior studies, we can start to see patterns emerging around the interplay between the key

M AN U

presage variables. For example, females are more likely to report adaptive approaches (cluster 2) if they have positive expectations of learning accounting and are majoring in accounting.

The contrasts identified across the three clusters formed on expectations of learning

TE D

accounting are less apparent when the profiles are examined in relation to the Deep/Accumulative scales of the RoLI.

As might be expected Cluster 2 (world of

engagement) is markedly more positive than cluster 3 (world of detachment), yet the When the Surface/Accumulative scales are examined,

EP

differences are attenuated.

differences between the ‘low’ (cluster 3) and ‘moderate’ (cluster 1) expectations cluster are

AC C

negligible for three scales (FAC, KDF, and MBU). A similar story operates for two (of the three) RoLI Pathology scales for all three clusters, where no differentiation is found for LBE and MAR.

Academic performance was not strongly positively correlated with any of the expectations of, or reflections on, learning variables in this study. The highest positive correlation (r = .11) was with ACH, a Study Process Questionnaire scale (Biggs, 1987a,b). This finding is

28

ACCEPTED MANUSCRIPT

consistent with previous research in accounting (Booth et al., 1999).

The other

Deep/Transformative variables of ELAcc and RoLI produced correlation coefficients of small magnitude (r < .06) and in one instance a negative direction (KOB, r = -.07). However, those

RI PT

variables identified as negative expectations or negative processes of learning produced small to medium-sized correlation coefficients, with ELAcc’s EXM and WOR (both r > - .15), and RoLI’s FAC, KDF, MBU, and FRA (all r > - .15). So while it may not be clear which

SC

expectations or learning processes produce good academic performance in first-year accounting, it is clear many Surface/Accumulative processes are associated with poor

M AN U

outcomes. This finding is consistent with Booth et al. (1999), where it was reported that higher surface approach scores were associated with less successful academic outcomes, but no association was found for deep approach scores in a second-year management accounting course.

TE D

An intriguing finding of the current study is that while clear differences exist in expectations of learning accounting, and these to some limited degree predict academic performance and Deep/Accumulative learning processes, the identified clusters have much less bearing

EP

on Surface/Accumulative or problematic learning processes, which the study identifies are indicative of poor performance.

That is, less variation is found between the identified

AC C

clusters and surface/accumulative RoLI scores, in contrast to deep/transformative RoLI scores where more marked variation is identified. Deep/Transformative expectations and learning processes have little impact on academic performance, yet Surface/Accumulative ELAcc and RoLI scales are associated with poor performance. The nature of first-year accounting, as an introduction to the subject, tends to favour the mechanical, over the conceptual, and the ‘how to do’ rather than ‘why’ (see Wygal, 2014 for

29

ACCEPTED MANUSCRIPT

a discussion of the first course in accounting, along with Pathways, 2012 and IAESB, 2013). As outlined earlier in the paper, the first-year accounting courses at both institutions in this study were like many introductory accounting courses, concerned with developing the

RI PT

fundamentals of accounting that are reflected in the assessment processes. Consequently positive expectations and reflections on learning are less likely to affect performance concerned with more procedural questions.

However, as consistently found in the

SC

literature, academic anxiety, and low-level learning processes focusing on memorisation

to result in low quality outcomes.

M AN U

without understanding, viewing knowledge as discrete and made up of facts, are more likely

A limitation of this investigation is the use of a single aggregate mark in assessing the product of learning. There are no simple solutions to overcoming this problem. Some prior research has attempted to decompose performance into discrete learning tasks with distinct

TE D

learning outcomes that reward the use of certain learning strategies (e.g., Davidson, 2002). However, decisions regarding the coding of different types of assessment and individual questions are ambiguous and highly subjective and are at odds with the quantitative, large-

EP

scale, dual-institution, and (relatively) objective methodological approach adopted by the present investigation. However, we see considerable potential for our work to be extended

AC C

by others using smaller samples and more qualitative assessment of the quality of learning outcomes rather than examination performance (e.g., Dahlgren, 1984; Jackling, 2005a).

Conclusions

30

ACCEPTED MANUSCRIPT

In conclusion, the cluster analytic results provide a clearer picture of undergraduate student learning in accounting as a vocational discipline. Students come to learning at university with some expectations of what learning that subject ‘is about’. It could be a ‘world of

RI PT

engagement’ with high expectations (cluster 2) or it could be a ‘world of detachment’ (cluster 3) where it is a subject to be feared, avoided and simply passed. When studying accounting is through choice, expectations are positive.

SC

Expectations also have some predictive validity with learning processes, with Deep/Transformative expectations positively correlated with their corresponding RoLI A similar, although slightly less marked, effect is noted between the

M AN U

scales.

Surface/Accumulative expectations and learning processes scales.

In particular less

variation is noted in the ELAcc surface/accumulative scale NUM (accounting is about numbers) or the RoLI study pathology scales LBE (learning by example) and MAR

accounting learners.

TE D

(memorising after rehearsal) each of which are strategies practiced by all forms of

EP

Although the findings of this investigation are potentially valuable, caution should be exercised about its generalisability. First, the study uses relatively high ability students from

AC C

two institutions in Australia, where the majority of students are non-native speakers of English. Lower ability students from different cultural backgrounds may differ, with less favourable expectations and learning processes. For example, ‘exam focus’ recorded the lowest scores across all three clusters in our study, indicating that whatever their perceptions of accounting, our high ability students at least expected to pass and were not overly affected by academic neuroticism. In addition, the study is cross-sectional and it

31

ACCEPTED MANUSCRIPT

would be interesting to undertake some form of longitudinal study to understand how expectations and beliefs about learning change over time.

RI PT

The accounting curriculum is highly prescribed by professional bodies - see Evans (2014) for a discussion on the interface between accounting education and professional training and Duff & Marriott (2012) for a discussion of UK academics’ acceptance of curricula largely

the United States of America.

SC

determined by the profession and Zeff’s (1989a, 1989b) earlier warnings of this trend from As described earlier in this paper, introductory level

M AN U

accounting is characterised by a significant technical content resulting in relatively didactic means of instruction, in part because of the large numbers of students enrolled in the subject at university. How would for example, expectations of learning finance, as a subject allied to accounting, influence student learning approaches and outcomes?

TE D

Our study then sheds more light on Haggis’ (2009) vexing question of why students utilise surface learning approaches. For those whom accounting is a ‘world of detachment’, with negative views of what accounting is likely to be about, are more likely to adopt ineffective

EP

learning strategies. In this study, they are also less likely to be native speakers of English or to have actively chosen accounting as a course of study. ELAcc, together with RoLI, then

AC C

provides a means of making students more aware of their preconceptions of the subject they are studying and making them more aware of their learning processes. Such an approach should allow them to take greater control of their learning, make them more aware of problems they encounter and seek help, and enhance their self-efficacy by allowing them to help their peers.

32

ACCEPTED MANUSCRIPT

The integration of peer-assisted study support, a technique widely adopted by other disciplines (e.g., Fuchs, Fuchs, Phillips, Hamlett & Karns, 1995; Fuchs, Fuchs, Mathes & Simmons, 1997; van der Meer & Scott, 2009), but which has made slower traction in

RI PT

accounting (exceptions being Adler & Milne, 1997; Crawford, Helliar, Monk & Stevenson, 2011; Fox, Stevenson, Connelly, Duff & Dunlop, 2010; and Jackling & McDowall, 2008), arguably because of high staff to student ratios which limit the degree of personal

2011), especially in large first-year classes.

SC

interaction in accounting education (Duff & Marriott, 2012; Parker, Guthrie & Linacre,

M AN U

Furthermore, given the results reveal that students with negative expectations of studying accounting are more likely to adopt ineffective learning strategies and have poorer learning outcomes, the challenge for accounting educators is to support these students to develop positive expectations of learning accounting. Many accounting educators have already

TE D

considered change to the way in which the first course is delivered with a move away from a didactic, accounting principles approach to a more blended learning method. Examples include the use of: case studies (Adams, Lea & Harston, 1999; Dikolli & Sedatole, 2003);

EP

skills-based approaches (Ainsworth, 1994; Hardy & Deppe, 1995; Saudagaran, 1996; Stoner & Milner, 2010); user-centred approaches with the focus on decision-making (Norgaard &

AC C

Hussein, 1990). Mladenovic (2000, p.148) found that comprehensive educational interventions are more likely to change students’ negative preconceptions: …an alignment of teaching methods, curriculum and assessments, and directly challenging students’ perceptions, appears to be more effective in changing students’ negative perceptions than changing teaching methods as the main intervention, as demonstrated in prior research (Friedlan, 1995 and Caldwell et al., 1996).

Finally, this study highlights the benefits of exploring all aspects of the Presage-ProcessProduct Model, and in particular the importance of including a comprehensive range of 33

ACCEPTED MANUSCRIPT

presage variables in the research. Previous studies that explore only one or few presage variables report mixed findings that make it difficult for educators to formulate effective changes to the learning environment that support student learning. Exploring the key

RI PT

presage variables identified on the literature of gender, expectations of learning, major, language, and prior study together provides insights into the relationships among these variables and their impact on approaches to learning and learning outcomes. This

and teaching practices to support student learning.

SC

knowledge in turn provides educators with insights into the kinds of changes to curriculum

M AN U

The manifest differences observed between students’ majoring in accounting and for those to whom the subject is an imposition suggests the promotion of differential teaching strategies. Teaching accounting students in different groups aside from non-accounting students was the traditional approach in Australia and the UK before the need for large class

TE D

sizes became apparent14. Our findings then provide support for the Pathways Commission (2012) and IAESB’s (2013) recommendations of encouraging different pathways to the study and practice of accounting. Positive and straightforward changes accounting programme

EP

designers could consider would include the creation of a short one- or two-day ‘camp’ prior to the commencement of studies whereby non-specialist could receive an intensive

AC C

introduction into the accounting cycle. Extensive prior literature documents the superior performance of accounting majors over non-majors (Al-Twaijry, 2010; Bonachi et al., 2010; Mo & Waples, 2011; Rankin et al., 2003). In addition separating non-specialists from accounting majors would be an effective and low-cost approach. Students studying an accounting degree, with largely positive expectations of the experience of learning accounting, may benefit from a technically-focused syllabus that aims

34

ACCEPTED MANUSCRIPT

to provide the building blocks for further study in accounting. Non-accounting students may benefit from a course that develops an understanding of the role of accounting in business and society, of accounting as a language, and its omnipresent role in human life. These

RI PT

different programmes suggest differential teaching and assessment strategies: one focused on techniques and principles; another focused on ideas and the embedding of accounting in

AC C

EP

TE D

M AN U

SC

social discourse.

35

ACCEPTED MANUSCRIPT

REFERENCES

AC C

EP

TE D

M AN U

SC

RI PT

Abeysekera, I. (2008). Preferred learning methods: a comparison between international and domestic accounting students. Accounting Education: an international journal, 17, 187-198. Adams, S.J., Lea, R.B., & Harston, M.E. (1999). Implementation of a serial case pedagogy in the introductory managerial accounting course. Issues in Accounting Education, 14 (4), 641-656. Adler, R.W. & Milne, M.J. (1997). Improving the quality of accounting students' learning through action-oriented learning tasks. Accounting Education: an international journal, 6, 191-215. Ainsworth, P. (1994). Restructuring introductory accounting courses: The Kansas State University experience. Journal of Accounting Education, 12 (4), 305-323. Al-Twaijry, A.A. (2010). Student academic performance in undergraduate managerialaccounting courses. Journal of Education for Business, 85 (6), 311-322. Atrill, P. & McLaney, E. (2012). Accounting for non-specialists. London: Pearson. Ballantine, J.A., Duff, A., & McCourt-Larres, P. (2008). Accounting and business students’ approaches to learning: a longitudinal study. Journal of Accounting Education, 26, 188201. Baeten, M., Kyndt, E., Struyven, K., & Dochy, F. (2010). Using student-centred learning environments to stimulate deep approaches to learning: Factors encouraging or discouraging their effectiveness, Educational Research Review, 5, 243–260. Beattie V., Collins B. & McInnes B. (1997). Deep and surface learning: a simple or simplistic dichotomy? Accounting Education: an international journal 6(1) pp.1-12. Beckwith, J. (1991). Approaches to learning, their context and relationship to assessment performance. Higher Education, 22, 17–30. Biggs J. (1987a). Student Approaches to Learning and Studying. Australian Council for Educational Research, Hawthorn, Victoria. Biggs J. (1987b). Study Process Questionnaire Manual. Australian Council for Educational Research, Hawthorn, Victoria. Biggs, J.B. (2003). Teaching for Quality Learning at University. Maidenhead: SRHE and Open University Press. Birkett, W. & Mladenovic, R. (2009). The Study Process Questionnaire: Theoretical and Empirical Issues for Accounting Education Research in Practices, Profession and Pedagogy in Accounting - Essays in Honour of Bill Birkett, ed. J Baxter and C Poullaos , Sydney University Press , Sydney , Australia, pp. 423-50. Bonachi, C.G. Mutiu, & Mustata, R.V. (2010). Influential factors of accounting students’ academic performance: a Romanian case study. Accounting and Managemnet Information Systems, 9 (4), 558-580. Booth, P. Luckett, P. & Mladenovic, R. (1999). The quality of learning in accounting education: the impact of approaches to learning on academic performance, Accounting Education: an international journal 8 (4), 277-300. Bourgen, P. (1994). Joking apart: the serious side to the accounting stereotype. Accounting, Organizations and Society, 19, 319-335. Browne, M.W. & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen & J. S.Long (Eds.), Testing structural equation models (pp. 136-162). Newbury Park, CA: Sage.

36

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Byrne, M., Flood, B., & Willis, P. (1999). Approaches to learning of Irish accounting students, Irish Accounting Review 6, 1-29. Byrne, M., Flood, B., & Willis, P. (2002). The relationship between learning approaches and learning outcomes: A study of Irish accounting students, Accounting Education: an international journal 11 (1), 2002, 27-42. Byrne, M. & Flood, B. (2008). Examining the relationships among background variables and academic performance of first year accounting students at an Irish University. Journal of Accounting Education, 26, 202-212. Byrne, M. & Willis, P. (2009). An exploration of tertiary accounting students’ prior approaches to learning accounting. International Journal of Management Education, 7, 35-46. Caldwell, M.B., Weishar, J., & Glezen, G.W. (1996) The effect of cooperative learning on student perceptions of accounting in the principles courses, Journal of Accounting Education, 14, 17-36. Chamorro-Premuzic, T., & Furnham, A. (2009). Mainly openness: The relationship between the Big Five personality traits and learning approaches. Learning and Individual Differences, 19, 524–529. Cohen, J. (1977). Statistical power for the behavioral sciences. San Diego: CA: Academic Press. Cooper, B.J. (2004). The enigma of the Chinese learner, Accounting Education: an international journal, 13, 289-310. Cooperstein, S.E. & Kocevar-Weidinger (2004). Beyond active learning: a constructivist approach to learning. Reference Services Review, 32, 141-148 Cowie, J., Shanahan, M., & Meyer, J.H.F. (1997). Measuring learning processes in first year economics: preliminary results, Research and Development in Higher Education, 20, 209-230. Crawford, L., Helliar, C., Monk, E., & Stevenson, L. (2011). SCAM: Design of a learning and teaching resource. Accounting Forum 35, 61-72). Elsevier. Dahlgren, L. O. (1984) Outcomes of learning, in: F. Marton et al. (Eds) The Experience of Learning, pp. 19–35. Edinburgh: Scottish Academic Press. Davidson, R.A. (2002). Relationship of study approach and exam performance, Journal of Accounting Education 20 (1), 29-44. Dikolli, S. & Sedatole, K. (2003). Schlotzky’s gourmet franchise store. Accounting Education: An international journal, 12 (4), 427-435.; Dimnik, T., & Felton, S. (2006). Accountant stereotypes in movies distributed in North America in the twentieth century. Accounting, Organizations and Society, 31, 129-155. Donald, J. & Jackling, B. (2007). Approaches to learning accounting: a cross-cultural study. Asian Review of Accounting, 15, 100-121. Drennan, L.G., & Rohde, F.H. (2002). Determinants of performance in advanced undergraduate management accounting: an empirical investigation, Accounting and Finance, 42, 27-40. Duff, A. (1996). The impact of learning strategies on academic performance in an accounting undergraduate course. In Gibbs, G. (ed.) Improving student learning: using research to improve student learning. Oxford: Oxford Centre for Staff Development, 50-62. Duff, A. (1997). Validating the learning styles questionnaire and inventory of learning process in accounting: a research note. Accounting Education: an international journal 6 (3), 263-272. 37

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Duff, A. (2004). Understanding academic performance and progression of first-year accounting and business economics undergraduates: the role of approaches to learning and prior academic achievement. Accounting Education: an international journal, 13, 409-430. Duff, A., & Marriott, N. (2012). Teaching and research: Partners or competitors? Edinburgh: Institute of Chartered Accountants of Scotland. Duff, A., & McKinstry, S. (2007). Students’ approaches to learning: a literature review, Issues in Accounting Education, 22, 183-214. Duff, A. & Mladenovic, R. (2014). Developing and validation of a scale measuring accounting students’ expectations of learning. Unpublished working paper. Dunkin, M.J. & Biddle, B. J. (1974). The study of teaching. Austin, TX: Holt, Rinehart & Winston. Dyson, J.R. (2010). Accounting for non-accounting students. London: FT/Prentice Hall. Edmunds, R., & Richardson, J. (2009). Conceptions of learning, approaches to studying and personal development in UK higher education. British Journal of Educational Psychology, 79, 295–309. English, L., Luckett, P. & Mladenovic, R. (2004). Encouraging a deep approach to learning through curriculum design, Accounting Education: an international journal 13 (4), 461– 488. Entwistle, N.J., Tait, H., & McCune, V. (2000). Patterns of response to an approaches to studying inventory across contrasting groups and contexts. European Journal of Psychology of Education, XV, 33-48. Evans, E. (2104). The interface between academic education and professional training in accounting. In: The Routledge Companion to Accounting Education (ed. R.M.S. Wilson.) London: Routledge (in press). Evans, E., Burritt, R., Guthrie, J. (Eds). (2010). Challenges for accounting education at a crossroad in 2010. Centre for Accounting, Governance and Sustainability, University of South Australia and The Institute of Chartered Accountants in Australia, Sydney pp. 915. Fox, A., Stevenson, L., Connelly, P., Duff, A., & Dunlop, A. (2010). Peer-mentoring undergraduate accounting students: The influence on approaches to learning and academic performance. Active Learning in Higher Education, 11, 145-156. Flood, B. & Wilson, R.M.S. (2008). An exploration of learning approaches of prospective professional accountants in Ireland. Accounting Forum, 32, 225-239. Friedlan, J.M. (1995) The effect of different teaching approaches on students’ perceptions of the skills needed for success in accounting courses and by practicing accountants, Issues in Accounting Education, 10(1), 47-63. Friedman, A. & Lyne, S. (2001). The beancounter stereotype: towards a general model of stereotype generation. Critical Perspectives on Accounting, 12, 423-451. Fuchs, L., Fuchs, D., Phillips, N., Hamlett, C., & Karns, K. (1995). of Study/Meta-analysis: Acquisition and transfer effects of classwide peer-assisted learning strategies in mathematics for students with varying learning histories. School Psychology Review, 24, 604-620. Fuchs, D., Fuchs, L.S., Mathes, P.G., & Simmons, D.C. (1997). Peer-assisted learning strategies: Making classrooms more responsive to diversity. American Educational Research Journal, 34, 174-206.

38

ACCEPTED MANUSCRIPT

M AN U

SC

RI PT

Garver, M.S. & Mentzer, J.T. (1990). Logistics research methods; Employing structural equation modelling for construct validity. Journal of Business Logistics, 20 (1), 33-57. Geiger, M.A. & Ogilby, S.M. (2000). The first course in accounting: students’ perceptions and their effect on the decision to major in accounting. Journal of Accounting Education, 18 (1), 63-78. Gibbs, G. (2010). Dimensions of Quality, Heslington, York: The Higher Education Academy. Guney, Y. (2009). Exogenous and endogenous factors influencing student performance in undergraduate accounting modules. Accounting Education: An international journal, 18 (1), 51-73. Haggis, T. (2006). Pedagogies for diversity: retaining critical challenge amidst fears of ‘dumbing down’, Studies in Higher Education, 31, 521-535. Haggis, T. (2009). What have we been thinking of? A critical overview of 40 years of student learning research in higher education, Studies in Higher Education, 34, 377-390. Hall, M., Ramsay, A. & Raven, J. (2004). Changing the learning environment to promote deep learning approaches in first-year accounting students, Accounting Education: an international journal 13 (4), 489–505. Hardy, J. & Deppe, L. (1995). A competency-based, integrated approach to accounting education. Accounting Education: An international journal, 4 (1), 55-75. Heiat, A. & Brown, D. (2007). An empirical analysis of underlying factors affecting the choice of accounting major, Journal of College Learning and Teaching, 4, 83-98. Hoe, S.L. (2008). Issues and procedures in adopting structural equation modeling technique, Journal of Applied Quantitative Methods, 3 (1), 76-115.

AC C

EP

TE D

International Accounting Education Standards Board (IAESB) (2013). International education standard 1 (revised): Entry requirements to professional accounting education programs. http://www.ifac.org/ Jackling, B. (2005a). Perceptions of the learning context and learning approaches: Implications for quality learning outcomes in accounting, Accounting Education: An International Journal 14 (3), 271-291. Jackling, B. (2005b). Analysis of the learning context, perceptions of the learning environment and approaches to learning accounting: a longitudinal study. Accounting and Finance, 45, 597-612. Jackling, B. & McDowall, T. (2008). Peer Mentoring in an Accounting Setting: A Case Study of Mentor Skill Development 1. Accounting Education: an international journal, 17(4), 447-462. Janssen, P.J. (1996). Studaxology: The expertise students need to be effective in higher education, Higher Education, 31, 117-141. Kline, R. B. (1998). Principles and practice of structural equation modeling. New York: Guilford.

Koh, H.C. (2014). Determinants of students’ performance in accounting programmes. In: The Routledge Companion to Accounting Education (ed. R.M.S. Wilson. London: Routledge. (in press). Koh, M. Y., & Koh, H. C. (1999). The determinants of performance in an accountancy degree programme, Accounting Education: An International Journal, 8, 13-29 Lucas, U. (2000). Worlds apart: students’ experiences of learning introductory accounting. Critical Perspectives on Accounting, 11, 479-504. Lucas, U. (2001). Deep and surface approaches to learning accounting: A phenomenographic study. Accounting Education: an international journal, 10, 1-24.

39

ACCEPTED MANUSCRIPT

M AN U

SC

RI PT

Lucas, U. & Meyer, J.H.F. (2005). Towards a mapping of the student world: the identification of variation in students’ conceptions of, and motivations to learn, accounting. The British Accounting Review, 37, 177-204. Lucas, U. & Mladenovic, R. (2006). Reflections on accounting education research; how accounting education research on approaches to learning has contributed to educational practice. In Reflections on Accounting Education Research, edited by L. Murphy Smith. Sarasota, FL: American Accounting Association, Teaching and Curriculum Section. Marton, F., & R. Säljö (1976). On qualitative differences in learning I - outcomes and processes. British Journal of Educational Psychology, 46, 4-11. Meyer, J.H.F. (2000). Variation in contrasting forms of ‘memorising’ and associated observables. British Journal of Educational Psychology, 70, 163–170. Meyer, J.H.F. (2004). An introduction to the RoLI. Innovations in Education and Teaching International (Special issue: Metalearning in Higher Education), 41, 491–497. Meyer, J. & Eley, M. (1999). The development of affective subscales to reflect variation in students' experiences of studying mathematics in higher education. Higher Education 37 pp.197-216.

AC C

EP

TE D

Meyer, J.H.F., Parsons, P., & Dunne, T.T. (1990). Individual study orchestrations and their association with learning outcome. Higher Education, 20, 67-89. Meyer J.H.F. & Sass, A.R. (1993). The impact of the first year on the learning behavior of engineering students. International Journal of Engineering Education, 9, 209-217. Mladenovic, R. (2000). An investigation into ways of challenging introductory accounting students’ negative perceptions of accounting. Accounting Education: an international journal, 9, 135-155. Mo, S. & Waples, E. (2011). Evidence on the association between major and performance in the introductory accounting course. Academy of Educational Leadership Journal, 15 Special Issue, 93-107. Norgaard, C.T., & Hussein, M. E. (1990). The managerial accounting component of elementary accounting. Journal of Accounting Education, 8 (1), 77-92. Parker, L., Guthrie, J., & Linacre, S. (2011). The relationship between academic accounting research and professional practice. Accounting, Auditing & Accountability Journal, 24, 5-14 Pathways Commission (2012). The Pathways Commission. http://commons.aaahq.org/ Paver, B. & Gammie, E. (2005). Constructed gender, approach to learning and academic performance, Accounting Education: an international journal, 14, 427-444. Ramburuth, P. & Mladenovic, R. (2004). Exploring the relationship between students' orientations to learning, the structure of students' learning outcomes and subsequent academic performance, Accounting Education: an international journal 13 (4), 507527. Ramsden, P. (1979). Student learning and perceptions of the academic environment. Higher Education, 8, 411-427. Ramsden, P. (2003). Learning to Teach in Higher Education. London: Routledge. Rankin, M., Silvester, M., Vallely, M., & Wyatt, A. (2003). An analysis of the implications of diversity for students’ first level accounting performance. Accounting and Finance, 43, 365-393.

40

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Saudagaran, S.M. (1996). The first course in accounting: An innovative approach. Issues in Accounting Education, 11, 83-94. Stoner, G. & Milner, M. (2010) Embedding generic employability skills in an accounting degree: Developments and impediments. Accounting Education: An international journal, 19(1-2), 167-187. Tait, H., Entwistle, N.J., McCune, V. (1998). ASSIST: A reconceptualisation of the approaches to studying inventory. In: Rust, C., (Ed.), Improving student learning: Improving students as learners, Oxford: Oxford Centre for Staff Development, pp. 262-271. Tan, K., & Choo, F. (1991). A note on the academic performance of deep-elaborative versus shallow-reiterative information processing students, Accounting and Finance May, 6781. Tan, L.M., & Laswad, F. (2008). Impact of prior content and meta-cognitive knowledge on students' performance in an introductory accounting course. Pacific Accounting Review, 20 (1), 63-74. Trigwell, K., & M. Prosser. (1991). Relating learning approaches, perceptions of context and learning outcomes. Higher Education 22 (1), 77-87. van der Meer, J., & Scott, C. (2009). Students’ experiences and perceptions of peer assisted study sessions: towards ongoing improvement. Journal of Peer Learning, 2(1), 3-22. Watkins, D., & Regmi, M. (1990). An investigation of the approach to learning of Nepalese tertiary students. Higher Education, 20, 459–469. Webb, G. (1997). Deconstructing deep and surface: Towards a critique of phenomenography, Higher Education, 33, 195-212. Wilson, K. L., Smart, R. M., & Watson, R. J. (1996). Gender differences in approaches to learning in first year psychology students. British Journal of Educational Psychology, 66, 59–71. Wygal, D. (2014). The first course in accounting. In: The Routledge Companion to Accounting Education (ed. R.M.S. Wilson) London: Routledge (in press). Zeegers, P. (2001). Approaches to learning in science: A longitudinal study. British Journal of Educational Psychology, 71 (1), 115–132. Zeegers, P. (2004). Student learning in higher education: A path analysis of academic achievement in science. Higher Education Research and Development, 23(1), 35–56. Zeff, S. A. (1989a). Does accounting belong in the university curriculum. Issues in Accounting Education, 4 (1), 203-210. Zeff, S. A. (1989b). Recent trends in accounting education and research in the USA: some implications for UK academics. The British Accounting Review, 21 (2), 159-176.

41

ACCEPTED MANUSCRIPT

Description of observable

SC

The student has a strong motivation to succeed The student is motivated by the idea that the study of accounting is expected to be enjoyable A view of knowledge that means that it is important to identify the underlying assumptions or principles on which it is based An intention to understand the reality/meaning behind accounting Accounting is seen as enabling a new view of (or changing understanding of) business, the economy or society

M AN U

Subscale Observable Panel A: Deep (transformative) subscales Achieving Motivation Enjoyment Motivation Questioning Epistemological belief Reality/meaning behind accounting Intention Social/economic importance of Epistemological accounting belief Panel B: Surface, accumulative subscales Exam focus Intention

RI PT

Table 1: Description of ELAcc subscales

The student’s main intention is to pass the examination There is a lack of personal interest and accounting is perceived to be a dull and boring subject An epistemological belief that accounting is mainly about the study of numbers

TE D

The student feels anxious about learning accounting

EP

Worry

Motivation Epistemological belief Motivation

AC C

Lack of interest Numbers

42

ACCEPTED MANUSCRIPT

Table 2: Description of RoLI subscales

+

Learning process

EP

TE D

M AN U

+ Learning process Knowing about learning (KAL) Seeing things differently + Conception of (SDI) learning Re-reading a text for ~ Learning process meaning (RER) Repetition aids ~ Learning process understanding (RAU) Panel B: Surface (accumulative) subscales Learning is fact based (FAC) - Conception of learning Knowledge is discrete and - Epistemological factual (KDF) belief Learning experienced as ~ Motivation duty (DUT) Memorising before - Learning process understanding (MBU) Panel C: Learning pathology subscales Detailed related pathology - Learning (anti)(DRP) process Fragmentation (FRA) - Learning (anti)process Memorising as rehearsal - Learning process (MAR) Learning by example (LBE) ~ Learning process

Process of committing to memory material which is simultaneously comprehended Process of committing to memory material whose meaning is comprehended An active, deep-level process of relating new ideas to other contexts/experiences A metacognitive process via the visualisation of knowledge as if it has a form/structure A metacognitive process of knowing when someone has learned something A transformative conception involving seeing something from a new perspective Meaning emerges through re-reading a text

RI PT

Knowledge objects (KOB)

Description of observable

SC

Subscale Observable Panel A: Deep (transformative) subscales Memorising with + Learning process understanding (MWU) Memorising after + Learning process understanding (MAU) Relating ideas (RID) + Learning process

Repetition is used to aid understanding

An accumulative conception of learning involving acquisition, reproduction of facts A belief that knowledge is discrete and fact-based

A motivational influence on learning experienced as amoral duty Committing to memory material whose meaning is not comprehended Over-reliance on detail and an inability to organise it into an overall picture A lack of an organising principle for processing new information to be learned Committing to memory with no intention to understand Learning is a process of copying others

AC C

Key: +, a generally positive learning attribute; ~ an equivocal learning attribute; - a generally negative learning attribute

43

ACCEPTED MANUSCRIPT

2

3

4

5

6

7

8

9

10

.37 .24 .40 .29 -.41 -.17 .02 -.06 .31 .18 .17 .24 .31 .25 .25 .03 .08 .01 .02 -.09 .08 .12 .19 -.07 .11

.28 .43 .42 -.34 -.52 .00 -.22 .31 .21 .18 .25 .28 .28 .27 -.04 .10 .05 .02 -.15 .09 .06 .17 -.09 .05

.42 .41 -.22 -.13 -.01 -.01 .28 .18 .29 .31 .31 .39 .39 .02 -.06 -.06 .02 -.06 .09 -.04 .14 -.06 .02

.51 -.48 -.25 .05 -.06 .51 .21 .33 .39 .49 .43 .49 -.10 -.03 -.11 -.05 -.22 .03 -.03 .21 -.16 .06

-.29 -.30 .03 .00 .36 .19 .24 .25 .37 .32 .42 .01 .07 .03 .09 -.07 .14 .07 .25 -.04 .02

.35 .14 .22 -.34 -.07 -.19 -.25 -.27 -.22 -.29 .32 .21 .31 .23 .44 .10 .12 -.07 .38 -.16

.17 .31 -.16 -.09 -.08 -.07 -.11 -.11 -.10 .15 -.01 .01 .07 .25 .03 .07 -.08 .19 -.06

.26 .06 .01 .04 .02 .05 .04 .03 .15 .27 .25 .19 .18 .09 .18 .13 .18 -.13

.04 .07 .12 -.04 .03 .03 .05 .28 .18 .14 .30 .38 .19 .27 .18 .42 -.20

.31 .43 .43 .47 .46 .48 -.03 .07 -.06 .05 -.14 .15 .10 .32 -.06 .05

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

.41 -.02 -.01 -.01 .03 -.10 .13 .03 .16 -.03 .02

-.06 -.06 -.11 .05 -.11 .11 .04 .23 -.08 .03

.32 .41 .31 .45 .20 .27 .16 .41 -.15

.62 .44 .31 .15 .37 .32 .24 -.15

.40 .39 .14 .29 .23 .34 -.16

.34 .20 .52 .42 .35 -.15

.18 .28 .11 .55 -.20

.22 .20 .25 -.06

.56 .29 -.07

.13 -.02

-.13

M AN U

SC

1

TE D

SD .69 .60 .55 .49 .54 .70 .78 .69 .77 .48 .63 .65 .52 .52 .44 .50 .72 .70 .66 .78 .64 .70 .78 .66 .56 16.11

EP

Mean 2.89 2.51 2.72 3.04 2.75 1.20 2.07 2.36 1.99 2.98 2.50 2.78 2.72 2.84 2.07 3.02 1.83 2.05 1.83 1.97 1.73 2.16 2.11 2.67 1.86 63.76

AC C

1. ACH 2. ENJ 3. QN 4. REAL 5. SOC 6. EXM 7. LINT 8. NUM 9. WOR 10. KAL 11. KOB 12. MAU 13. MWU 14. RER 15. RID 16. SDI 17. DUT 18. FAC 19.KDF 20. MBU 21. FRA 22. LBE 23. MAR 24. RAU 25. DRP 26. PER

RI PT

Table 3: Descriptive statistics of the subscales: means, standard deviations and correlation coefficients

.28 .33 .31 .36 .28 .15 .12 .11 .15 .06 .15 .17 .22 .07 -.07

.44 .40 .40 .33 .06 .08 .08 .15 -.01 .09 .05 .28 .04 .00

.42 .41 .38 -.04 .01 -.04 .02 -.15 .09 .00 .19 -.05 .05

.44 .45 -.01 .05 -.04 .08 -.08 .14 .18 .50 -.03 .05

r > .05, p <.05; r > .07, p < .01; r > .08, p < .001

44

RI PT

ACCEPTED MANUSCRIPT

Table 4: Comparisons among the three-cluster profiles, ELAcc

0.58 0.42 0.46 0.30 0.50 0.42 0.64 0.66 0.75

3.48 3.11 3.20 3.58 4.13 0.42 1.38 2.11 1.71

0.50 0.50 0.44 0.33 0.58 0.32 0.74 0.85 0.88

Cluster 3 (N = 488) Mean St. Dev

SC

3.01 2.59 2.74 3.09 3.50 0.85 2.00 2.39 1.91

TE D

Achieving Enjoyment Questioning Reality/Meaning Behind Accounting Social/Economic Importance of Accounting Exam Focus Lack of Interest Numbers Worry

2.38 2.05 2.41 2.64 2.98 1.43 2.56 2.42 2.27

0.61 0.54 0.54 0.47 0.62 0.50 0.66 0.62 0.69

F, (2, 1550), p

357.643, <.001 461.013, <.001 234.642, <.001 599.464, <.001 386.664, <.001 530.525, <.001 284.709, <.001 20.179, <.001 56.388, <.001

Partial η2

.32 .37 .23 .44 .33 .41 .27 .03 .07

Inequality 2 > 1, 3; 1 > 3 2 > 1, 3; 1 > 3 2 > 1, 3; 1 > 3 2 > 1, 3; 1 > 3 2 > 1, 3; 1 > 3 1, 3 > 2; 3 > 1 1, 3 > 2; 3 > 1 1, 3 > 2; 3 > 1 1, 3 > 2; 3 > 1

Table 5: Demographic membership of clusters

EP

Cluster, variable English, % EFL Gender, % female Major, % majoring acc’ing Studied accounting at school, % yes

AC C

1. 2. 3. 4. 5. 6. 7. 8. 9.

Cluster 2 (N = 273) Mean St. Dev

M AN U

Cluster 1 (N = 790 ) Mean St. Dev

Clustering variable

1 39.6 54.8 45.6 33.9

2 39.6 49.6 57.4 44.7

3 32.2 47.7 27.8 34.6

45

ACCEPTED MANUSCRIPT

Table 6: Comparisons among the three-cluster profiles, RoLI, and performance

Note: n.s. not significant at p < .05

.43 .71 .70 .56 .76 .50 .53 .46

2.72 2.35 2.55 2.51 2.50 2.58 2.51 2.76

.46 .59 .60 .50 .63 .54 .57 .49

171.02, <.001 28.51, <.001 58.21, <.001 87.49, <.001 26.08, <.001 135.73, <.001 111.11, <.001 164.84, <.001

.18 .04 .07 .10 .03 .15 .13 .18

2 > 1, 3; 1 > 3 2 > 3; 1 > 3 2 > 3; 1 > 3 2 > 1, 3; 1 > 3 1>3 2 > 1, 3; 1 > 3 2 > 1, 3; 1 > 3 2 > 1, 3; 1 > 3

1.60 1.93 1.65 1.80

.83 .80 .72 .86

1.96 2.09 1.93 2.07

.65 .64 .61 .72

20.92, <.001 4.81, < .001 15.04, <.001 9.82, <.001

.03 .01 .02 .01

3 > 1, 2; 1 > 2 3>1 3 > 1, 2; 1 > 2 3 > 1, 2; 1 > 2

.60 .69 .78

1.40 2.19 2.02

.63 .85 .90

1.98 2.13 2.14

.60 .64 .70

79.15,<.001 n.s n.s.

0.10 -

3 > 1, 2; 1 > 2 -

EP

68.00

14.54

62.60

14.59

14.53, < .001

.71 .71 .68 .79

1.69 2.16 2.13

M AN U

1.84 2.07 1.81 1.97

TE D

.42 .61 .63 .47 .62 .41 .54 .42

66.52

14.62

SC

3.33 2.71 3.06 3.00 2.82 3.17 3.14 3.37

3.01 2.51 2.81 2.75 2.73 2.88 2.79 3.06

1, 2,> 3, 2 > 1

AC C

Panel A: Deep/transformative 1. Knowing about learning 2. Knowledge objects 3. Memorise after understanding 4. Memorise with understanding 5. Repetition aids understanding 6. Rereading a text 7. Relating ideas 8. Seeing things differently Panel B: Surface/accumulative 9. Learning as a duty 10. Learning is fact-based 11. Knowledge is discrete and factual 12. Memorising before understanding Panel C: Pathology 13. Fragmentation 14. Learning by example 15. Memorising as rehearsal Panel D: Performance 1. Mark

RI PT

Cluster 1 (N = 790 ) Cluster 2 (N = 273) Cluster 3 (N = 488) Mean St. Dev Mean St. Dev Mean St. Dev F, (2, 1509), p Partial η2 Inequality

Clustering variable

46

Approaches to Learning Reflections on Learning Accounting

Academic Outcomes Performance

M AN U

SC

Students' perceptions Expectations of Learning Accounting

TE D

Background variables Gender, English First/Second Language, Major, Prior experience of accounting

RI PT

ACCEPTED MANUSCRIPT

----------- Process ---------

--- Product (consequences) ---

EP

------------------------- Presage (antecedents) ----------------------

AC C

Figure 1: Presage-process-product model (adapted from Dunkin & Biddle, 1974)

47

ACCEPTED MANUSCRIPT

Figure 2: Comparisons across three clusters – ELAcc 4 3.5 3

RI PT

2.5 2 1.5 1

0 ACH

ENJ

QN

REAL

SC

0.5 SOC

EXAM

LINT

NUM

WOR

_____ _____

Cluster 1 - average

Cluster 2 – world of engagement Cluster 3 – world of detachment

AC C

EP

TE D

_____

M AN U

Axis Title

48

ACCEPTED MANUSCRIPT

Figure 3: Comparisons across three clusters – RoLI Deep/Accumulative subscales 4.00 3.50 3.00

RI PT

2.50 2.00 1.50

0.50 KOB

MAU

_____ _____

RAU

RER

RID

SDI

Cluster 1 - average

Cluster 2 – world of engagement Cluster 3 – world of detachment

AC C

EP

TE D

_____

MAW

M AN U

KAL

SC

1.00

49

ACCEPTED MANUSCRIPT

Figure 4: Cluster analysis of RoLI Surface scores 2.50

RI PT

2.00

1.50

1.00

FAC

_____ _____

MBU

Cluster 1 - average

Cluster 2 – world of engagement

TE D

_____

KDF

M AN U

DUT

SC

0.50

AC C

EP

Cluster 3 – world of detachment

50

ACCEPTED MANUSCRIPT

Figure 5: Comparisons of three clusters on RoLI Pathology measures 2.50

RI PT

2.00

1.50

1.00

LBE

Cluster 1 - average

Cluster 2 – world of engagement Cluster 3 – world of detachment

AC C

EP

TE D

____ ____ ____

MAR

M AN U

FRA

SC

0.50

51

ACCEPTED MANUSCRIPT

1

AC C

EP

TE D

M AN U

SC

RI PT

Alongside the deep/surface dichotomy exists a third ‘strategic’ approach Ramsden (1979), a student seeks to maximize academic performance by effective study methods, such as analysing the structure and content of prior examinations to predict questions. Subsequent studies have typically failed to reproduce the strategic approach as a distinct approach to learning. Some authors (e.g., Janssen 1996; Entwistle, Tait, & McCune, 2000) consider a strategic approach as students’ extrinsic motivation to adopt a deep or surface approach so they may maximize their grades 2 Active learning and constructivist approaches to learning are synonymous with terms such as ‘studentcent(e)red learning’, ‘student-cent(e)red teaching’, ‘active learning’, ‘constructivist learning’, ‘studentactivating’, ‘problem-based learning’, ‘powerful learning environment’, ‘minimal guidance’, ‘discovery learning’, ‘open-ended learning environment’, ‘collaborative learning’, ‘cooperative learning’, ‘project-based learning’, and ‘case-based learning’ (Baeten et al., 2010 p.245). 3 Of the total sample of 1,553, a small number declined to report their gender (N=1), their major (N=17), whether English was their first or second language (N=2), and whether they had studied accounting at school (N=4). 4 The Association to Advance Collegiate Schools of Business (AACSB) provides internationally recognised, specialised accreditation for business and accounting programs at the bachelor’s, master’s, and doctoral level. 5 The ELAcc is available from the corresponding author. 6 To evaluate the fit of the measurement models, multiple indices of fit are usually recommended. Specifically, Garver and Mentzler (1999) recommend the use of three indices. Accordingly, the comparative fit index (CFI) is reported as an incremental fit index, i.e., one which evaluates of how much better the model fits the data, compared with a restrictive baseline model. Second, the root mean square error of approximation (RMSEA) is used and its associated 90% confidence intervals (CIs). RMSEA assesses how well the model fits the population covariance (Browne and Cudeck, 1993). Third, the study uses the non-normed fit index (NNFI), which 2 compares the proposed model’s fit to a baseline (null) model. For completeness, Chi-square (χ ), as a measure of the fit of the reproduced covariance matrix implied by the model to the covariance matrix of the sample data, is reported. Although no precise standards exist to indicate what value of fit indices are needed for a satisfactory fit, typical guidelines are that the CFI should exceed .9 (Hu & Bentler, 1995; Kline, 1998). Browne and Cudeck (1993) suggest that values of RMSEA less than .05 are indicative of a close fit. NNFI values in excess of .9 represent a good fit (Hoe, 2008). 7 Confidence intervals are shown for the RMSEA computations and are interpreted in a similar way to the RMSEA value. The confidence interval width advises of the precision of the RMSEA estimate. 8 The RoLI is subject to copyright and available from its author. 9 The CFA undertaken for the RoLI was separate to the CFA undertaken for the ELAcc reported above. A more detailed account of the psychometric qualities of scores yielded by the ELAcc is reported in Duff and Mladenovic (2014b). . 10 In this instance, the NNFI estimate just falls short of the recommended .9 cutoff, the estimates for CFI and RMSEA provide confidence that the data fit the model. 11 Eta-squared (ƞ2) is a measure of effect size, analogous to R2 in multiple regression, is interpreted as .02 = small; .13 = medium; and .26 = large (Cohen, 1977). 12 Table 4 and 6 report the F-ratios and partial η2 statistics for replication purposes. The inequalities identify which of the group differences are statistically significant. 13 Table 5 reports the percentage of students in each cluster. For example in cluster 1, 39.6% are categorised as EFL and correspondingly, 60.4% are ESL. 14 Textbooks for non-accounting students, of course, have been in existence for many years, for example (Dyson, 2010; Atrill & McLaney, 2012).

52