Development and use of an instrument to measure students’ perceptions of a business statistics...

ORI GIN AL PA PER

Development and use of an instrument to measurestudents’ perceptions of a business statistics learningenvironment in higher education

Thuyuyen H. Nguyen1 • Michael Newby1 •

Panayiotis G. Skordi1

Received: 25 October 2013 / Accepted: 5 May 2014 / Published online: 26 July 2015� Springer Science+Business Media Dordrecht 2015

Abstract Statistics is a required subject of study in many academic disciplines, including

business, education and psychology, that causes problems for many students. This has long

been recognised and there have been a number of studies into students’ attitudes towards

statistics, particularly statistical anxiety. However, none of these studies investigated the

classroom learning environment in relation to students’ attitudes towards statistics. This

paper describes the development and validation of two new instruments, the Business

Statistics Computer Laboratory Environment Inventory and the Attitude towards Business

Statistics instrument. The former measures appropriate aspects of the learning environment

of a computer-based statistics class, whilst the latter measures students’ attitudes towards

business statistics. The instruments were administered to undergraduate business students

at a university in Southern California. When exploratory factor analysis was carried out, it

was found that the items of both instruments loaded onto their a priori scales. All the scales

exhibited both internal reliability and discriminant validity. Correlation analysis between

the scales of the two instruments demonstrated a strong relationship between aspects of the

learning environment and students’ attitudes towards business statistics.

Keywords Attitude � Business statistics � Computer-based learning environments �Instrument

& Michael [email protected]

1 Information Systems and Decision Sciences, California State University Fullerton, 800 N. StateCollege Blvd., Fullerton, CA 92831-3599, USA

123

Learning Environ Res (2015) 18:409–424DOI 10.1007/s10984-015-9192-3

http://crossmark.crossref.org/dialog/?doi=10.1007/s10984-015-9192-3&domain=pdf

http://crossmark.crossref.org/dialog/?doi=10.1007/s10984-015-9192-3&domain=pdf

Introduction

Statistics is a subject that is studied in high school, as well as at university from the

undergraduate to the doctoral levels. In higher education, it is a required course for a large

number of disciplines including actuarial science, business, econometrics, epidemiology,

environmental studies, geology, engineering and psychology. In all of these fields, statistics

is used as a tool to analyse and summarise large amounts of data. In other words, because

these disciplines use applied statistics, statistics is seen as a service subject in the teaching

context. Many students do not know why they have to take a course in statistics or how it is

related to their main discipline (Petocz and Reid 2005). One student in an earlier study by

Petocz (1998) was reported as saying: ‘‘Statistics is something strange and useless. I was

studying engineering, and didn’t know why I had to study statistics’’. Even when students

have some appreciation of the importance of statistics in their future courses and careers,

they tend to dismiss it as being dull, boring and tedious and most students would not study

it if it were not compulsory (Gordon 2004). It would seem that many students view

statistics courses as something that they have to pass so that they can proceed to focus on

the real subjects in their degree programme.

In most business degree programmes, at least one statistics course is mandatory, and this

is certainly the case for those programmes that are accredited. The purpose of such courses

is to provide skills in ‘‘statistical data analysis and management science as they support

decision-making processes throughout an organization’’ (Association to Advance Colle-

giate Schools of Business [AACSB] 2012). Unfortunately, many students find statistics

very difficult, causing them anxiety which can result in poor performance and lead to

business statistics courses having relatively high failure rates (Cook 2010). This is not a

recent phenomenon (Williams et al. 2008) but, with the increase in the number of

undergraduate and graduate business degree programmes over the past three decades, it is

becoming more prevalent. One interesting observation that has been made is that, although

statistics is a subject requiring mathematics, business statistics is not a subject that can be

taught in the same way as mathematics (Ramsey 1999). Just getting an answer is not

enough; that answer has to be interpreted to determine what it means in the context of the

application. In addition, many students believe that there is one correct answer from

analysing data (Onwuegbuzie and Wilson 2003), just as there is one correct answer to a

mathematical problem. In many ways, business statistics has more in common with verbal

reasoning than with mathematics (Williams 2010). Having said this, it is still the case that

prior mathematical courses do have a positive effect on student grade performance in

statistics (Green et al. 2009), probably because of the need to manipulate numerical data.

One phenomenon with regard to student learning of statistics that has received a great

deal of attention in recent years is statistics anxiety (Carlson and Winquist 2011; Francis

and Lipson 2010; Hsu et al. 2009; Meletiou-Mavrotheris et al. 2007; Nguyen and Hampson

2012; Onwuegbuzie 2000; Onwuegbuzie and Wilson 2003; Pan and Tang 2005; Williams

et al. 2008). The main underlying causes of statistics anxiety among students would appear

to be a lack of numeracy skills, a lack of confidence in doing anything mathematical, and

low self-esteem (Zeidner 1991), culminating in a feeling that they do not really understand

the meaning or purpose of what they are doing (Wilensky 1997). This anxiety sometimes

manifests itself as fear of statistics with students making statements such as ‘‘This

[statistics] course is scary’’ (Nguyen and Hampson 2012) and ‘‘I’ve never been frightened

of math. In fact, I received an A in my last math class. Yet I am terrified of statistics’’

(Onwuegbuzie et al. 1997).

410 Learning Environ Res (2015) 18:409–424

123

Anxiety is just one aspect of attitude, which also includes perceived usefulness and

enjoyment as well as others. Attitude in general, not just anxiety, has been shown to be

associated with achievement in science (Freedman 1997), mathematics (Hemmings et al.

2011; Ma and Kishor 1997), business computing (Newby and Fisher 2000) and statistics

(Onwuegbuzie 2000; Onwuegbuzie and Wilson 2003; Pan and Tang 2005). In an educa-

tional setting, a factor that has an impact on attitude is the classroom learning environment.

The relationship between the learning environment and attitude towards mathematics was

investigated by Taylor and Fraser (2013) and between environment and attitude to business

computing by Newby (2002). To our knowledge, this is the first study of the relationship

between the learning environment of a computer-based statistics classroom and attitude

towards statistics. This paper describes the development and validation of two instruments,

one for measuring students’ perceptions of the learning environment of a computer

classroom used for statistics classes, and another for measuring the attitudinal outcomes

from the class. The next section deals with the literature on learning environments and this

is followed by a brief review of attitudes towards statistics, in particular, statistics anxiety.

Learning environments

Student learning and achievement are affected by many factors, including student char-

acteristics (Onwuegbuzie and Wilson 2003), teacher characteristics (Williams 2010),

formative assessment approaches (Newby and Nguyen 2010; Nguyen and Ulbrich 2013)

and the classroom learning environment (Fraser 1998). Of these, classroom learning

environments have received the most attention, having been an area of academic study for

over half a century (Dorman and Fraser 2009). It is recognised that the environment has a

significant impact on all aspects of learning. Classroom environment research concerns

psychosocial aspects of a single classroom, and this has succeeded in conceptualising

learning environments. Although the concept of a classroom environment is a subtle one,

teachers have always been aware of it in an informal manner. Different classes have

different characteristics arising from the ways in which individuals interact with each

other, with the teacher and with their environment.

The use of objective measures for classroom environments began with two research

programmes that started in the late 1960s. As part of the evaluation of the Harvard Physics

Project (Anderson et al. 1968), the Learning Environment Inventory (LEI) was developed

(Anderson and Walberg 1968; Walberg and Anderson 1968). At about the same time, but

arising from research into social climate scales in different settings, the Classroom

Environment Scale (Trickett and Moos 1973) was developed for measuring the environ-

ment of a classroom. The scales of these instruments measure various aspects of students’

perceptions of the classroom environment including interactions between students, inter-

actions between students and teacher, course organisation, class organisation and even the

physical environment. Since that time, many different instruments for measuring a class-

room learning environment have been developed. There are instruments that focus on

university settings, others for high schools and also for elementary schools (Fraser 1998).

A particularly robust instrument for general use, the What is Happening In this Class?

(WIHIC) questionnaire developed by Fraser et al. (1996) has been used in a variety of

settings, achieving what Dorman (2008) calls bandwagon status in the assessment of

classroom environments. The WIHIC has seven scales: Student Cohesiveness, Teacher

Support, Involvement, Investigation, Task Orientation, Cooperation, and Equity. As well

as instruments developed for general classroom environments, there are also many

instruments used in specialist settings such as high school science laboratories (Fraser et al.

Learning Environ Res (2015) 18:409–424 411

123

1993; Taylor and Fraser 2013), higher education seminars (Fraser and Treagust 1986),

distance learning (Jegede et al. 1995), language laboratories (Okan 2008) and medical

universities (Kek and Huijser 2011).

With the growth in the use of information technology in educational settings over the

past two decades, instruments have been developed to measure many aspects of the

environment of computer laboratory classrooms (Newby and Fisher 1997, 2000), to

research the environment of on-line web-based instructional systems (Chang and Fisher

2003), blended web-based environments (Chandra and Fisher 2009) and technology-in-

tegrated science classrooms (Wu et al. 2009). The widespread use of this technology has

given rise to a category of learning environments known as Technology Rich Learning

Environments, and a number of examples of these are given in Khine and Fisher (2003).

Dorman and Fraser (2009) developed the Technology-Rich Outcomes-Focused Learning

Environments Inventory (TROFLEI), based on the WIHIC instrument (Fraser et al. 1996)

with scales added for computer usage, differentiation and young adult ethos.

Classroom learning environment instruments have been used in a number of studies to

investigate the relationship between learning environment and attitudinal variables, such as

anxiety, enjoyment and perceived usefulness of a course or a subject. Newby and Fisher

(2000) demonstrated a causal relationship between students’ perceptions of their computer

laboratory environment and their attitudes towards computers and computing courses.

Their findings were supported and extended by a later study (Dorman and Fraser 2009) of

secondary school students in Australia in which they showed a relationship between

antecedents (age, school grade and computer usage), the learning environment and atti-

tudes to both the subject and computer use. Other studies have found associations between

the learning environment and attitudes towards science (Allen and Fraser 2007; Wolf and

Fraser 2008) and attitudes towards mathematics (Afari et al. 2013).

Attitude towards statistics

Attitude towards anything, whether physical or abstract, is a concept that is recognised as

having three components: cognitive, affective and behavioural (McGuire 1969). This

applies to attitude towards certain individuals, computers, learning, statistics or, even,

learning statistics. A cognitive component is what an individual believes, such as

‘‘statistics is a useful tool in business’’. An affective component is how the individual feels

and is based on emotions such as anxiety or enjoyment. An example of this is ‘‘business

statistics scares me’’ or ‘‘I enjoy this statistics class’’. A behavioural component is a

predisposition to act based on the cognitive and affective components. An example of this

type is procrastination regarding taking a statistics course. Students could believe that the

course is difficult and therefore they feel worried about having to do it in case they get a

low grade, and so they delay in enrolling in the course. In order for someone to have any

emotions about a situation, it is necessary for them to have a belief about that situation

(Kek and Huijser 2011). Moreover, this belief can be associated with pleasant or

unpleasant events, or desirable or undesirable goals. It is at this point that the individual’s

emotions are affected by the situation, which can lead to a predisposition to behave in a

certain way. If the events are pleasant or the goals are perceived to be desirable, the

individual will have a positive attitude towards the situation, and vice versa (McGuire

1969).

Of all the attitudinal outcomes from learning statistics, the one that has received the

most attention is anxiety and this has been defined by a number of authors. For example,

Cruise et al. (1985) define it as ‘‘a feeling of anxiety when taking a statistics course or


123

doing statistical analysis; that is gathering, processing and analyzing data’’. Zeidner (1991)

expands on this by highlighting the accompanying effects of worry, tension and stress

when students are taking or have to take a statistics course. Onwuegbuzie et al. (1997)

define statistics anxiety simply as an anxiety that occurs when a student encounters

statistics in any form or at any level. Statistics anxiety is related to mathematics anxiety,

which has been studied intensively over many years (Pan and Tang 2005; Williams et al.

2008), but they are considered to be different (Williams 2010). The main difference is that,

while mathematics anxiety relates to manipulation of numbers and symbols, statistics has

the added factor of interpretation of the results with the insecurity of there possibly being

no one correct answer.

There have been a number of studies of statistics anxiety related to various factors,

which are classified by Onwuegbuzie and Wilson (2003) as situational, dispositional and

environmental. Situational factors refer directly to the specific course and include course

status (whether required or elective), student major, course structure, instructor attitude and

prior statistics knowledge. Dispositional factors are related to an individual student’s

attitude; they include self-esteem, perceived importance of statistics, mathematical self-

concept, and perceived mathematical and intellectual ability. Finally, environmental fac-

tors relate to attributes of the students such as gender, age, learning style, ethnicity and

whether or not they are studying overseas.

Statistics anxiety has been found to be related to a large number of factors including

academic major, learning style, expected grade, calculator use, mathematical ability and

the length of time since the student’s previous mathematics course (Pan and Tang 2005).

Ethnicity has also been shown to be related to anxiety (Onwuegbuzie and Wilson 2003),

with a study by Liu et al. (2012) showing that college students in China had lower anxiety

than their Western counterparts in the US and the UK. Williams (2010) demonstrated that

the instructor’s behaviour also affected student anxiety. She showed that instructor

immediacy, which refers to a set of communicative behaviours (Bell 1998), is directly

associated with many aspects of statistics anxiety. The performance of low-achieving

students is particularly affected by statistics anxiety.

Many studies have revealed that statistics anxiety has a negative impact on performance

(Lalonde and Gardner 1993; Onwuegbuzie and Seaman 1995; Pan and Tang 2005; Wil-

liams 2010; Zanakis and Valenzi 1997), and in many cases this anxiety leads to students to

delaying enrolling in such courses until the last possible moment (Onwuegbuzie 2004).

This is not desirable at either undergraduate or graduate level, because it could lead to

students being unprepared for other courses that require the use of statistical techniques. As

a potentially complicating factor, most statistics courses now have a computer laboratory

component, and students use software such as Microsoft Excel or SPSS (Hsu et al. 2009),

which means that in many cases they do not do the calculations by hand (Onwuegbuzie and

Wilson 2003). This approach relies even more on the interpretation of the results than

using hand calculations, and this in turn could lead to even greater statistics anxiety.

Instruments have been developed for measuring different aspects of attitude in several

disciplines and settings such as mathematics (Hannula 2002), science (Fraser 1978) and

computer usage (Newby and Fisher 1997). Ramirez et al. (2012) review a number of

instruments for measuring students’ attitudes towards statistics, and these cover those having

a single construct, such as Self-Efficacy to Learn Statistics instrument (Finney and Schraw

2003), to six-construct instruments such as the Statistical Anxiety Rating Scale (STARS)

(Cruise et al. 1985) and the Survey of Attitudes Towards Statistics-36 (Schau 2003). The

instrument that has been used in the majority of studies into attitudes towards statistics is the

STARS. It has six scales; four of them measure an aspect of anxiety or fear (interpretation


123

anxiety, test anxiety, fear of asking for help, and fear of the instructor); worth of statistics

measures the student’s perceptions of the usefulness of statistics; and computational self-

concept measures the student’s belief in his/her ability to perform the required calculations.

Methods

Sample

The sample consisted of 302 students enrolled in a core statistics course as part of an

undergraduate business degree programme at a public university in Southern California. As

this is an upper level course, all business students are required to take it. The characteristics

of the sample are shown in Table 1.

Course, procedures and intervention

Students are taught in classes of up to 45 over one semester of 15 weeks and each section of

the course involves a lecture session and a practical laboratory session. The assessment for the

course consists of four tests: three in-class tests and a final examination, each worth 25 % of

the final score of 100. To remove instructor bias, the students surveyed were all taught by the

same instructor. Data were collected over three semesters, with the surveys being adminis-

tered in the twelfth week of each semester. A one-way ANOVA was performed on the data to

test for differences in overall achievement in each of the three semesters to test if the semester

in which the student took the course could affect their score.

Instruments

For this study, two instruments were developed, both of them adapted from existing

instruments. To measure the learning environment of a computer based statistics

Table 1 Description of sampleDescriptor Values Frequency (%)

Gender Male 46.4

Female 53.6

Age 24 or under 15.9

Between 25–29 61.9

30 and over 22.2

Prior knowledge of statistics All 6.6

Some 80.8

None 12.6

Major Accounting 32.8

Business studies 14.6

Economics 6.6

Finance 11.6

Information systems 6.6

Management 10.6

Marketing 17.2


123

classroom, the Business Statistics Computer Laboratory Environment Inventory (BSCLEI)

was developed. The selection of appropriate constructs for this instrument was based on the

type of class being surveyed. These were practical computer workshops, in which students

were required to apply the techniques learned in the lecture sessions. They were supervised

by the instructor, who provided exercises for the students to work through but, in many

ways, these classes were relatively informal. Given the nature of the class, the following

five constructs were selected as being appropriate for the instrument: Student Cohesive-

ness, Integration, Technology Adequacy, Involvement and Task Orientation. The first three

were adapted from the Computer Laboratory Environment Inventory (CLEI) (Newby and

Fisher 1997) and the latter two from the WIHIC (Fraser et al. 1996). Student Cohesiveness

was chosen as a construct because it had been observed that students in such classes

collaborate with each other, Integration was included because the exercises were designed

to be based on material covered in the lectures, Technology Adequacy was selected

because the exercises were computer based and appropriate hardware and software were

necessary, Task Orientation was appropriate because the laboratory exercises are struc-

tured and students need to know what they are required to do and, finally, Involvement was

chosen because students need to do the exercises in order to gain proficiency in applying

statistical techniques to business problems. Each construct has five items, on a five-point

Likert scale ranging from Strongly Disagree to Strongly Agree. A description of the

constructs, together with a sample question from each and a reference to where they were

derived from, is given in Table 2.

The Attitude towards Business Statistics (ABS) instrument consists of four constructs:

Anxiety, Enjoyment, Perceived Usefulness of the Course and Perceived Usefulness of

Statistics. These constructs were adapted from the Attitude towards Computing and

Computing Courses (ACCC) instrument (Newby and Fisher 1997) and the Statistical

Anxiety Rating Scale (STARS) (Cruise et al. 1985). Again, each construct has five items

Table 2 Business Statistics Computer Laboratory Environment Inventory (BSCLEI)

Construct Description Sample item References

StudentCohesiveness(SC)

Extent to which students know, helpand are supportive of each other

I get on well with the otherstudents in my workshopsessions

Newby and Fisher(1997)

Integration (IN) Extent to which the IT workshopsessions are linked to non-laboratory and theory classes

What I do in the lecture isrelated to my workshopexercise


TechnologyAdequacy(TA)

Extent to which IT facilities(computers and statisticalsoftware) adequately support theclass activities

The computer softwareavailable enables me towork on my exercisesefficiently


Involvement(IV)

Extent to which students participateactively and attentively in classactivities

I put effort into what I do inthe workshop sessions

Fraser et al.(1996)

TaskOrientation(TO)

Extent to which it is important tocomplete planned activities and tostay on the subject matter

I know exactly what has tobe done in our workshopsessions

Fraser et al.(1996)


123

on a five-point Likert scale from Strongly Disagree to Strongly Agree. A description of the

constructs, together with a sample question and a reference to where they were derived

from is given in Table 3.

Instrument validation procedures

Exploratory factor analysis using principal component analysis with varimax rotation was

performed to validate the BSCLEI. For the validation of the ABS, principal component

analysis followed by a direct oblimin rotation with d = 0 was performed. The reason for

direct oblimin rotation being carried out is that, because we believed that the factors within

ABS would be correlated, using oblique rotation could produce clearer axes than using

varimax rotation (Tabachnick and Fidell 2010).

Results

Sample

The results of the one-way ANOVA performed on the students’ overall scores across

semesters to test for differences in the average achievement are shown in Table 4. This

analysis shows that the semester in which the students took the course was not significant

with respect to students’ scores, F (2299) = 0.362, p = 0.696. This implies that students in

each semester had a similar range of abilities.

Instrument validation

The results of the principal component analysis with a varimax rotation on the 25 items of

the BSCLEI showed that the Kaiser–Meyer–Olkin (KMO) sampling adequacy measure-

ments (Kaiser 1958) was 0.852, indicating that the matrix is factorable. Using eigenvalues

greater than 1 as the criterion, with a cutoff point for the loadings at 0.4, five components

were extracted for the BSCLEI. The extracted five factors explained 50.3 % percent of the

Table 3 Attitude towards Business Statistics (ABS)

Construct Description Sample item References

Anxiety (AX) Extent of a student’snervousness or lack ofcomfort with statistics

Working with statistics makesme nervous

Cruise et al. (1985),Newby and Fisher(1997)

Enjoyment (EN) Extent to which a studentenjoys learning statistics

Learning about statisticalanalysis is interesting


PerceivedUsefulness ofthe Course(PUC)

Extent to which the studentfound the course useful

This course provided me withskills I expect to use in futureemployment


PerceivedUsefulness ofStatistics (PUS)

Extent to which the studentbelieves statistics is usefulin general

Statistical analysis is animportant factor in the successof a business

Cruise et al. (1985)


123

variance, which is satisfactory for social sciences studies (Kline 1994). The final results of

principal component analysis for the BSCLEI are reported in Table 5.

Table 5 shows that all the hypothesised factors of the BSCLEI had items with high

loadings on to their a priori factors. For Student Cohesiveness, loadings were between

0.593 and 0.773, for Integration loadings were between 0.524 and 0.658, for Technology

Adequacy loadings were between 0.453 and 0.705, for Involvement loadings were between

Table 4 One-way ANOVAresults for achievement over thethree semesters

Semester N Mean SD Minimum Maximum

S01 111 77.8 10.0 53 99

S02 134 76.5 15.3 20 98

S03 57 76.3 12.9 42 100

F-value = 0.362df = 299, 2Sig. = .696

Table 5 Factor loadings ofrotated component matrixa forBSCLEI constructs

Extraction method: Principalcomponent analysis

Rotation method: Varimax withKaiser normalizationa Rotation converged in 6iterationsb Variance explained as apercentage

Item Factor 1 Factor 2 Factor 3 Factor 4 Factor 5

SC_01 0.593

SC_02 0.773

SC_03 0.715

SC_04 0.771

SC_05 0.675

IN_01 0.623

IN_02 0.531

IN_03 0.658

IN_04 0.570

IN_05 0.524

TA_01 0.521

TA_02 0.453

TA_03 0.705

TA_04 0.702

TA_05 0.520

IV_01 0.617

IV_02 0.780

IV_03 0.626

IV_04 0.472

IV_05 0.524

TO_01 0.692

TO_02 0.576

TO_03 0.555

TO_04 0.541

TO_05 0.529

VEb 24.66 9.04 6.38 5.39 4.83

Eigenvalue 6.17 2.26 1.60 1.35 1.21


123

0.472 and 0.780, and for Task Orientation loadings were between 0.529 and 0.692. Table 6

shows the internal reliabilities and mean correlations for the scales of the BSCLEI. The

internal reliabilities varied from 0.70 to 0.79, indicating that each scale had satisfactory

internal consistency (Cronbach 1951); the mean correlation between the scales varied from

0.08 to 0.25, suggesting that the scales measured distinct aspects of the learning

environment.

The results of the principal component analysis with direct oblimin rotation for the 20

items of the ABS showed a KMO of 0.916 indicating that the matrix is factorable. Again,

using eigenvalues greater than 1 as the criterion, with a cutoff point for the loadings at 0.4,

four components were extracted for the ABS. The extracted four factors explained 61.8 %

percent of the variance. The final results of principal component analysis for the ABS are

reported in Table 7.

The items in hypothesised factors of the ABS loaded on to their a priori factors with

high loadings. For Anxiety, loadings were between 0.747 and 0.832, for Enjoyment

between 0.637 and 0.769, for Usefulness of the Course between 0.595 and 0.765, and for

Usefulness of Statistics between 0.584 and 0.683. Table 8 shows the internal reliabilities

and mean correlations for the scales of the ABS. The internal reliabilities varied from 0.72

to 0.86, indicating that each scale had satisfactory in terms of internal consistency; the

mean correlation between the scales varied from 0.44 to 0.56, suggesting that the scales

measured distinct but correlated aspects of the students’ attitude towards statistics.

Correlations between learning environment and attitudes

As the reason for developing these instruments was to investigate the relationship between

the learning environment and students’ attitudes, Pearson correlation coefficients between

the scales of the BSCLEI and the scales of the ABS were calculated. These are given in

Table 9. An examination of the correlation coefficients shows that there were 18 statisti-

cally significant associations between the scales of the instruments.

Involvement correlated strongly with all attitudinal variables, particularly with Per-

ceived Usefulness of Statistics and Enjoyment, which suggests that, if students participate

in the workshop sessions and do the class exercises, they are likely to appreciate the

usefulness of statistics and enjoy the class more. They also are likely to be less anxious and

to perceive the class as useful. Technology Adequacy also correlated with all attitudinal

variables, most strongly with Perceived Usefulness of Statistics positively and Anxiety

negatively. This highlights the importance of appropriate hardware and software for a

computer-based classroom. It appears that students had a positive view of statistics if the

software produced output that was easy to understand, and were less anxious if software

performs well and the computer works properly. The relationship between technology

adequacy and anxiety is consistent with an earlier study on use of computers in

Table 6 Internal reliabilitiesand mean correlation coefficientof the scales of the BSCLEI

Scale N Reliability Mean correlation

Student Cohesiveness 302 0.79 0.08

Integration 302 0.72 0.12

Technology Adequacy 302 0.75 0.18

Involvement 302 0.70 0.25

Task orientation 302 0.70 0.20


123

programming classes (Newby and Fisher 1997). Task Orientation correlated with all

attitudinal factors, but strongly only with Perceived Usefulness of Statistics and Perceived

Usefulness of the Course. Again, this would imply that clear instructions for the exercises

are likely to be lead to students perceiving both the course and statistics to be useful, to

being less anxious and to having greater enjoyment.

With the exception of Enjoyment, Integration was associated with all attitudinal factors,

suggesting that a course in which the relationship between lecture and workshop is clear

tends to be associated with reduced anxiety and a perception of increased usefulness of

both the course and statistics itself. Finally, Student Cohesiveness correlated with all

attitudinal scales except Anxiety, implying that students perceive statistics and the course

as useful and enjoyable if they are allowed to work together. At first glance, the non-

Table 7 Factor loadings ofrotated component matrixa forABS constructs

Extraction method: Principalcomponent analysis

Rotation method: Oblimin withKaiser Normalizationa Rotation converged in 6iterationsb Variance explained as apercentage

Item Factor 1 Factor 2 Factor 3 Factor 4

AX_01 0.827

AX_02 0.832

AX_03 0.756

AX_04 0.777

AX_05 0.747

EN_01 0.775

EN_02 0.676

EN_03 0.769

EN_04 0.703

EN_05 0.637

PUC_01 0.703

PUC_02 0.595

PUC_03 0.765

PUC_04 0.634

PUC_05 0.738

PUS_01 0.626

PUS_02 0.683

PUS_03 0.594

PUS_04 0.605

PUS_05 0.584

VEb 38.44 10.86 6.80 5.71

Eigenvalue 7.69 2.17 1.36 1.14

Table 8 Internal reliabilities and mean correlation coefficient of the scales of the ABS

Scale N Reliability Mean correlation

Anxiety 302 0.86 -0.44

Enjoyment 302 0.86 0.56

Perceived Usefulness Course 302 0.77 0.56

Perceived Usefulness Statistics 302 0.72 0.52


123

significant association between Student Cohesiveness and Anxiety seems counterintuitive,

given the nature of these workshops, but this result supports findings from a previous study

(Newby and Fisher 1997). It seems that, if students have anxiety about statistics, then

collaborating with other students does not change that feeling. Indeed, if all students in a

group are anxious, working together could well make it worse.

Looking at the correlations from the perspective of attitudinal variables, what stand out

are the significant relationships between all environment variables and Perceived Useful-

ness of the Course and Perceived Usefulness of Statistics. This implies that how students

perceive their computer classroom environment is strongly associated with their belief

about the usefulness of both the course and statistics.

Conclusions

This paper described two instruments, the BSCLEI for measuring aspects of the learning

environment of a computer-based business statistics workshop and the ABS for measuring

students’ attitudes towards a business statistics course. For both instruments, exploratory

factor analysis resulted in all items loading onto their postulated scales, with the five

factors of the BSCLEI explaining over 50 % of the variance and the four factors of the

ABS explaining over 60 % of the variance. The scales of the BSCLEI were shown to be

reliable and that they measured separate aspects of the learning environment. For the ABS,

the scales again were shown to be reliable. The mean correlations between the scales of the

ABS were high, which is to be anticipated with an instrument that measures attitudes.

However, it should be noted that the mean correlations are much smaller than the corre-

sponding reliability coefficients, indicating that the scales do measure separate aspects of

the students’ attitudes.

Correlation analysis indicated that there were associations between most of the learning

environment variables and the attitudinal variables. The aspects of the learning environ-

ment that had the strongest relationships with attitudes were Technology Adequacy,

Involvement and Task Orientation. This would imply that, to improve attitudes towards

business statistics, instructors should ensure that workshop exercises are well-designed,

that students do the exercises during the workshop sessions, and that the software chosen as

the vehicle for these classes is up to the task. Anxiety was not only associated with

Technology Adequacy, Involvement and Task Orientation, but it had a significant negative

relationship with Integration, suggesting that anxiety about statistics also might be reduced

Table 9 Correlation coefficients between the scales of the BSCLEI and the ABS

ABS scale Correlation with BSCCEI scale

StudentCohesiveness

Integration TechnologyAdequacy

Involvement Taskorientation

Anxiety -0.066 -0.202** -0.302** -0.254** -0.133*

Enjoyment 0.146* 0.079 0.257** 0.328** 0.144*

Perceived UsefulnessCourse

0.171** 0.131* 0.239** 0.268** 0.168**

Perceived UsefulnessStatistics

0.201** 0.172** 0.355** 0.445** 0.289**

* p\ 0.05; ** p\ 0.01


123

by designing a course in which the student can see the relationship between theory and

practice. Student Cohesiveness was correlated with all attitudinal variables except Anxiety,

which would imply that students in a class where they are more helpful to and supportive

of each other are more likely to enjoy the class and perceive both the course and statistics

to be useful.

As statistical skills are now required in a number of disciplines and the study of

statistics is no longer restricted to particular subjects (British Academy 2012), these

findings have implications for both instructors and educational administrators. For

instructors who are designing and running a service statistics course, it is important that the

course is structured in such a way that the exercises for laboratory sessions are based on the

topics covered in the lectures and that the requirements are clear, easy to follow and

manageable in the time available. It is also desirable that the exercises are derived from

realistic problems appropriate to the discipline (business, in this case). Further, it must be

made clear to students that these exercises are very important as practice is the only way to

gain proficiency in a subject such as statistics. They should be encouraged to work through

these exercises during the laboratory sessions when the instructor is there to give them

assistance. Allowing students to collaborate during these sessions could be particularly

beneficial to students who do not see the course or statistics as relevant. It is also rec-

ommended that educational administrators ensure that computer laboratories are equipped

with suitable computers together with appropriate statistical software. Finally, because it

was observed that perceived usefulness of statistics was strongly correlated with all

environment variables, improving the classroom learning environment could lead to a

greater appreciation of the relevance and usefulness of statistics.

There are, of course, limitations with this study. Although the sample size was adequate,

the sample was taken from students in an undergraduate programme at one university in

Southern California. It would be of value to administer the instruments to students in

graduate programmes, particularly because this seems to be a major problem area (On-

wuegbuzie 2004; Onwuegbuzie and Wilson 2003; Williams 2010). Also, the scales of the

BSCLEI were selected based on the researchers’ experience as being most likely to be

associated with attitudes for computer workshops. It would be useful to include other

aspects of the learning environment in a future investigation. Despite these limitations, this

research has demonstrated a relationship between aspects of classroom learning environ-

ment and attitude towards statistics, and has led to the development and validation of two

instruments which can be used in future research.

Acknowledgments The authors wish to thank Professor Darrell Fisher of Curtin University, WesternAustralia for his insightful comments on an early draft of this paper.

References

Afari, E., Aldridge, J. M., Fraser, B. J., & Khine, M. S. (2013). Students’ perceptions of the learningenvironment and attitudes in game-based mathematics classrooms. Learning Environments Research,16(1), 131–150.

Allen, D., & Fraser, B. J. (2007). Parent and student perceptions of classroom learning environment and itsassociation with student outcomes. Learning Environments Research, 10(1), 67–82.

Anderson, G. J., & Walberg, H. J. (1968). Classroom climate and group learning. International Journal ofEducational Sciences, 2, 178–180.

Anderson, G. J., Walberg, H. J., & Welch, W. W. (1968). Curriculum effects on the social climate oflearning: A new representation of discriminant functions. American Educational Research Journal, 6,315–328.


123

Association to Advance Collegiate Schools of Business. (2012). Eligibility procedures and accreditationstandards for business accreditation. http://www.aacsb.edu/accreditation/standards-busn-jan2012.pdf

Bell, J. A. (1998). International students have statistics anxiety too! Education, 118, 634–636.British Academy. (2012). Society counts. http://www.britac.ac.uk/policy/Society_Counts.cfmCarlson, K. A., & Winquist, J. R. (2011). Evaluating an active learning approach to teaching introductory

statistics: A classroom workbook approach. Journal of Statistics Education, 191, 1–22.Chandra, V., & Fisher, D. L. (2009). Students’ perceptions of a blended web-based learning environment.

Learning Environments Research, 12(1), 31–44.Chang, V., & Fisher, D. L. (2003). The validation and application of a new learning environment instrument

for online learning in higher education. In M. S. Khine & D. L. Fisher (Eds.), Technology-rich learningenvironments: A future perspective (pp. 1–20). Singapore: World Scientific.

Cook, A. (2010). Improving the success rate in statistics (No. 415). Brisbane: University of Queensland,School of Economics.

Cronbach, L. J. (1951). Coefficient alpha and the internal structure tests. Psychometrika, 16, 297–334.Cruise, R. J., Cash, R. W., & Bolton, D. L. (1985, August). Development and validation of an instrument to

measure statistical anxiety. Paper presented at the annual meeting of the statistical education sec-tion. Proceedings of the American Statistical Association, Chicago, IL.

Dorman, J. P. (2008). Use of multitrait-multimethod modeling to validate actual and preferred forms of theWhat Is Happening In this Class? (WIHIC) questionnaire. Learning Environments Research, 11,179–193.

Dorman, J. P., & Fraser, B. J. (2009). Psychosocial environment and affective outcomes in technology-richclassrooms: Testing a causal model. Social Psychology of Education, 12(1), 77–99.

Finney, S. J., & Schraw, G. (2003). Self-efficacy beliefs in college statistics courses. Contemporary Edu-cational Psychology, 28(2), 161–186.

Francis, G., & Lipson, K. (2010, July). The importance of teaching statistics in a professional context. InProceedings of the eighth international conference on teaching statistics, Ljubljana, Slovenia.

Fraser, B. J. (1978). Development of a test of science-related attitudes. Science Education, 62(4), 509–515.Fraser, B. J. (1998). Classroom environment instruments: Development, validity and applications. Learning

Environments Research, 1(1), 7–34.Fraser, B. J., Fisher, D. L., & McRobbie, C. J. (1996, April). Development, validation and use of personal

and class forms of a new classroom environment instrument. Paper presented at the annual meeting ofthe American Educational Research Association, New York.

Fraser, B. J., McRobbie, C. J., & Giddings, G. J. (1993). Development and cross-national validation of alaboratory classroom environment instrument for senior high school science. Science Education, 77(1),1–24.

Fraser, B. J., & Treagust, D. F. (1986). Validity and use of an instrument for assessing classroom psy-chosocial environment in higher education. Higher Education, 15(1–2), 37–57.

Freedman, M. P. (1997). Relationship among laboratory instruction, attitude toward science, and achieve-ment in science knowledge. Journal of Research in Science Teaching, 34(4), 343–357.

Gordon, S. (2004). Understanding students’ experiences of statistics in a service course. Statistics EducationResearch Journal, 3(1), 40–59.

Green, J. J., Stone, C. C., Zegeye, A., & Charles, T. A. (2009). How much math do students need to succeedin business and economics statistics? An ordered probit analysis. Journal of Statistics Education, 17(3),1–22.

Hannula, M. S. (2002). Attitude towards mathematics: Emotions, expectations and values. Educationalstudies in Mathematics, 49(1), 25–46.

Hemmings, B., Grootenboer, P., & Kay, R. (2011). Predicting mathematics achievement: The influence ofprior achievement and attitudes. International Journal of Science and Mathematics Education, 9(3),691–705.

Hsu, M. K., Wang, S. W., & Chiu, K. K. (2009). Computer attitude, statistics anxiety and self-efficacy onstatistical software adoption behavior: An empirical study of online MBA learners. Computers inHuman Behaviour, 252, 412–420.

Jegede, O. J., Fraser, B. J., & Fisher, D. L. (1995). The development and validation of a distance and openlearning environment scale. Educational Technology Research and Development, 43(1), 89–94.

Kaiser, H. F. (1958). The varimax criterion for analytic rotation in factor analysis. Psychometrika, 23,187–200.

Kek, M., & Huijser, H. (2011). Exploring the combined relationships of student and teacher factors onlearning approaches and self-directed learning readiness at a Malaysian University. Studies in HigherEducation, 36(2), 185–208.


123

http://www.aacsb.edu/accreditation/standards-busn-jan2012.pdf

http://www.britac.ac.uk/policy/Society_Counts.cfm

Khine, M. S., & Fisher, D. L. (Eds.). (2003). Technology-rich learning environments: A future perspective.Singapore: World Scientific.

Kline, P. (1994). An easy guide to factor analysis. London: Routledge.Lalonde, R. N., & Gardner, R. C. (1993). Statistics as a second language? Model for predicting performance

in psychology students. Canadian Journal of Behavioral Science, 25, 108–125.Liu, S., Onwuegbuzie, A. J., & Meng, L. (2012). Examination of the score reliability and validity of the

Statistics Anxiety Rating Scale. The Journal of Educational Enquiry, 11(1), 29–42.Ma, X., & Kishor, N. (1997). Assessing the relationship between attitude toward mathematics and

achievement in mathematics: A meta-analysis. Journal for Research in Mathematics Education, 28(1),26–47.

McGuire, W. J. (1969). The nature of attitudes and attitude change. In G. Lindzey & E. Aronson (Eds.), Thehandbook of social psychology (pp. 136–314). Reading, MA: Addision-Wesley.

Meletiou-Mavrotheris, M., Lee, C., & Fouladi, R. T. (2007). Introductory statistics, college student attitudesand knowledge—A qualitative analysis of the impact of technology based instruction. InternationalJournal of Mathematical Education, 19, 65–83.

Newby, M. (2002). An empirical study comparing the learning environments of open and closed computerlaboratories. Journal of Information Systems Education, 13(4), 303–314.

Newby, M., & Fisher, D. L. (1997). An instrument for assessing the learning environment of a computerlaboratory. Journal of Educational Computing Research, 16, 179–190.

Newby, M., & Fisher, D. L. (2000). A model of the relationship between university computer laboratoryenvironment and student outcomes. Learning Environments Research, 3(1), 51–66.

Newby, M., & Nguyen, T. H. (2010). Using the same problem with different techniques in programmingassignments: An empirical study of its effectiveness. Journal of Information Systems Education, 21(4),375–382.

Nguyen, T. H., & Hampson, P. (2012, March). Improving students’ learning outcomes in university businessstatistics modules: Increase contact hours or increase relevance? Paper presented at the 5th AnnualNortheast Universities (3 Rivers Consortium) Regional Learning and Teaching Conference, Mid-dlesbrough, Tees Valley, UK.

Nguyen, T. H., & Ulbrich, F. (2013, August). The effectiveness of different grading strategies: An empiricalstudy. In Proceedings of the 19th Americas conference on information systems, AMCIS 2013, 15–17Chicago, IL.

Okan, Z. (2008). Computing laboratory classes as language learning environments. Learning EnvironmentsResearch, 11(1), 31–48.

Onwuegbuzie, A. J. (2000). Statistics anxiety and the role of self-perceptions. Journal of EducationalResearch, 93, 323–335.

Onwuegbuzie, A. J. (2004). Academic procrastination and statistics anxiety. Assessment & Evaluation inHigher Education, 29(1), 3–19.

Onwuegbuzie, A. J., Da Ros, D., & Ryan, J. M. (1997). The components of statistics anxiety: A phe-nomenological study. Focus on Learning Problems in Mathematics, 19(4), 11–35.

Onwuegbuzie, A. J., & Seaman, M. (1995). The effect of time and anxiety on statistics achievement. Journalof Experimental Psychology, 63, 115–124.

Onwuegbuzie, A. J., & Wilson, V. A. (2003). Statistics anxiety: Nature, etiology, antecedents, effects andtreatments—A comprehensive review of the literature. Teaching in Higher Education, 8, 195–209.

Pan, W., & Tang, M. (2005). Students’ perceptions on factors of statistics anxiety and instructionalstrategies. Journal of Instructional Psychology, 32, 205–214.

Petocz, P. (1998). Effective video-based resources for learning statistics. In Statistical education—Ex-panding the network, fifth international conference on teaching statistics (pp. 985–991).

Petocz, P., & Reid, A. (2005). Something strange and useless: Service students’ conceptions of statistics,learning statistics and using statistics in their future profession. International Journal of MathematicalEducation in Science and Technology, 36(7), 789–800.

Ramirez, C., Schau, C., & Emmioglu, E. (2012). The importance of attitudes in statistics education.Statistics Education Research Journal, 11(2), 57–71.

Ramsey, J. B. (1999, August). Why do students find statistics so difficult? In Proceedings of the 52th Sessionof the International Statistical Institute, Helsinki, Finland.

Schau, C. (2003). Students’ attitudes: The ‘‘other’’ important outcome in statistics education. In Proceedingsof the joint statistical meetings (pp. 3673–3681).

Tabachnick, B. G., & Fidell, L. S. (2010). Using multivariate statistics (6th ed.). Boston, MA: Pearson.Taylor, B. A., & Fraser, B. J. (2013). Relationships between learning environment and mathematics anxiety.

Learning Environments Research, 16, 297–313.


123

Trickett, E. J., & Moos, R. H. (1973). Social environment of junior high and high school classrooms.Journal of Educational Psychology, 65(1), 93–102.

Walberg, H. J., & Anderson, G. J. (1968). Classroom climate and individual learning. Journal of Educa-tional Psychology, 59, 414–419.

Wilensky, U. (1997). What is normal anyway? Therapy for epistemological anxiety. Educational Studies inMathematics, 33(2), 171–202.

Williams, A. S. (2010). Statistics anxiety and instructor immediacy. Journal of Statistics Education, 18,1–18.

Williams, M., Payne, G., Hodgkinson, L., & Poade, D. (2008). Does British sociology count? Sociologystudents’ attitudes toward quantitative methods. Sociology, 42, 1003–1021.

Wolf, S. J., & Fraser, B. J. (2008). Learning environment, attitudes and achievement among middle-schoolscience students using inquiry-based laboratory activities. Research in Science Education, 38(3),321–341.

Wu, W., Chang, H. P., & Guo, C. J. (2009). The development of an instrument for a technology-integratedscience learning environment. International Journal of Science and Mathematics Education, 7(1),207–233.

Zanakis, S. H., & Valenzi, E. R. (1997). Student anxiety and attitudes in business statistics. Journal ofEducation for Business, 73, 10–16.

Zeidner, M. (1991). Statistics and mathematics anxiety in social science students: Some interesting parallels.British Journal of Educational Psychology, 61, 319–328.


123

Development and use of an instrument to measure students’ perceptions of a business statistics...

Documents

Transcript of Development and use of an instrument to measure students’ perceptions of a business statistics...