Personality and Individual Di€erences 29 (2000) 841±856

www.elsevier.com/locate/paid

Relationship between Thinking Styles Inventory and Study

Process Questionnaire
Li Fang Zhang*
Department of Education, The University of Hong Kong, Pokfulam Road, Hong Kong
Received 12 May 1999; received in revised form 11 October 1999; accepted 22 October 1999

Abstract

This study investigated further the relationship between thinking styles as de®ned by Sternberg's
theory of mental self-government and learning approaches as de®ned by Biggs's model of student
learning. Participants were two independent groups of American university students …N1 ˆ 67, N2 ˆ 65).
Participants responded to the Thinking Styles Inventory and the Study Process Questionnaire. It was
found that the two inventories generally were correlated in predictable ways. This ®nding con®rmed the
one obtained in an early study of two Chinese populations. 7 2000 Elsevier Science Ltd. All rights
reserved.

Keywords: Thinking styles; Learning approaches

1. Introduction

Theories of styles ¯ourished during the late 1950s and the early 1970s. However, the
seemingly permanent research on theorization of styles subsided partially because of the
overwhelming output from the ®eld and partially because of its lack of internal dialogue
(Jones, 1997). Riding and Cheema (1991, p. 193) explained that `` . . . many researchers working
within the learning/cognitive style research, fail[ed] to mention the existence of other types of
styles''. In addition, the `style' construct, by the year 1984 when Messick discussed the nature
of cognitive styles, has been identi®ed as labelled in 19 di€erent ways and by the year 1991

* Tel.: +852-28-59-2522; fax: +852-28-59-2522.

E-mail address: lfzhang@hkucc.hku.hk (L.F. Zhang).

0191-8869/00/$ - see front matter 7 2000 Elsevier Science Ltd. All rights reserved.
PII: S 0 1 9 1 - 8 8 6 9 ( 9 9 ) 0 0 2 3 6 - 6
842 L.F. Zhang / Personality and Individual Di€erences 29 (2000) 841±856

when Riding and Cheema overviewed the `style' construct, 30 labels were found to be used in
the literature. As a result, we were left with a research ®eld that encompasses a variety of
di€erent, and yet similar constructs.
For the past decade or so, two types of e€orts have been made in the study of the
relationships among the di€erent labels for the style construct. One is to make conceptual
integration. The other is to conduct empirical studies.
Among the e€orts in conceptual integration, three stand out. The ®rst is by Curry (1983),
the second by Riding and Cheema (1991) and the third by Sternberg (1997). Curry (1983)
proposed a three-layer `onion' model of style measures. He believed that all style measures can
be grouped into three types resembling an onion. The outmost layer of the onion includes
styles that characterize an individual's instructional preference, that is, an individual's choice of
environment in which to learn (e.g. The Learning Preference Inventory, Rezler & Rezmovic,
1981). The middle layer contains style measures that assess an individual's information
processing style (e.g. The Learning Style Inventory, Kolb, 1976). The innermost layer involves
style measures that assess an individual's cognitive personality style, that is, an individual's
approaches to adapting and assimilating information (e.g. The Embedded Figures Test, Witkin,
Oltman, Raskin & Karp, 1971).
A second major attempt at integrating the work on styles was made by Riding and Cheema
(1991) who organized the style-based work along two style-dimensions and one family of
learning strategies. The ®rst is the wholist±analytic style dimension which is concerned with
whether individuals perceive a whole or separate parts (e.g. converging±diverging, Guilford,
1967; ®eld dependence±independence, Witkin, 1959, 1964). The dimension is the verbal±
imagery style dimension which pertains to whether individuals think with words or pictures
(e.g. verbalizer±visualizer, Richardson, 1977; verbalizer±imager, Riding & Taylor, 1976). The
family of learning strategies contains such work as Biggs's (1987) model of surface±deep
learning approaches and Kolb's (1977) four types of learners: divergent, convergent,
assimilating and accommodating.
The third major endeavor made to integrate models of styles was that of Sternberg (1997).
Sternberg discussed three approaches of the study of styles: cognition-centered, personality-
centered and activity-centered. Styles in the cognition-centered approach most closely resemble
abilities. Furthermore, like abilities, these styles have often been measured by tests of maximal
performance with `right' and `wrong' answers. Two models of styles that have generated the
most interest are Witkin's (1964) ®eld dependence±independence and Kagan's (1976)
re¯ection±impulsivity. The personality-centered approach views styles as closer to personality
traits. In addition, styles in this tradition are measured via typical- rather than maximum-
performance tests. Major work in this approach has been done by Myers and McCaulley
(1988, based on Jung's, 1923 work on theory of types) and Gregorc (1979) who proposed four
main types of styles, based on all possible combinations of two dimensions, concrete versus
abstract and sequential versus random. The activity-centered approach focuses on the notion
of styles as mediators of various forms of activities that tend to arise from aspects of cognition
and personality. One major group of work in this tradition is represented by similar theories of
deep- and surface-learning approaches proposed by Biggs (1979), Entwistle (1981), Marton
(1976) and Schmeck (1983). Renzulli and Smith (1978) suggested di€erent learning styles, with
 L.F. Zhang / Personality and Individual Di€erences 29 (2000) 841±856 843

each corresponding to a method of teaching such as projects, discussion and drill and
recitation.
However, as pointed out by Fine (1991) and Marton (1999, personal communication),
empirical research with the aim of clarifying the relationship between styles underlying di€erent
theories has been much limited. All studies found in the literature (through a PsycLit search)
are reviewed here. First, the studies that have been reviewed by Riding and Cheema (1991) will
be recapitulated. Second, eight individual studies will be reviewed.
In an e€ort to clarify the nature of work on styles, Riding and Cheema (1991) reviewed the
major theoretical models of styles. Apart from providing a description of the style labels
concerned, the authors discussed the relationships of a central style label under consideration
with other style labels. These relationships are shown by empirical studies. The ®rst group of
relationships is centered around the style label `impulsivity±re¯ectivity'. The second is centered
around the style label `convergent±divergent' thinking.
Riding and Cheema stated that after studying a sample of lower class children aged 3 to 6
years, Banta (1970) concluded that there was a highly signi®cant correlation between ®eld
dependence±independence and impulsivity±re¯ectivity. In a study of middle class and
disadvantaged preschoolers using the Embedded Figures Test (Witkin et al., 1971) and the
Matching Familiar Figures Test (Kagan, 1965), Schleifer and Douglas (1973) found identical
signi®cant relationships for both samples. In reviewing impulsivity±re¯ectivity style, Messer
(1976) discussed other studies that revealed a `moderate relationship' (Riding & Cheema, 1991,
p. 199) between ®eld dependence±independence and impulsivity±re¯ectivity. Re¯ective
individuals are signi®cantly more ®eld independent than are impulsive individuals (Campbell &
Douglas, 1972; Keogh & Donlon, 1972; Massari, 1975; Neimark, 1975).
Riding and Cheema (1991) also reviewed empirical studies that revealed a relationship
between convergent±divergent thinking and ®eld dependence±independence. Research indicated
that although all individuals who scored high on ®eld independence were not necessarily more
creative, those obtaining high scores on divergent thinking tended to score signi®cantly higher
on ®eld independence than did those who scored lower on divergent thinking (Bloomberg,
1971; Noppe & Gallagher, 1977; Spotts & Mackler, 1967). It was also found that individuals
who were studying in disciplines that tend to allow more creativity were more ®eld independent
than were those who were studying in disciplines that require less creativity. For example,
Bergum (1977) and Morris and Bergum (1978) reported that students of architecture rated
themselves as more creative than did business students and were also more ®eld independent
than were business students.
As shown above, empirical work reviewed by Riding and Cheema (1991) on the relationships
among the style labels was limited in one major way. That is, the studies invariably involved
Witkin's work on ®eld dependence±independence that has been proved to be essentially the
same as perceptual ability. Furthermore, all the studies were conducted more than two decades
ago. The following review, therefore, provides a more recent picture of what has been done
empirically to examine the relationships between some of the style labels.
The studies reviewed here are put into three di€erent groups based on what Sternberg (1997)
calls ``three approaches to the study of styles''. Each of the ®rst three studies (Cantwell &
Moore, 1998; Kember & Gow, 1990; Wilson, Smart & Watson, 1996) involves two theoretical
models, both using the `activity-centered approach'. Each of the next three studies (Beishuizen,
844 L.F. Zhang / Personality and Individual Di€erences 29 (2000) 841±856

Stoutjesdijk & Van-Putten, 1994; Ford, 1995; Sadler-Smith, 1997) involves two theoretical
models of two di€erent approaches, one being the `cognition-centered' and the other being the
`activity-centered'. The seventh study (Alvi, Khan, Hussain & Baig, 1988) also involves two
di€erent models, one being cognition-centered, the other being personality-centered. The last
study (Harasym, Leong, Juschka, Lucier & Lorscheider, 1996) involves two theoretical models
from the personality-centered approach.
In a study of 207 ®nal-year Diploma of Nursing Australian students, Cantwell and Moore
(1998, p. 100) identi®ed ``theoretically consistent associations'' between the scales in Biggs's
(1987) Study Process Questionnaire and those in Cantwell and Moore's (1996) Strategic
Flexibility Questionnaire. The adaptive control strategy was positively correlated with both the
deep and the achieving approaches. Similarly, a surface approach was positively correlated
with the in¯exible and irresolute control strategies.
Kember and Gow (1990) administered both the Study Process Questionnaire and the
Approaches to Studying Inventory (ASI, Entwistle, 1981) to Hong Kong university students.
Factor analysis results indicated that although all three factors (surface, deep and achieving)
appeared in the SPQ, only two factors (deep and achieving) appeared in the ASI. The surface
factor was replaced by a factor labelled `narrow orientation' (Harper & Kember, 1989), which
also has been variously called `operation learning' by Watkins (1982, p. 80) and `disorganized
study' by Ramsden and Entwistle (1981, p. 372).
A third study was conducted by Wilson et al. (1996), who also studied the relationship
between the SPQ and the ASI. Analyzing the data collected from 283 Australian university
students, the authors found signi®cant correlations between the scales in the two inventories.
They concluded that the two inventories are measuring similar constructs.
Beishuizen et al. (1994) studied the relation between cognitive levels of task accomplishment
and deep versus surface processing of learning material. These scholars expected deep-
processing students to bene®t from metacognitive support and surface-processing students to
bene®t from cognitive support. They found that students who processed at a surface level
tended to bene®t from cognitive support. Students who combined self-regulation with deep
processing and students who combined external regulation with surface processing
outperformed students who showed the opposite pairings of type of regulation with type of
processing.
In Ford's (1995) study, thirty-eight university students were tested for ®eld dependence±
independence using Riding's (1991) computer-administered Cognitive Styles Analysis (CSA).
Students were also taught the computerised version of Pask and Scott's (1972) original testing
materials designed to suit holist and serialist learning strategies. A computerised test was
utilized for assessing learning performance. It was found that students' holist and serialist
competence could be predicted by the CSA scores.
Sadler-Smith (1997) conducted a more comprehensive study of the relationships among the
style labels by including four instruments, three being activity-centered and one being
cognition-centered. Two hundred and forty-®ve university undergraduates in business studies
participated in the study. The participants responded to the Cognitive Styles Analysis (CSA,
Riding, 1991), the Learning Preferences Inventory (LPI, Riechmann & Grasha, 1974), the
Learning Styles Questionnaire (LSQ, Honey & Mumford, 1986, 1992) and the Revised
Approaches to Studying Inventory (RASI, Entwistle & Tait, 1994). After examining the
 L.F. Zhang / Personality and Individual Di€erences 29 (2000) 841±856 845

correlation coecients among the scales of the di€erent instruments, the authors concluded
that the study suggested some overlap between the dimensions measured by the LSQ and the
RASI. However, no statistically signi®cant relationships were identi®ed between cognitive styles
(as measured by the CSA) and any of the other `style' constructs investigated.
A study conducted by Alvi et al. (1988) focused on the relationship between Holland's
(1973) personality types as assessed by the Self-Directed Search and Witkin et al.'s ®eld
dependence±independence as assessed by the Group Embedded Figures Test (GEFT, Oltman,
Raskin & Witkin, 1971). Participants were 179 Canadian senior high school students and 376
Pakistani college and university students. Similar results were observed for the Canadian and
Pakistani samples: students with two- or three-letter codes consisting of R (realistic), I
(investigative) and A (artistic) in any order, obtained higher GEFT scores than did those with
two- or three-letter codes composed of S (social), E (enterprising) and C (conventional).
Harasym et al. (1996) investigated the relationship between the Myers±Briggs Type Indicator
(MBTI, Myers & McCaulley, 1988) and the Gregorc Style Delineator (Gregorc, 1982). It was
found that each learning style assessed by the Gregorc Style Delineator corresponded to certain
traits assessed by the MBTI. For example, individuals who scored higher on the concrete
sequential learning style scale tended to have traits of sensing and judging on the MBTI,
whereas individuals who scored higher on the concrete random learning style scale tended to
have the traits of intuition and perceiving on the MBTI.
As shown in the above review, recent empirical studies on the relationships among the `style'
constructs are clearly moving toward being based on theoretical models that have more
implications for the teaching±learning process. Moreover, we begin to see studies (e.g. Ford,
1995) that are based on a model of styles, which is more general (e.g. Riding, 1991), rather
than a model that addresses only one dimension of cognitive styles, for which scores often are
dichotomized (e.g. Witkin, Dyk, Faterson, Goodenough & Karp, 1962).
The current study is based on two theoretical models. One is Sternberg's (1988, 1990, 1994,
1997) theory of mental self-government; the other is Biggs's (1987, 1992) model of student
learning approaches.
In an e€ort to resurrect previous work in the ®eld of styles, Sternberg (1988, 1990, 1994,
1997) constructed a more comprehensive theory of thinking styles, the theory of mental self-
government. Sternberg's theory addresses people's thinking styles, which may be used in many
settings, including university, home and community. At the heart of this theory is the notion
that people need somehow to govern or manage their everyday activities. There are many ways
of doing so; whenever possible, people choose styles of managing themselves with which they
are comfortable. Still, people are at least somewhat ¯exible in their use of styles and try with
varying degrees of success to adapt themselves to the stylistic demands of a given situation.
Thus, an individual with one preference in one situation may have a di€erent preference in
another situation. Moreover, styles may change with time and with life demands. Thinking
styles are at least in part socialized (Sternberg, 1994, 1997), suggesting that, to some extent,
they can be modi®ed by the environment in which people reside. The theory of mental self-
government describes 13 thinking styles that fall along ®ve dimensions of mental self-
government: (a) functions (including the legislative, executive and judicial styles), (b) forms
(including the monarchic, hierarchic, oligarchic and anarchic styles), (c) levels (including the
global and local styles), (d) scopes (including the internal and external styles) and (e) leanings
846 L.F. Zhang / Personality and Individual Di€erences 29 (2000) 841±856

(including the liberal and conservative styles) of government as applied to individuals. One key
characteristic of each of the 13 thinking styles is presented in Appendix A (for details, see
Sternberg, 1988, 1990, 1994, 1997).
The theory of mental self-government possesses several characteristics. The styles it speci®es
fall along ®ve dimensions, rather than along one. Second, styles are viewed as falling along
continua rather than as being dichotomous. Third, styles are not viewed as `good' or `bad' in
themselves. The utility of a style for a person interacts with the task the person is performing
and the situation in which the task is performed. Finally, the theory of mental self-government
yields a pro®le of styles for each individual, rather than merely the identi®cation of a single
style.
The theory of mental self-government has been operationalized through several instruments,
including the Thinking Styles Inventory (TSI, Sternberg & Wagner, 1992). The TSI is a self-
report measure in which respondents rate themselves on a 7-point scale ranging from 1 (low) to
7 (high) on a number of preferences. Examples of items from the inventory are: (1) `I like tasks
that allow me to do things my own way' (legislative), (2) `I like situations in which it is clear
what role I must play or in what way I should participate' (executive) and (3) `I like to
evaluate and compare di€erent points of view on issues that interest me' (judicial). The TSI
went through a translation and back-translation procedure between Chinese and English in
1996.
Results indicated that the Thinking Styles Inventory is a reliable and valid measure for
assessing the thinking styles proposed in the theory of mental self-government among
students and teachers in Hong Kong, mainland China, as well as in the United States.
The internal-consistency reliabilities of the scales are generally satisfactory, ranging from
the high '50s to the low '80s. Factor analyses were generally, although not completely,
supportive of the structure of the theory. In two of the studies, ®ve factors accounted for
77 (Sternberg, 1994) and 78% (Zhang, 1999) of the variance, respectively. The remaining
studies indicated that the TSI scales are correlated with one another in the expected
directions, with most of the correlation coecients being statistically signi®cant.
The theory also has showed good external validity. Sternberg (1994) has also examined the
correlates of the TSI with other tests, both of styles and of abilities. For example, with the
Myers±Briggs Type Indicator, 30 of 128 correlations were statistically signi®cant, whereas with
Gregorc's measure, 22 of 52 were signi®cant. These correlations go well beyond the levels that
would be expected by chance, thus suggesting that the various style measures partition a
similar space of the intelligence±personality interface, but in di€erent ways (Sternberg, 1997).
On the contrary, the correlation of the measure of mental self-government with IQ was not
signi®cant, nor was the correlation with grade point average overall. Thus, ``styles are not a
proxy for intelligence'' (Sternberg, 1998, personal communication).
Early work in learning approaches can be found in the frequently cited deep±surface
distinction in approaches to learning initiated by Marton (1976). To this early work, Biggs
(1987), Entwistle (1981) and Entwistle and Ramsden (1983) each added a third approach. Biggs
called the third approach the `achieving' approach whereas Entwistle and Ramsden described it
as the `strategic' approach. According to Biggs, there are three common approaches to
learning: surface, which involves a reproduction of what is taught to meet the minimum
 L.F. Zhang / Personality and Individual Di€erences 29 (2000) 841±856 847

requirements; deep, which involves a real understanding of what is learned and achieving,
which involves using such a strategy that would maximize one's grades.
One of the measures used to assess university students' learning approaches is the Study
Process Questionnaire (SPQ, Biggs, 1987, 1992). The SPQ is a self-report test composed of 42
items falling into 6 subscales. For each item, the respondents rate themselves on a 5-point scale
ranging from 1 (low) to 5 (high). The 6 subscales are: surface-motive, surface-strategy, deep-
motive, deep-strategy, achieving-motive, and achieving-strategy. Examples of items from the
questionnaire are: (1) `even when I have studied hard for a test, I worry that I may not be able
to do well on it' (surface motive), (2) `while I am studying, I think of real life situations to
which the material that I am learning would be useful' (deep strategy) and (3) `I summarize
suggested readings and include these as part of my notes on a topic' (achieving strategy). The
three overall scales are surface, deep and achieving. The SPQ was originally designed to assess
the learning approaches of Canadian and Australian students. In 1992, the SPQ was translated
and back-translated between Chinese and English and the Hong Kong norms were established
(Biggs, 1992). The psychometric properties of the SPQ scales have been summarized by a few
researchers, including Albaili (1995), Watkins (1998) and Zhang and Sternberg (in press).
These researchers reported that the SPQ is a reliable and valid instrument for assessing
university students' learning approaches in practically all the cultures in which the SPQ has
been used.
There are two types of validity evidence for the SPQ, the internal and the external. The
internal validity is obtained through examining the internal structure of the instrument.
Whereas some studies supported Biggs's original argument that the SPQ assesses three
approaches to learning (surface, deep and achieving, e.g. Bolen, Wurm & Hall, 1994; O'Neil &
Child, 1984), other studies supported a two-factor (surface and deep) model (e.g. Niles, 1995;
Watkins & Dahlin, 1997). The two-factor model is consistent with the model proposed by
Marton (1976) who used a phenomenographic method in studying students' learning
approaches.
The external validity is obtained through examining the SPQ against other instruments
assumed to be based on similar constructs with the SPQ. It has been identi®ed that the SPQ
assesses similar constructs as Entwistle's (1981) Approaches to Studying Inventory (Wilson et
al., 1996) and Cantwell and Moore's (1996) Strategic Flexibility Questionnaire (Cantwell &
Moore, 1998). As discussed earlier, Kember and Gow's (1990) factor analysis of the SPQ and
of the Approaches to Studying Inventory yielded two factorial structures that are slightly
di€erent. Detailed descriptions of the nature of the SPQ can be found in Biggs's, 1993 work.
Both the Thinking Styles Inventory and the Study Process Questionnaire have been assessed
regarding their heuristic values in educational settings cross-culturally. Majoring ®ndings
regarding the Thinking Styles Inventory can be found in Grigorenko and Sternberg (1997),
Sternberg and Grigorenko (1995), Zhang (1999), Zhang and Sachs (1997) and Zhang and
Sternberg (1998). Results from these studies indicate the following statistically signi®cant
relationships: (1) teachers' thinking styles with teacher characteristics (e.g. length of teaching
experience, subject areas taught), (2) teachers' thinking styles with school environment, (3)
students' thinking styles with student characteristics (e.g. age, sex, socio-economic status, birth
order, ®eld of study, leadership experience, work and travelling experience), (4) match and
848 L.F. Zhang / Personality and Individual Di€erences 29 (2000) 841±856

mismatch of teachers' and students' thinking styles with student academic achievement and (5)
students' thinking styles and academic achievement beyond abilities.
The value of the Study Process Questionnaire also has been assessed in educational settings
of many cultures. Considerable work has been done to investigate the impact of student
characteristics and learning context on the learning approaches that students take (e.g. Biggs,
1988; Sadler-Smith & Tsang, 1998; Watkins & Hattie, 1981). Meanwhile, a great deal of work
has focused on the relationship between students' learning approaches and their academic
achievement (e.g. Albaili, 1997; Biggs, 1988; Watkins, 1998; Zhang, in press).
In summary, the constructs of both the Thinking Styles Inventory and the Study Process
Questionnaire have been well researched. Furthermore, both the theory of mental self-
government and the theory of learning approaches have signi®cant implications for the
teaching±learning process. However, there is a lack of study of the relationship between the
constructs from the two inventories. One such study was conducted by Zhang and Sternberg
(in press). In their study of the relationship between the Thinking Styles Inventory and the
Study Process Questionnaire among 854 Hong Kong university students and 215 mainland
China students, Zhang and Sternberg found that the scales in the two inventories were
correlated in predictable ways. The authors concluded that the two inventories measure similar
constructs.
The current study aimed at verifying further the nature of the relationship between the two
inventories among two samples of students from a Western culture, in this case, the United
States. It was anticipated that the relationships between the two inventories identi®ed in Zhang
and Sternberg's (in press) study would hold in the current study.
Based in part on the work of Beishuizen et al. (1994) and on a previous study of the
relationship between the TSI and SPQ scales, two hypotheses were made. First, the surface
approach (either motive or strategy) to learning should be signi®cantly positively related to the
use of thinking styles that involve less complexity (e.g. executive, monarchic and conservative).
In contrast, the surface learning approach should be signi®cantly negatively related to the use
of thinking styles that involve more complexity (e.g. legislative, judicial, liberal and
hierarchical). Second, the deep approach should be signi®cantly positively correlated with the
use of thinking styles that involve more complexity (e.g. legislative, judicial, liberal and
hierarchical). On the contrary, the deep learning approach should be signi®cantly negatively
correlated with the use of thinking styles that involve less complexity (e.g. executive, monarchic
and conservative).

2. Method

2.1. Participants

Data were collected from two groups of American university students. In 1997, 67 (19 male
and 48 female; 33 sophomore, 21 junior and 13 senior) students studying in an introductory
psychology class from a mid-west Big Ten university participated in the study. The ages of the
participants ranged from 19 to 27, with a mean of 21. In 1998, the study was replicated among
a di€erent cohort group of students registered in the same class in the same university.
 L.F. Zhang / Personality and Individual Di€erences 29 (2000) 841±856 849

Participants were 65 (14 male and 51 female; 32 sophomore, 23 junior, 8 senior and 2
graduate) students. The ages of the participants ranged from 19 to 42, with a mean of 21.

2.2. Measures

The Study Process Questionnaire (SPQ, Biggs, 1987) and the Thinking Styles Inventory
(Sternberg & Wagner, 1992) were administered to participants in a normal class period.

3. Results

3.1. Reliability

For both groups, the reliability coecients of the TSI scales were in line with those in
previous studies (e.g. Sternberg, 1994; Zhang, 1999). The scale internal correlation
coecients for the 1997 group (hereafter called group one) ranged from 0.51 (oligarchic
scale) to 0.86 (liberal scale), with a median of 0.71 (judicial scale) and a mean of 0.71. The
ones for the 1998 group (hereafter called group two) ranged from 0.48 (anarchic scale) to
0.86 (conservative scale), with a median of 0.75 (judicial and external scales) and a mean of
0.72. By the same token, the internal scale correlation coecients of the SPQ scales are
also in line with those reported in previous studies (e.g. Biggs, 1992; Albaili, 1995; Watkins,
1998). The scale reliabilities ranged from 0.59 (surface motive) to 0.78 (achieving strategy),
with a median of 0.68 and a mean of 0.69 for group one. The ones for group two ranged
from 0.63 (deep motive) to 0.82 (deep strategy), with a median of 0.72 and a mean of
0.73.

Table 1
Oblimin-rotated two-factor model for the Study Process Questionnaire …N1997 ˆ 67, N1998 ˆ 65)a

Subscale 1997 1998

factor 1 factor 2 factor 1 factor 2

Surface motive 0.76 0.87

Surface strategy 0.82 0.89
Deep motive 0.89 0.82
Deep strategy 0.76 0.88
Achieving motive 0.66 0.48 0.61 0.79
Achieving strategy 0.82 0.84
% of variance 41.6 27.0 45.5 29.1
Cumulative % 41.6 68.6 45.5 74.6
Eigenvalue 2.50 1.62 2.73 1.75
a
Variables with factor loadings of less than |0.30| have been omitted.
850 L.F. Zhang / Personality and Individual Di€erences 29 (2000) 841±856

3.2. Validity

Table 1 presents the results obtained from a principal-axis factor analysis, with an
Oblimin rotation for each of the two samples. Based on Biggs's original theory, there
should be three factor solutions, with each of the three learning approaches (deep, surface,
achieving) being one of the three factors. However, many studies (e.g. Bolen et al., 1994;
Watkins & Dahlin, 1997; Zhang & Sternberg, in press) suggested two factor solutions (deep
and surface) from the SPQ. The achieving subscales were usually found to either load on
one of the other two factors (often the deep approach) or be divided between the other
two factors. In the current study, a two-factor model resulted for both samples. For the
1997 sample, both achieving subscales loaded on the deep-approach factor. For the 1998
sample, the two achieving subscales were divided between the deep- and surface-approach
factors. The two factors explained 69% and 75% of the variance in the 1997 and 1998
samples, respectively. These results indicated good validity of the SPQ for measuring
students' learning approaches based on Biggs's theory.

3.3. Relationships between TSI and SPQ

As shown in Table 2, expected statistically signi®cant correlations are observed between

many scales from the two inventories for both samples. There were two distinct patterns of
correlations. First, the motive and strategy scales within the deep approach and within the
surface approach tended to be correlated with a particular thinking style in a similar fashion.
For example, for group one, both the deep motive …r ˆ 0:41, p < 0.01) and deep strategy
…r ˆ 0:39, p < 0.01) scales were signi®cantly positively correlated with the legislative thinking
style. By the same token, for group two, both the deep motive …r ˆ 0:49, p < 0.01) and deep
strategy …r ˆ 0:48, p < 0.01) scales were signi®cantly positively correlated with the legislative
thinking style. Complementarily, both the surface motive and strategy scales were negatively
correlated with the legislative style (although not statistically signi®cant) for both groups (see
Table 2).
The second pattern of correlations is that the two achieving scales (motive and strategy)
were individually correlated with thinking styles that are theoretically the opposite of each
other. That is, the achieving motive tended to be signi®cantly positively correlated with
thinking styles that involve less complexity (e.g. executive and conservative styles); the
achieving strategy, on the other hand, tended to be signi®cantly correlated with thinking
styles that involve more complexity (e.g. legislative and liberal styles). For instance, for
group one, the correlation coecient of the legislative thinking style was 0.16 (non-
signi®cant) with the achieving motive scale and 0.27 ( p < 0.05) with the achieving strategy
scale. Similarly, for group two, the correlation coecient of the legislative style was 0.23 (non-
signi®cant) with the achieving motive scale and 0.30 ( p < 0.05) with the achieving strategy.
However, compared with the legislative style, the conservative style had a reversed pattern
of correlations with the achieving scales. Speci®cally, for group one, the correlation
coecient of the conservative style was 0.36 ( p < 0.01) with the achieving motive scale and
0.24 (non-signi®cant) with the achieving strategy scale. By the same token, for group two, the
 L.F. Zhang / Personality and Individual Di€erences 29 (2000) 841±856 851

Table 2
Pearson correlation matrix for the subscales in the Study Process Questionnaire and Thinking Styles Inventory
…N1 ˆ 67; N2 ˆ 65)a

Subscale AM AM AS AS DM DM DS DS SM SM SS SS
group one two one two one two one two one two one two

Legislative 0.16 0.23 0.27b 0.30b 0.41c 0.49c 0.39c 0.48c ÿ0.09 ÿ0.19 ÿ0.13 ÿ0.10
Executive 0.43c 0.28b 0.26b 0.14 0.20 0.04 ÿ0.01 0.03 0.20 0.44c 0.43c 0.39c
Judicial 0.33c 0.20 0.47c 0.20 0.42c 0.48c 0.33c 0.36c ÿ0.09 ÿ0.09 0.05 ÿ0.03
Global 0.37c 0.14 0.11 0.06 0.33c 0.20 0.19 0.18 0.12 0.04 0.10 0.00
Local 0.18 0.27b 0.38c 0.45c 0.02 0.40c ÿ0.04 0.43c 0.00 0.16 0.29b 0.10
Liberal 0.17 0.01 0.26b 0.37c 0.41c 0.50c 0.37c 0.45c 0.01 ÿ0.24 ÿ0.22 ÿ0.26b
Conservative 0.36c 0.23 0.24 0.02 0.15 ÿ0.04 ÿ0.07 ÿ0.05 0.22 0.39c 0.41c 0.42c
Hierarchical 0.34c 0.38c 0.44c 0.47c 0.44c 0.49c 0.34c 0.52c ÿ0.07 ÿ0.08 0.08 ÿ0.10
Monarchic 0.30b 0.43c 0.13 0.32b 0.14 0.20 0.05 0.15 0.20 0.17 0.18 0.34c
Oligarchic 0.36c 0.19 0.17 ÿ0.02 0.16 0.14 0.06 0.09 0.12 0.29b 0.46c 0.27b
Anarchic 0.31b 0.21 0.45c 0.36c 0.46c 0.45c 0.38c 0.50c ÿ0.11 ÿ0.17 0.02 -0.03
Internal 0.29b 0.14 0.15 0.12 0.33c 0.22 0.22 0.12 ÿ0.26b ÿ0.01 ÿ0.12 0.11
External 0.14 0.09 0.44c 0.20 0.28b 0.43c 0.10 0.45c 0.12 ÿ0.01 0.07 ÿ0.06
a
SM=surface motivation, DM=deep motivation, AM=achieving motivation, SS=surface strategy, DS=deep
strategy, AS=achieving strategy.
b
Indicates that correlation is statistically di€erent from zero at 0.05 level.
c
Indicates that correlation is statistically di€erent from zero at 0.01 level.

correlation coecient of the conservative style was 0.23 (non-signi®cant) and 0.02 (non-
signi®cant).
It should be noted that no speci®c hypothesis was made regarding the pattern of relationship
between the achieving scales and particular thinking styles. Previous studies of the internal
structure of the SPQ have indicated that the achieving approach did not stand as a separate
factor. Instead, the achieving motive and strategy scales often were either split between the
surface and deep scales or clustered with one of the other two factors (e.g. Watkins & Dahlin,
1997; Wong, Lin & Watkins, 1996; Zhang, in press). The current study con®rmed these
previous ®ndings because the two achieving scales were correlated with thinking styles that are
directly opposite of each other. One possible explanation of this ®nding is that achieving
motive and achieving strategy are two separate constructs. Thus, individuals whose motive is to
achieve were more eager to follow instructions (executive style) and be conservative when they
carry out certain tasks. On the other hand, the use of an achieving strategy may simply be a
manifestation of being self-directed (legislative style) and of not worrying about rules (liberal
style).

4. Discussion

The primary objective was to examine the relationship between students' reported thinking
852 L.F. Zhang / Personality and Individual Di€erences 29 (2000) 841±856

styles and their reported learning approaches. Results indicated that there were clear and
consistent associations between particular TSI and SPQ scales. It was evident that thinking
styles that require more complexity (legislative, judicial, liberal and hierarchical) were
signi®cantly positively related to the deep approach scales, but negatively related to the surface
approach scales. It was also obvious that thinking styles that require less complexity (executive
and conservative) were signi®cantly positively related to the surface approach scales, but
negatively related to or had little relationship with the deep approach scales. These results
indicated that the two inventories overlap in at least one dimension underlying their respective
theories. That is, a continuum of complexity involved in the process of perceiving and
performing tasks.
These results not only fully supported the hypotheses made, but also con®rmed the ®ndings
of a previous study (Zhang & Sternberg, in press) that used the same two inventories.
Moreover, compared with the results from Zhang and Sternberg's (in press) study, the
magnitudes of the correlation coecients of the current study are stronger. Moreover, these
results also supported previous studies reviewed earlier that studied the relationships between
style constructs proposed by di€erent theories (e.g. Alvi et al., 1988; Beishuizen et al., 1994;
Ford, 1995; Wilson et al., 1996).
The correlations found in the current study also make substantive sense. Both theories
describe the manners in which tasks are perceived and performed. For example, someone
who is motivated to understand the learning material and use a deep strategy to learn
would inevitably use his or her abilities in a creative and non-traditional way. Vice versa,
someone who prefers to use the legislative or liberal thinking style would automatically
take a deep approach to learning (also see Sadler-Smith, 1996). Thus, the thinking styles
and learning approaches under consideration seem to have an overlapping conceptual role.
The overlapping conceptual role of thinking styles and learning approaches coupled with
the similar inference with regard to the empirical data encourage the author to conclude
that the two labels attached to the two theories may be considered as two similar
constructs.
In conclusion, the current study has made one major contribution. That is, through
providing empirical data, the study enhanced the understanding of the relationships between
two theories of `style' constructs. It was found that the Thinking Style Inventory and the Study
Process Questionnaire overlap in measuring at least one construct, that is, the degree of
complexity involved in task perception and performance.

Acknowledgements

Research for this project was supported by the Committee on Research and Conference
Grants as administered by The University of Hong Kong.

Appendix A

Sample items from the Thinking Styles Inventory are shown in Table 3.
 L.F. Zhang / Personality and Individual Di€erences 29 (2000) 841±856 853

Table 3
Sample items from the Thinking Styles Inventory

Sample items Scale type Key characteristic

I like tasks that allow me to do things my own way. legislative being creative
I like situations in which it is clear what role I must play executive being conforming
or in what way I should participate.
I like to evaluate and compare di€erent points of view on judicial being analytical
issues that interest me.
I like to complete what I am doing before starting monarchic dealing with one task at a time
something else.
When undertaking some task, I like ®rst to come up with hierarchic dealing with multiple prioritized tasks
a list of things that the task will require me to do and to
assign an order of priority to the items on the list.
I usually know what things need to be done, but I oligarchic dealing with multiple non-prioritized tasks
sometimes have trouble deciding in what order to do
them.
When working on a written project, I usually let my mind anarchic dealing with tasks at random
wander and my pen follow up on whatever thoughts cross
my mind.
Usually when I make a decision, I don't pay much global focusing on abstract ideas
attention to details.
I like problems that require engagement with details. local focusing on concrete ideas
I like to be alone when working on a problem. internal enjoying working independently
I like to work with others rather than by myself. external enjoying working in groups
I like to do things in new ways, even if I am not sure they liberal using new ways to deal with tasks
are the best ways.
In my work, I like to keep close to what has been done conservative using traditional ways to deal with tasks
before.

References

Albaili, M. A. (1995). An Arabic version of the Study Process Questionnaire: reliability and validity. Psychological
Reports, 77(3), 1083±1089.
Albaili, M. A. (1997). Di€erences among low-, average- and high-achieving college students on learning and study
strategies. Educational Psychology, 17(1 and 2), 171±177.
Alvi, S. A., Khan, S. B., Hussain, M. A., & Baig, T. (1988). Relationship between Holland's typology and cognitive
styles. International Journal of Psychology, 23, 449±459.
Banta, T. J. (1970). Tests for the evaluation of early childhood education: the Cincinnati Autonomy Test Battery
(CATB). In J. Hellmuth, Cognitive studies, 1. New York: Brunner-Mazel.
Beishuizen, J., Stoutjesdijk, E., & Van-Putten, K. (1994). Studying textbooks: e€ects of learning styles, study tasks,
and instructions. Learning and Instruction, 4, 151±174.
Bergum, B. O. (1977). Undergraduate self-perceptions of creativity and independence. Perceptual and Motor Skills,
44, 187±190.
Biggs, J. B. (1979). Individual di€erences in study processes and the quality of learning outcomes. Higher Education,
8, 381±394.
Biggs, J. B. (1987). Student approaches to learning and studying. Hawthorn: Australian Council for Educational
Research.
Biggs, J. B. (1988). Assessing student approaches to learning. Australian Psychologist, 23(2), 197±206.
854 L.F. Zhang / Personality and Individual Di€erences 29 (2000) 841±856

Biggs, J. B. (1992). Why and how do Hong Kong students learn? Using the Learning and Study Process
Questionnaires. Education Paper No. 14, Faculty of Education, The University of Hong Kong.
Biggs, J. B. (1993). What do inventories of students' learning processes really measure? A theoretical review and
clari®cation. British Journal of Educational Psychology, 63, 3±19.
Bloomberg, M. (1971). Creativity as related to ®eld independence and mobility. Journal of Genetic Psychology, 118,
3±12.
Bolen, L. M., Wurm, T. R., & Hall, C. W. (1994). Factorial structure of the Study Process Questionnaire.
Psychological Reports, 75(3), 1235±1241.
Campbell, S. B., & Douglas, V. I. (1972). Cognitive styles and responses to the threat of frustration. Canadian
Journal of Behavioral Science, 4, 30±42.
Cantwell, R. H., & Moore, P. J. (1996). The development of measures of individual di€erences in self-regulatory
control and their relationship to academic performance. Contemporary Educational Psychology, 21, 500±517.
Cantwell, R. H., & Moore, P. J. (1998). Relationships among control beliefs, approaches to learning, and the
academic performance of ®nal-year nurses. The Alberta Journal of Educational Research, 44(1), 98±102.
Curry, L. (1983). An organization of learning styles theory and constructs. ERIC Document 235, 185.
Entwistle, N. J. (1981). Styles of teaching and learning: an integrated outline of educational psychology for students,
teachers, and lecturers. Chichester: Wiley.
Entwistle, N. J., & Ramsden, P. (1983). Understanding student learning. London: Croom Helm.
Entwistle, N. J., & Tait, H. (1994). The revised approaches to studying inventory. University of Edinburgh, Center for
Research into Learning and Instruction.
Fine, B. J. (1991). Field-dependence and extraversion: univariate or multivariate research orientation. Perceptual and
Motor Skills, 72, 1044±1046.
Ford, N. (1995). Levels and types of mediation in instructional systems: an individual di€erences approach.
International Journal of Human-Computer Studies, 43, 241±259.
Gregorc, A. F. (1979). Learning/teaching styles: potent forces behind them. Educational Leadership, 36, 234±236.
Gregorc, A. F. (1982). Gregorc Style Delineator. Maynard, MA: Gabriel Systems.
Grigorenko, E. L., & Sternberg, R. J. (1997). Styles of thinking, abilities, and academic performance. Exceptional
Children, 63(3), 295±312.
Guilford, J. P. (1967). The nature of human intelligence. New York, NY: McGraw-Hill.
Harasym, P. H., Leong, E. J., Juschka, B. B., Lucier, G. E., & Lorscheider, F. L. (1996). Relationship between
Myer±Briggs Type Indicator and Gregorc Style Delineator. Perceptual and Motor Skills, 82, 1203±1210.
Harper, G., & Kember, D. (1989). Interpretation of factor analyses from the approaches to studying inventory.
British Journal of Educational Psychology, 59, 66±74.
Holland, J. L. (1973). Making vocational choices: a theory of careers. Englewood Cli€s, NJ: Prentice-Hall.
Honey, P. & Mumford, A. (1986 and 1992). The manual of learning styles. Maidenhead: Peter Honey.
Jones, A. E. (1997). Re¯ection±impulsivity and wholist±analytic: two ¯edglings? . . . or is R±I a cuckoo? Educational
Psychology, 17(1 and 2), 65±77.
Jung, C. (1923). Psychological types. New York: Harcourt Brace.
Kagan, J. (1965). Matching familiar ®gures test. Cambridge, MA: J. Kagan, Harvard University.
Kagan, J. (1976). Commentary on re¯ective and impulsive children: strategies of information processing underlying
di€erences in problem solving. Monographs of the society for research in child development, 41(5) (Serial No. 168).
Kember, D., & Gow, L. (1991). A challenge to the anecdotal stereotype of the Asian students. Studies in Higher
Education, 16(2), 117±128.
Keogh, B. K., & Donlon, G. (1972). Field dependence, impulsivity and learning disabilities. Journal of Learning
Disabilities, 5, 331±336.
Kolb, D. A. (1976). The learning style inventory: technical manual. Boston, MA: McBer.
Kolb, D. A. (1977). The learning styles inventory: a self description of preferred learning modes. Boston, MA: McBer.
Marton, F. (1976). What does it take to learn? Some implications on an alternative view of learning. In N. J.
Entwistle, Strategies for research and development in higher education (pp. 200±222). Amsterdam: Swets and
Zeitlenger.
Massari, D. J. (1975). The relation of re¯ection±impulsivity to ®eld-dependence±independence and internal±external
control in children. Journal of Genetic Psychology, 126, 61±67.
 L.F. Zhang / Personality and Individual Di€erences 29 (2000) 841±856 855

Messer, S. B. (1976). Re¯ectivity±impulsivity: a review. Psychological Bulletin, 83, 1026±1053.

Morris, T. L., & Bergum, B. O. (1978). A note on the relationship between ®eld-independence and creativity.
Perceptual and Motor Skills, 46, 1114.
Myers, I. B., & McCaulley, M. H. (1988). Manual: a guide to the development and use of the Myers±Briggs type
indicator. Palo Alto, CA: Consulting Psychologists Press.
Neimark, E. D. (1975). Individual di€erences and the role of cognitive style in cognitive development. Genetic
Psychology Monographs, 91, 171±225.
Niles, F. S. (1995). Cultural di€erences in learning motivation and learning strategies: a comparison of overseas and
Australian students at an Australian university. International Journal of Intercultural Relations, 19(3), 369±385.
Noppe, L. D., & Gallagher, J. M. (1977). A cognitive style approach to creative thought. Journal of Personality
Assessment, 41, 85±90.
Oltman, P. K., Raskin, E., & Witkin, H. A. (1971). Group Embedded Figures Test. Palo Alto, CA: Consulting
Psychologists Press.
O'Neil, M. J., & Child, D. (1984). Biggs' SPQ: a British study of its internal structure. British Journal of Educational
Psychology, 54, 228±234.
Pask, G., & Scott, B. C. E. (1972). Learning strategies and individual competence. International Journal of Man±
Machine Studies, 4, 242±253.
Ramsden, P., & Entwistle, N. J. (1981). E€ects of academic departments on students' approaches to studying.
British Journal of Educational Psychology, 51, 368±383.
Renzulli, J. S., & Smith, L. H. (1978). Learning styles inventory. Mans®eld Center, CT: Creative Learning Press.
Rezler, A. G., & Rezmovic, V. (1981). The learning preference inventory. Journal of Applied Health, 10, 109±126.
Richardson, A. (1977). Verbaliser±visualiser: a cognitive style dimension. Journal of Mental Imagery, 1, 109±126.
Riding, R. (1991). Cognitive style analysis user manual. Birmingham: Learning and Training Technology.
Riding, R., & Cheema, I. (1991). Cognitive styles: an overview and integration. Educational Psychology, 11(3 and 4),
193±215.
Riding, R. J., & Taylor, E. M. (1976). Imagery performance and prose comprehension in 7 year old children.
Educational Studies, 2, 21±27.
Riechmann, S. W., & Grasha, A. F. (1974). A rational approach to developing and assessing the construct validity
of a student learning styles scale instrument. Journal of Psychology, 87, 213±223.
Sadler-Smith, E. (1996). Approaches to studying: age, gender and academic performance. Educational Studies, 22(3),
367±379.
Sadler-Smith, E. (1997). `Learning style': frameworks and instruments. Educational Psychology, 17(1 and 2), 51±63.
Sadler-Smith, E., & Tsang, F. (1998). A comparative study of approaches to studying in Hong Kong and the
United Kingdom. British Journal of Educational Psychology, 68, 81±93.
Schleifer, M., & Douglas, V. I. (1973). Moral judgments, behavior and cognitive style in young children. Canadian
Journal of Behavioral Science, 5, 133±144.
Schmeck, R. R. (1983). Learning style of college students. In R. F. Dillon, & R. R. Schmeck, Individual di€erences
in cognition, (Vol. 1) (pp. 233±279). New York: Academic Press.
Spotts, J. V., & Mackler, B. (1967). Relationships of ®eld-dependent and ®eld-independent cognitive test
performance. Perceptual and Motor Skills, 24, 239±268.
Sternberg, R. J. (1988). Mental self-government: a theory of intellectual styles and their development. Human
Development, 31, 197±224.
Sternberg, R. J. (1990). Metaphors of mind: conceptions of the nature of intelligence. New York: Cambridge
University Press.
Sternberg, R. J. (1994). Thinking styles: theory and assessment at the interface between intelligence and personality.
In R. J. Sternberg, & P. Ruzgis, Intelligence and personality (pp. 169±187). New York: Cambridge University
Press.
Sternberg, R. J. (1997). Thinking styles. New York: Cambridge University Press.
Sternberg, R. J., & Grigorenko, E. L. (1995). Styles of thinking in the school. European Journal for High Ability, 6,
201±219.
Sternberg, R. J. & Wagner, R. K. (1992). Thinking Styles Inventory. Unpublished test, Yale University.
856 L.F. Zhang / Personality and Individual Di€erences 29 (2000) 841±856

Watkins, D. (1982). Identifying the study process dimensions of Australian university students. The Australian
Journal of Education, 26(1), 76±85.
Watkins, D. (1998). approaches to learning: a cross-cultural perspective on the Study Process Questionnaire. In B.
Dart, & G. Boulton-Lewis, Teaching and learning in higher education (pp. 124±144). Melbourne: Australian
Council for Educational Research.
Watkins, D. A., & Dahlin, B. (1997). Assessing study approaches in Sweden. Psychological Reports, 81, 131±136.
Watkins, D., & Hattie, J. (1981). The learning processes of Australian university students: investigations of
contextual and personological factors. British Journal of Educational Psychology, 51, 384±393.
Wilson, K. L., Smart, R. M., & Watson, R. J. (1996). Gender di€erences in approaches to learning in ®rst year
psychology students. British Journal of Educational Psychology, 66, 59±71.
Witkin, H. A. (1959). The perception of the upright. Science American, 200, 50±56.
Witkin, H. A. (1964). Origins of cognitive style. In C. Sheerer, Cognition, theory, research, promise. New York, NY:
Harper and Row.
Witkin, H. A., Dyk, R. B., Faterson, H. F., Goodenough, D. R., & Karp, S. A. (1962). Psychological di€erentiation.
New York, NY: Wiley.
Witkin, H. A., Oltman, P. K., Raskin, E., & Karp, S. A. (1971). Embedded Figures Test, Children's Embedded
Figures Test: manual. Palo Alto, CA: Consulting Psychologists Press.
Wong, N. Y., Lin, W. Y., & Watkins, D. (1996). Cross-cultural validation of models of student learning.
Educational Psychology, 16, 317±327.
Zhang, L. F. (1999). Further cross-cultural validation of the theory of mental self-government. Journal of
Psychology, 133(2), 165±181.
Zhang, L. F. (in press). University students' learning approaches in three cultures: an investigation of Biggs's 3P
model. Journal of Psychology.
Zhang, L. F., & Sachs, J. (1997). Assessing thinking styles in the theory of mental self-government: a Hong Kong
validity study. Psychological Reports, 81, 915±928.
Zhang, L. F., & Sternberg, R. J. (1998). Thinking styles, abilities, and academic achievement among Hong Kong
university students. Educational Research Journal, 13(1), 41±62.
Zhang, L. F. & Sternberg, R. J. (in press). Are learning approaches and thinking styles related?: a study in two
Chinese populations. Journal of Psychology.